 A very good evening everybody. My name is Sonakshi Varma and I along with Rohail Amin, the senior editor exchange for media, welcome you all to the day two of Matic India Bridge One, scheduled from 23rd to 26th of June. A virtual event that will uncover insights about how marketers in the real world will have to rethink what technologies really need, which ones will really help them save money and which ones can transform the businesses that have been altered in the current crisis. The event is a precursor to Matic India 2020, which is tentatively scheduled on 24th of September and slated to be the biggest Matic event in India we have ever seen. It is also the next step towards building a robust marketing community in India. Today's sessions focus on marketing technologies for the new world. Our thought leaders for the day are Jim Stern, he is the founder of marketing analytics summit and has written a book called artificial intelligence for marketing practical applications. Then we have Varit Saurabh, he is the vice president product management and marketing technology manthan. Along with him we have Kashyap Kampala, he is the CEO, RPA2AI and also the co-author of practical artificial intelligence. Then Prasad Ayur, he is the vice president digital e-commerce distribution and rewards at lemon tree hotels. And then we have Alan Pell Shah, he is the founder of deep analysis and co-author of practical artificial intelligence. If you have any questions from the audience, you can put them down on the Zoom chat or you can also use the hashtag martechindia and post them on Twitter. We also have a comment section on our Facebook page where we are live currently and as an Instagram too. Now we begin the first session of the day with Jim Stern. Just a quick introduction about Jim, he is an internationally known speaker and a consultant to Fortune 500 companies and an entrepreneur with over 25 years of experience in sales and marketing. Most of that on measuring the value of digital media for creating and strengthening customer relationships. He is the author of dozen books and on marketing, customer experience, email marketing and web analytics. So Jim I welcome you on board and by the way it is 4.30 a.m. in Santa Barbara, California. So Jim having a cup of coffee right now. Yes. Thank you Jim for joining us and over to you and we will come back to you with questions at the end. Very good. Well it is an honor, very much so an honor to be here. I was excited to be able to join you live. I had hope to join you in person and I send you greetings from well what do I usually say is sunny Southern California but that's not going to be for another hour yet. So good morning from here. So I've had that very nice introduction. Thank you. I started a conference back in 2002 and the audience created the professional association, the digital analytics association in order to keep track of how well are we doing and we've had more data, we have more sophisticated tools and now the most sophisticated tool is artificial intelligence. I don't want to make it very clear that what we imagine as artificial intelligence as we have been told in the movies is simply not the case. These systems are not going to take on human form. They're not going to become sentient. They're not going to take over your job. So don't worry about that. You will have a job and I will can assure you by the end of this presentation you'll understand why and how but do focus on narrow AI or functional AI. Specific tasks that AI can perform are going to be very valuable. This is the to-do list. So for those of you who are just give me the answers, there will not be a quiz. So here are the answers. Number one, what do you want to do in this whole AI realm? What role do you want to play? What tools are you going to use and understanding AI tools is crucial? How are they different? Getting to know your data is essential because the data is so much more important now than it's ever been. You can use artificial intelligence tools to augment yourself and become an augmented marketer. Branding becomes even more important. And finally, well, you know B2B and you know B2C, now we're going to have M2M and you'll just have to wait to find out what that is. So stick around. I want to start with choosing your role. The last company I worked for, which is now 38 years ago, I had a terrible boss. Could not stand this man and I worked. We worked very hard together, trying to get along, not successfully. So I went to his boss and I said I need anything else to do in this company. Love the company. Love the product. Want to participate but can't stand the man. And the boss's boss said, well, what do you like to do? And I was shocked. I had never been asked that before. It never occurred to me that my job could be something that I liked to do. You're paying me to do this. Why would it be something I enjoy? And I thought about it long enough and realized that I really enjoyed giving presentations and educating people. So here I am, 40 years later doing exactly that. What is it that you love to do? What turns you on? What kind of activity excites you the most? That's the thing to focus on. That's where you can apply yourself, enjoy the work you're doing and add value to the company. So what do you like? What do you enjoy? Do you like being in charge? Perhaps you're an entrepreneur. Perhaps you are a business manager making business decisions. But maybe you like the idea of taking business problems and translating them into data problems. That's where the analyst lives. Or you like building very complex predictive analytics models in order to try to simulate the real world in order to make predictions about what might happen. Or maybe you like building things. Maybe you like the idea of building a data pipeline and a platform that other people can use to build their models. Maybe you like to invent algorithms. Maybe you like to get deep into the math. That can be very exciting. Or maybe you just want to explore the new tools. You want to find out how artificial intelligence and machine learning works and dive deep into all of that. These are all available and they're all very much needed. I myself enjoy the analyst role. So as you could have guessed from digital analytics association, I like this idea of understanding the business and understanding the data and trying to make sense out of it with the help of the data scientists. But you have to choose for yourself. Next comes the tools. Now you know the classic statement that if all you have is a hammer, everything looks like a nail. We have an enormous number of tools available to us. The MarTech stack is beyond comprehension. But machine learning and AI fit in a particular area. And I want to explain this historically. You've learned programming, maybe C++ or in my case started with basic and then Fortran and cobalt. Yes, I am that old. And it is just straight logic. You tell the machine exactly what to do and how to do it. And it performs marvelously. If this happens, it will do that. If that happens, it will do the other thing. If it doesn't understand, it gives you an error. And if you've really done a bad job, it's the blue screen of death. So that's just programming straightforward. It's fragile. It requires a lot of attention from the human to make sure it's right. And then we move along to the mathematical model. Now this is where you take values and their relationships and formulas and you create a mathematical model of maybe your marketing budget or your household budget or what sales predictions might be. Yeah, I'm talking about an Excel spreadsheet. It is a different kind of software. You are programming the spreadsheet with formulas for yourself. Then we move on to predictive analytics. This is where we take the data that we have and build a model to try to reflect the real world so that it can assume what might happen next. It predicts. Also very fragile. Lots of assumptions built in requires lots of statistical rigor. And then comes machine learning. And this is a completely different animal. This is where we have the computer look at the data, derive rules and find structure in the data, make its own model from the data. And when new data comes in, it can change its mind. And that's why we call it artificial intelligence because intelligence is the ability to adapt to change. Now machine learning comes in three flavors. Supervised, unsupervised and reinforcement. And these are all valuable for marketing because they do slightly different things. Supervised is where you know the answer and you're teaching the machine with lots and lots of examples. Lots and lots of examples of is this a picture of a cat? Give it 100,000 pictures of cats and then give it a random picture and it will say yes, that's a cat or no, it's not. And you have to teach it over and over and over. Lots of really good labeled data. So that might be these are the emails that we've received from people who become customers. When you see an email that looks like this, we can assume we can predict that person is going to become a customer because we've trained it to recognize what kind of emails come in. Unsupervised, oh, this is a lot of fun. Unsupervised is where you give the machine lots of information and you ask it to tell you something that you didn't know. You say, do you see a pattern? Do you see something unusual? Can you draw my attention to some part of this data that I'm not thinking about? And then reinforcement and this is probably the most practical version where you give the machine a goal and you give it agency. You give it the ability to impact the environment. The goal is I want more people to open my email. So I give it control over the subject line and it sends out tens of thousands of emails with different subject lines and sees which one caused more emails to be opened. It's learning. That is reinforcement. We give it a goal and we hope that it figures it out for us. Supervised, you know the answer in your teaching machine. Unsupervised, it's telling you things you didn't know. Reinforcement, you give it a task and a goal and turn it loose. So why is machine learning so powerful then? What can it do better than humans? It's great at finding correlations. It's amazing at finding obvious things. When the weather is bad, people buy more things online. Well, yes, that's true, but I knew that. So tell me something I didn't know. Well, people who open this email and look at that page and have used that search term are 75% more likely to be customers within 30 days. Oh, now that's useful. That I can use. And segmentation. We have different kinds of customers and clustering. Here's how different behaviors look in the real world and how we might address them as separate segments. And then finally, there's outliers. Here's where, here's something we did not know at all. This is a very strange anomaly. Perhaps there's a problem with the data, but maybe that's a business opportunity. So the machines are really good at elevating things for you to look at, but it's for you to decide, is it interesting? Is it useful? Can I use it to help my business? If I am ranking, if I am sorting, looking for patterns, these are the things that machine learning can do better than other programming and certainly better than humans. In marketing, that means machine learning is very good at A-B testing, for example, or lead scoring. Which leads should the marketing department send on to the sales department? Even meeting scheduling now. There's a tool that will, if we want to get together for lunch, I'll send you an email and I'll copy this AI system and you in the AI system trade emails about when I might be available until there's an appointment set and it gets added to my calendar. Content personalization. I show up at your website. You have seen me a couple of times before. You've seen some of the interchanges through email. Oh, you know what I'm interested in and you can automatically surface the content that might be most interesting to me, et cetera, so these ranking and sorting and clustering activities, finding correlations, that's what the machine is really good at. Then you should use it if you have lots and lots and lots of data well labeled. So advertising, for instance, it is a huge amount of data because you're doing perhaps millions of ads. It is small consequence. If you put the wrong ad in front of the wrong person, it's not like a self-driving car. That's a big consequence. But if you put the wrong ad in front of the wrong person, that's okay. It doesn't cost that much. But the machine learns how to put the right ad in front of the next person. You have good data. It's well labeled and it's not just, it's not the tool for everything. There is a cost to it. So don't just use machine learning because it's there. Use it if you have to. If an Excel spreadsheet can do the job, use an Excel spreadsheet by all means. Then there's the question of what language do I use? And the running joke is if it's written in R, it's probably statistical analysis. If it's written in Python, it's probably machine learning. But you know it is artificial intelligence if it's written in PowerPoint. So you have to decide which language you want to learn. But do learn one of them. And by all means, if you're just a straight marketing person, please at least spend an hour reading Wikipedia about what SQL is all about just so you know your way around data. But do think about solving new problems because this is a new capability. Now, the data becomes so important because we're trusting the machine to find the rules and find the structure and create its own model. So you have to be really confident in your data. And there's lots of ways that data can go wrong. Now as a marketing person, I want to trust my data engineers. But I have to know something about it, just like as a manufacturer of soup. I have to trust the person responsible for the celery and the person responsible for the salt and the person responsible for the chicken. Because if the soup makes somebody sick, I have to go back and find out which one of those was the problem. Same with data. I need a data steward for each data stream. Because I'm going to bet my job on this. I'm going to say, I think this is the right answer. This is the recommendation the machine made. And I trust the data. Therefore I trust the recommendation over and over the top two success criteria are doing the work to build a data infrastructure and a willingness to learn from failure. The first is tactical. The second is cultural. Neither is optional. You have to be confident in your data. And you have to be willing to let the machine try things that don't work out the right way. I mean, learning from failure is actually the definition of machine learning. It tries stuff and it learns. All right, let's get personal. You need to be an augmented marketer. That means instead of just a t-shaped professional where you're really good at marketing but you know something about all the rest of these, you also know a lot about your industry and you understand analytics really well. It's becoming harder to be a professional in any industry and marketing, yeah, we're just layering on more stuff. But we have tools that can help us individually. So I write books and I teach Microsoft Word about my language. These are the words I use. I add it to the dictionary or in my spreadsheet, I'm creating the ability to count or the ability to rank or I write macros. And these are my ways of interpreting that data. I talked about the system that will set up meetings through email. That's called x.ai. It's free. You can try that. You can use something I've used for 10 years, auto hockey to create my own keyboard shortcuts. I mean, how often a day do you type your email address and I just use two keys and boom, it's done. So you can also use Google now Gmail will fill in the rest of your sense. It will guess what you're trying to say and fill in the sentence for you. Your job is to take all of these tools and make yourself an augmented marketer and put all of those on your LinkedIn profile and you become very interesting to prospective employers. And then we're going to brand, oh boy, we need to brand because we've got these voice activated systems. If I ask Alexa, oops, shouldn't say that out loud because you probably, you might have one in your home. If I ask the Amazon Echo to send me batteries, I will get Amazon branded batteries and perhaps Samsung will send me Samsung branded batteries. So the people at Duracell have to make sure that I ask for their product by name. So I have to brand. But then we get to machine to machine marketing, B2B, B2C, M2M. Marketing is going to get a little strange. So I just want you to see over the horizon to see what's coming. Number one, customers own their own data and they store their own data. Right now that's not the case. Right now I'm telling all of these apps and all of these websites, my information, here's my email address, here's my password and here's my shipping address and here are my preferences. I'm going to flip that around. I'm going to store my own information in a secure location and I'm going to give these apps access. So if I do move to another apartment, I enter that once and all of those apps when they need it can be updated. And then I'm going to run my own agent on top of that data. It is my agent. It's not Google, Apple, Facebook, Twitter, Amazon, GAFTA, and it knows everything about me. And I give access to the people and the websites and the apps that I want to. I give my accountant access to all of my financial data and I give my doctor access to all of my medical data but not vice versa. And I let this system work on my behalf because the ultimate convenience is having to do nothing. Right? So today I subscribe to the filters for my air conditioner, the ink for my printer, the soap that my wife likes. It just shows up on my doorstep without me having to remember when it shows up. Oh yeah, that's right. I needed that. I didn't know that. We used to shop and then ship and we're switching that around to ship and then shop. There's this company called Stitch Fix that started doing this years ago. They send you a box of clothes in your size according to your style. You choose the ones you want and send back the rest. They ship, then you shop, then you send back the stuff you don't want. Finally, bonus number seven. You will always be needed. AI is not going to take your job. That's a myth. It is going to help you do your job but you will always be needed for three things. Number one, what problem do you want to solve? What question do you want answered? What do you try to accomplish? What is the business goal? You have to decide what you want the computer to do. You have to decide what data you want the computer to look at so that it can do machine learning but you have to tell it what the training data is and then finally you have to evaluate the output. We can't always trust these machines because they don't have common sense. They don't have reason. They just have statistical analysis. It might give you an answer that is statistically correct but absolutely useless. I call this the smell test and it's something that's incredibly obvious the minute you look at it. This I came across in a hospital recently. It's pretty obvious that that's not installed correctly and any human being looks at that and goes, what on earth were they thinking? That's wrong but you can understand how it happened. Same thing with artificial intelligence. Your job is to perform the smell test. Now I like being between the analyst and the between the business stakeholder and the data. This is the part that's fun for me. Data science, data engineers, some people love that stuff. It's not me. I love doing the analysis, trying to figure out what the data means and make it purposeful and valuable. If you want to be an analyst, it turns out it's not always the person with the best technology. You just have to have good brains. It's not necessarily the person with the best algorithm. You can do it with a spreadsheet. It's not even necessarily the person with the best data or even the best vision of the future. The best analyst is the person who asks the best questions. You learn critical thinking and you learn how to ask good questions and the one who asks the best questions wins. That's what I wanted to cover today. I'm so happy I was able to join you and I'm delighted to take any questions if there are any. Absolutely. Thank you so much, Jim, for this wonderful conversation. A lot of insightful points that you've made. A lot of takeaways for our viewers here today. We're getting a lot of questions. We have just shortlisted three in the interest of time. The first one is how important is social or cultural or linguistic factor in any AI program? Enormously important. The biggest problem that we have with data for AI is that there is bias in the data. The quick example is Amazon had the machine rank all of the resumes that came in to see who might be the most successful in the company. Success was a number of promotions, the amount of raises you got, how many direct reports you have, and so the machine looked at the resumes and chose white males as being the most successful. Amazon said, oh, it's broken, doesn't pass the smell test. So culture, language, these things are very important. Next question is data comes from a large number of sources, your own third party platforms, et cetera. How does one integrate all of that? Painfully. I wish there were an answer. This is a business opportunity. If somebody can figure out a way to use machine learning to help integrate all of those data sets, boy, everybody needs that because right now we do it by hand. It is fragile. It breaks constantly. It's a business opportunity. It's also a professional track. That is what a data engineer does for a living. The market is a nightmare kind of a thing. Yes. Final question is which companies or brands are playing AI game the best on top of the AI game and any that are your personal favorites in that space? Boy, everybody is playing, not all of them successfully. Right. That's an impossible question to answer. It's sort of like who has the best website? Who ran the best ad campaign? There are some projects inside some companies that are doing really well and then the same people in the same company can go to the next project and it doesn't work. So there's no absolute, oh, always use this tool or these are the leaders in the field. There are case studies, one case at a time. People get it right, but it takes a lot of work and a lot of education in making those mistakes and learning how to make it all happen. I wish I could just give you, oh, here's the list, but boy, I'm afraid not. Thank you so much, Jim, for this lovely session and the fact that you're up there at 4 a.m. Now time for the morning walk, I guess. Great talking to you. Over to you. Thank you so much. It's been a pleasure. Now we will move on to the next session for the day. It's a dialogue between Varit Saurabh and Kashif Kampala on AI for Marketing in India. Just a little brief about Varit. He is the Vice President, Product Management and Marketing Technology at Manthan. His focus on creating innovative products that use analytics and deep customer insights into driving management, works on personalization and targeted marketing. He has also helped consumers facing businesses such as retailers, restaurants and airlines across the global leverage, leverage their data better and achieve better ROI for the marketing dollars. Kashif is the CEO of RPA2AI. He's also the co-author of Practical Intelligence, Practical Artificial Intelligence. He is an industry analyst and has advised marquee global brands in retail, high tech, telecom and financial services on marketing, AI and emerging tech. He has been a part of the Thinker 360 ranks and is among the top 10 thought leaders globally on artificial intelligence, customer experience and marketing technologies. Hey, so I welcome both of you today and looking forward to this session, can we start with your dialogue? Thank you, Sonakshi. It's a pleasure to be here. Good morning and good evening to everybody wherever you may be and welcome Varage. Look forward to an interesting session. Yeah, happy to be here. Yeah, thank you. So the format of this is I'll give some, I'll share some opening remarks and then we'll get into a dialogue with Varage. So if you're on Twitter, please use the tag hash martikindia to share some of the highlights from this session. Any questions also you can leave there. So this is the day two, AI day of this excellent conference. So I was listening in some sessions yesterday and Anurag, Anurag Bhattra has made a very perceptive comment. He said, AI is really nothing new. It's been around for 65 years is what he said. I agree with that. In fact, marketers have been doing AI. They just don't call it AI, but they've been doing it AI when you do things like segmentation, personalization, clustering. I mean, marketers are using AI every day without calling it actually AI. But having said that, what's new is that we now have more data than we can handle. And there are also some advances in machine learning techniques itself, so that enables new possibilities and new use cases. So if you look at the defining graphic of the martik industry, which is the 8,000 product landscape of this industry, you'll see that there are probably about a few dozen categories, 50, 60, 70, I don't remember how many, but there are some categories. You'll see that there is no specific category for AI or machine learning. That's because in most of those products, AI or machine learning capabilities are getting embedded or already embedded, so that that's how most of the marketers experience AI through the products that you buy. What I find useful though is, I mean, I have an MBA in marketing, so I come from a little bit of a theoretical background of marketing. So it's useful to think of the four P's of classical, the classical four P's of so there are thousands, literally thousands of use cases of AI, but it helps to look at it from the four P's. So from the four P's of product, pricing, promotions and place. So you'll see that in very many products, the machine learning itself is becoming core to the product or the product itself. When we take self-driving cars, for example, I mean, no car can drive itself in Indian roads at any point in time, I think. But you can see that self-driving features are becoming parking, parking car, parking on its own or doing paddle parking. Those are becoming features of the product itself. In healthcare, it's still a long way to go, but increasingly there is a cognitive software that assists in the diagnosis and provides second opinions. In financial services, you'll see that credit ratings, risk profiles of audiences of customers, they're all increasingly relying on machine learning and so on so forth in product. When it comes to pricing, this is like, I mean, bread and butter for marketers. It's quite mature in practice. You see in verticals like airlines, e-commerce and hotels. So this is, I mean, very well understood, I would think. So when you next P is a promotion and a lot of the machine learning applications that you can buy fall into this category. You'll find like a lot of tools for audience segmentation, dynamic web content generation, targeted offers, discounts, campaigns and whatnot. So here, I think we're seeing some pretty good advances in terms of what's not possible before. The place, I mean, I found it a little confusing to think of it as place, but I think of it as the omnichannel experience is raised. So how and where customers look for your product, how they become aware of your product both across physical stores and digital channels. So AI can significantly shape these customer interactions and journeys, be it in terms of recommendations, ad placements, new channels and interfaces. So that's the context, a useful context to think in terms of for what are the AI applications for marketers. So we've talked about the four P's, but I think in 2020, there is a fifth P, which is the pandemic. So we find that it's still early days, but it's changing consumer preferences, consumer behavior, and ultimately the customer journey. So some of these behaviors are going to be longer lasting while some will be transient and will revert to the original behaviors. We'll talk a little bit about that a little later. So with that context, let me invite Varage. Sonakshi has already given his intro, but he is with Monthan Software as VP of product management, and he creates innovative products using analytics and uses customer insights, helps his clients undermine customer insight to drive engagement, personalization and marketing. So he's worked with a lot of B2C businesses, retailers, restaurants and airlines across the globe to help better leverage their data and achieve the bang for the buck for their marketing dollars. So welcome once again, Varage. So let's kick off this discussion. Let me ask you, customer journey, that's becoming a very important topic. So what's the best way to think about customer journeys? Is there a framework to think about customer journeys? How do you look at it? Thanks, Aishwarya. And thanks for the intro. A customer journey is definitely a topic that's gaining a lot of interest. It started off from a point of view where people were thinking about how a customer would get a huge result when they call a call into your call center. But over time, it has really moved ahead and people are looking at all kinds of journeys from how somebody downloads your app, consumes it for the first time and becomes hooked to it. Or how does a customer engage with your brand in general? Right? At month, because we deal more with a longer term data set where the problem that we are trying to solve is how do you increase the average lifetime value of a customer? The basic arithmetic is the more customer you have, the more money you can get each customer to spend with you, the bigger business you have. So what's the journey that the customer is taking? When they discover you, they discover your products, explore, make their first purchase and then start engaging with the brand after that first purchase. And then finally, you hope that most of them land up in that loyalty bucket where they are very deeply engaged with you and are coming back to your business again and again or becoming a good champion for your business. So for us, customer journey is really this macro customer journey of how somebody moves across the entire chain. And if you think of the kinds of analytics that each of these steps can power, can benefit from, there's actually quite a bit of diversity in what you can do. When the customer is discovering you and you are trying to get them to see more of your website, understand more of your products, you are to some extent dealing a lot more with the marketplace. You are dealing much more in the third party world. Where you are trying to identify customers are most likely to become prospects are most likely to become a customer. It has as Jim pointed out, there could be data science techniques that you're using to sift through incoming emails and identify the prospects that you should really focus on. So starting from there, once they're engaging with you, how do you surface, understand the preferences, surface their expectations before they actually know. So whether it's recommendation engines and so on, that really helps the customer explore and buy. Once they have bought the products from you and then they are using it, that's when you have a lot more first party. You have your own datasets and you can start understanding which segment do they belong to, what kind of lifestyle do they leave, what kind of persona do they have. And then based on that, you can start looking at what sets your best customers apart from the rest of the group. How can you make some of the lower engaged customers move them into that best customer bucket. So that's the kind of analysis, that's the kind of questions you can start asking. And finally, you are hoping to keep as many customers as loyal as possible. So that is where looking at the biggest drivers of churn. When does somebody are disengaged? When do they start buying? What is organic churn? For example, if you are selling a barrel for kids, clearly there's a life cycle after which there will be organic churn and you should be happy about that. You can't do anything to fix that. But how do you identify in organic churn? How do you arrest it? How do you preempt it? And those are some of the techniques that started becoming really... I think these definitely are a lot of things that keep chief marketing officers awake at night trying to crack some of these problems. So marketing is a pretty location or regional specific. You work across both international Indian clients. So is there something unique or different about the customer journeys in India or this landscape that you paint here that that's different? It is definitely... Yes, as you pointed out, each region has its own nuances. India in particular, I have to start with because many of us come with an engineering background. I see Indians have much better words with data and data science techniques than you would see in the rest of the world. So you expect a lot more pointed questions, a lot more educated questions. Well, they are definitely well educated. They do their homework and they are always a step ahead. But the marketing landscape that they work in is actually quite challenging at least compared to the West. One of the reasons is that the market data in India is still in a place where it's getting better. It does not have the same quality that's available outside. So we see a lot more customers, a lot more of our clients focusing a lot more on their own first party data as the primary source of insights and analytics and doing a lot less with third party. That's one big area. And the second one I would say is more around channel engagement. In the West, email is still the king. Email is the channel that everybody has to get right. In India, that's not the case. In India, phones have always been, phones came in before the desktops. So because of that, SMS and mobile apps have a lot more penetration. And with that, the challenge that I see a lot of women marketers facing is that there is a lot less of policing. In the email world, if you send too many emails, you quickly get tagged as a spammer and you lose your reputation. But with SMS that has not happened to the point that there is a lot of channel fatigue that has come in. Most marketers are pulling back from SMS as the primary channel. And I see many Indian businesses doing very well with mobile I think we probably have some of the best case studies of mobile app and mobile app engagement in India than in the struggle. I think that that's a pretty good insight. Smarter clients, smart clients, more engagement on the mobile, not as much on email so far. Data is a little sporty, still not as mature. The data environment, the laws around that also probably not as contingent compared to GDPR. Yeah, absolutely. And the Indian Privacy Act is coming. So I'm sure that could change a lot of things and then hopefully the data quality would get better because there will be a lot more explicit intent from the customer to share that data. So one of my favorite questions to ask all analytics experts, I mean, invariably always ask them because when I hear analytics, people say, oh yeah, we already knew this, it just validated it. So I want to know from you, were there any aha moments in your work? This is very interesting. We wouldn't have known without analytics out of those aha moments. Yeah, actually, there are quite a few. My favorite one was with a client who was doing a lot of email and their challenge was unsubscription rates. They were seeing at least one to two percent of the customers unsubscribe every month. And given that it's a compounding problem, you start losing your customer base, your contactable customer base very quickly. It's the problem the CMO throughout us was to explain or to understand why and when does somebody unsubscribe and what they could do to fix it. When we threw the data into our system and we were running some of those likelihood to churn likelihood to unsubscribe algorithms. The biggest insight for me was that the typical marketers behavior actually makes the problem goes. If a customer is not responding, if the customer does not respond to two of your emails, the marketers first reaction is to send two more and see if that would get that click. But what our data showed was that the days since last purchase, that's the metric that we were looking at. And the longer it has been for a customer since the last one is actually the more likely they are to churn. And we actually saw distinct steps in about 90 days, 180 days when that likelihood increased. So our feedback to them was that if a customer has not bought in six months, you have to really cut down and you have to really reduce the number of communications you are sending to them and just focus on quality at that point. A number of communications are actually hurting you. And based on that insight, they completely changed their journey marketing strategy and that helped them quite a bit in bringing out that unsubscription rates. Another one, which is more of an oh-ho and not an aha, how many of your customers are actually just a one-timer or somebody who has shopped a couple of times but is not really engaging with your brand? And these are some of the automated things. That's me most of the time. I pick up a discount coupon and then disappear after. So while everyone acknowledges the problem, the extent of it is where we saw with one of our Indian clients, for example, about 70% of the customers had just done one or two transactions. And this was when the marketing team saw this insight for the first time. They did not believe it. It was a very classic case of how a group of people go through those five stages of grief, starting with disbelief, the anger, and then finally acceptance. So it's just getting a sense of what's your 80% of your customers are going to be in that very low value pocket. But that 20% of the customers were in your long term and are spending a lot with you are really the ones who are most valuable. So being able to get this insight to the CMO, to the point that he actually made sure that thousands of those customers were called to validate if they have changed their phone numbers or addresses, or is there a data glitch because of which they were coming out as one-timers. And only after that he accepted what we had to say. Let me actually throw a question to you here. As an industry analyst, you have been looking at a number of different industries and how AI is moving. Do you see certain companies or industries doing this well, getting on the AI driven growth better than some of the other companies? That's an interesting question because I've been looking at AI case studies and go to so many sessions, conferences. The number one AI application that always comes back is Amazon. Amazon is the only company and Amazon's recommendations that that seems to be the case. A joke's a part. So I get a lot of calls from tens saying, particularly this happens at the end of the year, saying, hey, we need to have an AI roadmap for the next year types. So maybe we get talking. So I explained, AI is really nothing but a machine learning model application is nothing but the data that you train on. So do you have data? Where are you in your current stage? So when it's like CMO is calling, it'll be like, okay, we need to have AI in marketing types. Do you have a marketing automation in system in place? Where is the data? So I try to figure out discussions. When it turns out, usually people are in various stages. So they are probably, I mean, many of the times it turns out that people are trying to implement a CRM application. I mean, those are just my clients maybe. So we say, okay, let's put the basics in place this year and we'll talk next year. So this sounds like a little bit facetious, but that's what happens. So what I find is that I find that a lot of companies are struggling to implement foundational marketing technology tools. But at the same time, because of the media hype, because of the external and pressure from prop, they're expected to be ready for new applications that are way more complex. Having said that, in general, there are certain industries that are early adopters of technology that invest in new technologies, emerging technologies, innovation, etc. So banking and financial services, e-commerce, to an extent. So and high tech media, the adoption of AI also tends to mirror that pattern as well. So banking and financial services, they are regulated, but they have it still a little easier because their products are not as much physical or omni-channel. It's a digital product we can play around with that a little bit. So those are some of the things that we see. But in general, marketing and a lot of marketers, like you said, I mean, not just in any way, they have a sort of quantitative background, they understand data. So they are able to appreciate what is possible with AI, what is not possible AI, what are the narrow use cases they should look at at this point. So that's what we find. Any specific type trends you would highlight that got your attention? Broadly, this is applicable to a broad swath of companies in different industries. Data is the biggest challenge. Data is the biggest challenge because the systems are siloed. And to be able to run machine learning models, you needed to have collected data in a specific manner. Even if you had data warehouses running for the last 10 years or so, to be able to get that data integrated with other data stores and label it, put it in a way that you can use is very difficult as time consuming a lot of effort goes in. Like I said, data scientists, let's say is like one of the success jobs of the 20th century or the 21st century. But only if you're not doing it, and somebody else is doing it, because bulk of the effort goes into that. So data is a big challenge. Then there is definitely this technology complexity. This access to expertise, there is a certain machine learning way of thinking things to translate business problems into machine learning problems. But expertise problem is easily solvable because they can either hire you or me. We can solve that problem. So that's the experience of one. So in some, it's still early days for using some of the newer machine learning techniques. So that I mean, I keep talking about the what is possible, there is an enormous list of use cases, et cetera. So how do you select what you do? Where do you start in this journey of trying to use AI? The way so the my suggestion to most of the marketers would be to just map out how customers are moving through this journey. And start looking at where your biggest step is, where your customers are, are they buying and not coming back or you're getting the traffic and not the transaction. So once you understand the gap that automatically becomes the, from a business perspective, that's the head at which you have to really work at. When you're looking at the tech stack, as you mentioned, we actually have a, we flip the flow and have a slightly different way of thinking about it, which is we break it into sort of a six layer problem. Obviously the foundational layer is customer data. Then you get into how do you segment customers, what kind of insights are you put into like, how do you measure marketing outcome? How do you personalize? And then finally, what level of optimization are you able to achieve in an automated manner? And so we typically ask clients and prospects to score them on this framework. On each of the roadway, do they see themselves as beginners or level three as a pro? And that sort of immediately starts throwing some light on where their challenges could be. Do they have all the data in place, but they're not using it today and they could directly pick up some personalization tools? Are they personalizing? But a lot of the complexity of the business is not reflecting in the data. As with the pandemic, every organization is now an omnichannel digital first business. So many of the retailers are thinking, how do I take my off 80% of offline customers online? So if your data is not connected, so you might have done a lot of fancy work in online by itself, but now how do you bring that offline data and also connected and get customers to know? It's anomaly when somebody scores them on this framework, that also immediately starts throwing some ideas on the techniques, the tech stack that they should invest in. I think interesting that you mentioned the pandemic, because what we have seen is that I mentioned before that it is changing some customer preferences and behavior. So machine learning is also dependent on the data on which it is trained. I mean, you need large amounts of data to be training this model song. Now suddenly, see if you take recommendations engine, the data on which it has been trained is not as applicable in this current period. So people buy different things, people buy different quantities, what's in the basket is different, why they buy is different. So it's really interesting to see that machine learning models are not as applicable to a crisis situation like this. So that's one point. The second point also to note is that customer journeys, companies are having to change their customer journey. They take a simple example of buying clothes online. So one big problem in clothes is you're always worried whether it'll fit you or not. So that's why you find a lot of easy returns is the hallmark of a lot of online shoppers. But now because of the hygiene reasons, people don't want to return goods. So all of a sudden that that step of the customer journey is broken. So it happens online, it happens in retail also. So a lot of questions related to where is really machine learning models very replicable. How do you realize that this is no longer applicable or it's going back to business as usual when we do that. So there are very interesting questions to think about. Do you have any thoughts on that? So one on the quality of how has pandemic changed the value of analytics? Well, yes, obviously, if you were looking at last five years of data to figure out who your best customers have been, that model is not going to be very helpful right now. But at the same time, you have certain techniques that require very, very literate or most recent data. For example, buying hand sanitizers, nobody has bought as many earlier. But when you start browsing for it, you can immediately make a send, you know, get that sniff that this person is looking to sanitize their house and the cars, right? So that immediately that just last three or five clicks might be enough to tell you what's on mind for this customer in this particular trip that they're on. So yes, there are certain techniques that become less important, more important in turbulent times. But you have to pick the battles and make sure that you're working with data that's more recent and you're trying to solve problems that today's problem and not really a long-term problem. That's what you should be doing in a phase like what we are in 2020. So if I were to summarize what you said, you're saying like AI can also be used to respond faster because in a situation like this. So I tend to agree with that. But that calls for, I mean, being agile, agile AI is what we're talking about. So I want to look at a couple of questions that have come in. So this talked about poor masks. AI can be used in customer journeys. Can it also be used for defining customer journeys or influencing them? So my short answer is AI or AI techniques are mostly used for identifying customer journeys and then sort of making sure that the customer experience that you're delivering matches to that customer journey that you identified. Yes. So earlier, I think Jim pointed out that when you have unsupervised learning techniques, you're throwing a lot of data at the engine and you're asking the engine to find a pattern. So that is those are the techniques that can help you identify the journeys that the customers are on because you are not really thought about what journeys the customers are making and are going through just throw the data and get some longitudinal analysis of how most of the customers are performing. So in that sense, AI or analytics in general can help you identify what your dominant journeys are. And then reinforcement learning techniques can actually help you start experimenting and tweaking and making sure that you are impacting the journeys towards so much more favorable. So definitely different tunes for different answers. Thank you, Valerj. Thank you, Kashyap for this lovely discussion. We have a lot of questions, but we are short on time. Maybe we can take it offline and then get them answered. Thank you again for this lovely conversation. Thank you. Thank you. Thank you. It's been great learning for me. Thank you, Kashyap and Varaj. We now move on to the next session for the day, which is a success story, CX and hospitality and Maatek. We have with us Mr. Prasad Ayer. He is the vice president, digital e-commerce distribution and awards at lemon tree hotels. He is a digital and e-commerce professional with a master's degree in business administration and e-business marketing with over 17 years of experience in senior management positions across digital e-commerce and e-business. He has taken proficiency and leading efforts to provide key insights into development of digital marketing and e-commerce, digital transformation projects, online sales, marketing strategies, the venue support and implementation plans for online media campaigns. He has a proven track record of outstanding accomplishments in diverse competitive, high pressure and evolving environments. He is responsible for Indian hotel company. It's digital initiatives including digital marketing, social media and ORM, web and mobile platforms, digital transformation projects and he is also the head of the digital center of excellence to enhance guest facing as well as associate facing technologies. So over to you Rohail and Thank you Prasad for being here. We had a chat earlier also. It's something I was looking forward to. When we talk about customer experience, of course there are demands and challenges in every sector but in your sector especially, customer experience is not compared to, it's a different ball game altogether. I think the expectations are not higher in this hospitality industry and which has been now impacted and you are trying to recreate those experiences through technology which also means there is no verbal communication. The aura is not around. All that warmth is not around and yet you have to make sure you deliver on those fronts. Tell me first of all, how is your industry dealing with this demand that tech has to solve these situations in front of you? That's a wonderful question to start off with. So again, thank you for having me here with this panel and the hospitality industry is always sort of overlooked when it comes to tech. It's also a chicken and egg story. It's largely because if you statistically look at tech adoption in general, hospitality and travel has always been the slowest adopter of tech in general. Largely also because the business is such that it's very capital intensive. The operational experiences or optics is very high and most of it is towards keeping the physical structure up and running. Service delivery is largely to do with people so the cost also sort of largely is full of wages for instance as an expense and that's largely also because the industry such as service-based. Having said that, the industry that's hit the worst or hit the hardest through the pandemic has been travel and hospitality. Our business is largely also dependent on the airline. Thankfully for us and the company that I work for now and represent on this fall with lemon tree hotels, we've been managing at about 25 to 30 percent occupancy even now, even given the current situation. It's largely because we work closely with the Indian government on many things with whether it is with the Ministry of Health with regards to asymptomatic quarantineers, whether it's business continuity plans that a lot of tech companies use for, whether it is medical staff and support services that use us through the course of the lockdown as well. We've been able to sustain this business very well and invariably we've also sort of gone a little lean with regards to how we spend. The industry in any case keeps its digital spend largely from a B2C or a go-to market strategy perspective. But in the larger scheme of things, I think where most of the spends are coming now are largely from companies trying to sort of future-proof themselves as opposed to sort of really looking at conversion-based investments in the current scheme of things. Does that answer your question? Absolutely, absolutely. Tell me, this quick adjustment, this nimble footedness that is required to adjust to this scenario and make your customers secure. From the tech side, can you tell me what lemon tree is doing? What are the new technologies you have in deploying to make sure that that customer connect remains, you know, that that site is developed and it is contemporary, it delivers the kind of solutions that you want Another interesting question, but let me, I mean I'll give you the easy version first. I'll just tell you what it is that we as a company sort of are doing and where the aspiration value is and where we intend to take this. For starters, about a few months ago, thankfully we set off on setting up the new CRM system. I heard Kashyap and Virat speak a lot with regards to sort of setting up the base and largely knowing your customers a little better and the dearth of data, if I may say so. We still need to sort of run a proper KYC because that's where we are. We are still in a position and I would sort of very comfortably assume that that's where most of the other companies in the hospitality and travel space are as well, largely because the volume of the database is such. We may have five points out of 20 that are known to us but we may still sort of not necessarily have the other 15 points that requires us to sort of set a profile and then personalize or customize because I think words like personalization is sort of thrown around quite a bit in there and in order to sort of add the cherry on top and then say it's nice a little bit more, I think people speak about AI and personalization in the same breath. I think we're still far away from it. With regards to sort of, I think step one would always be to sort of have a strong backbone CRM platform. Step two is to sort of really get all your data points and complete and thorough KYC done for your customers. Their likes, their dislikes, their preferences, meal preferences, vegetarian, non-vegetarian, allergies, so on, so forth, smoking, non-smoking. I think the list is endless because ideally when we have that is when we can sort of really make use of it. Preferences are something that go a long way in the travel and hospitality space and it's something that we expect the company that we choose whether it's an airline or a hotel company to know that about us. We often hear things like, oh you know I stay with you a dozen times a year or I stay with you every month. You should know this by now. The problem and the reality of the matter is not every company has got that right unless it's probably about 0.5% of your total platinum guess who might know it and that's about it and that too requires a whole lot of on-ground service delivery. So what we've done is we've invested well in a good CRM system that's starting to fire on all cylinders but unfortunately we've had to sort of put a few roadblocks in there largely because we want to sort of do this when it's safe to travel again. Let me put it that way. So while you are putting the CRM in place, what does the front side look like, the customer facing side? For example I read that you know people can now when once they reach hotels they don't have to go to the desk, they will be checking in directly, they will be ordering contactless meals. So what does the customer facing end look like in a new CRM? It's again this is ideally where reality meets fiction and I want to sort of really put that into perspective as well. There are companies that I've worked for in the past that have sort of piloted this to a massive extent. In fact I've been a part of some of these pilots which include the keyless entry, web checking, mobile checking and so on and so forth. The idea is these pilots at that point in time which was the pre-pandemic time, the adoption rate was negligible. People still prefer the touch and feel key card, they still prefer to go speak to the hostess at the front desk, they still prefer to sort of have someone usher them around to the room and so on and so forth. But I think the pandemic and the post-COVID world is going to be slightly different. I'm still optimistic or call me pessimistic if you must but I see a lot of this going back to the way it was. I think an important part about hospitality is that smile, that greeting that is ideally what makes it all better, that justifies the price tag, that justifies your arrival experience, that justifies the pre-stay, post-stay, in-stay and all of it. I still don't see that going away completely and that will eventually come back and I hope it does for all of us. But realistically, Keyless Entry requires us to do a whole lot of things. It's not as simple as just creating a mobile app and figuring out the property management system, which again is a big capex investment and to have to change, let's say, Lemon Tree Hotels today has 8,000 rooms, 8,000 keys, that's 8,000 keys for us to change locks on. It's not as simple as just changing the RFID on the lock and not as simple as just creating a mobile application that can talk to that RFID lock. It requires a whole lot of things. It also sort of threads on the lines of security, for instance. Even I think identity proof and photo identity is something that's still required. There are companies that have actually done this. We piloted this ourselves, but is it easy? Is it going to sort of be done overnight? Absolutely not. It's going to take time, but I definitely see that having its own niche and its own takers definitely over the next 18 months and even more sure over the next three years or so, I see this becoming a reality across the board. But in North America and Europe, we see the usage being quite high. But also I think that that depends on the kind of customer service that customers in let's say our part of the world expect versus customers in North America and Europe would expect because for them contactless might be okay. I say this with experience because I've traveled extensively during my time with Marriott. These parts of the world, we like people opening doors for us. We like people taking our bags up to our rooms and so on and so forth. But I don't think it's the same in North America and Europe. So I think this is definitely the need of the R, but whether all hotel companies have an appetite or putting that level of capital expenditure given the current recession that we are in, I think it's a matter of time, but it eventually comes there, but it's going to take its time. Right. One side of the conversation is that the solution being offered by hotels and the industry to customers. The other side of the conversation is the acceptance of the solution by customers. Absolutely. That's also important because as you said earlier, they like people to open the doors. They like to be served to be greeted. I think they're looking for it. They're not coming to a robotic place. Tell me why you're offering this solution. What has been your understanding of the customer acceptance of this solution? Are they okay with it? Do you need to build it further? Are you just keeping it as a temporary arrangement? What is it like? So like I said, we haven't gone full board with this because currently while it's still on the drawing board for us, but as a customer myself, I see myself using some of these products. I see myself using a hotel company app that I use very frequently, but this is largely again only going to be to that subset of customers that use your services or use your hotel more than let's say 12 times a year or maybe four times a year even if that was the case. But in terms of adoption, I still think that people will want to sort of stay conventional because I think we tend to believe things that we can touch and feel as opposed to things that we don't touch and feel good for instance. But in the largest scheme of things, key cards and things are sort of slowly getting replaced. We are trying a few things with regards to sort of trying to sort of find an alternate solution to sort of having as much contact less service delivery as possible. Like I said, there is probably only a handful of hotels in this part of the world in India at least that give you contactless service. It's largely because adoption rate is another story altogether. I remember when Bombay had its first metro, Reliance Industries that set this up here had to sort of deploy about a thousand people across just seven or eight stations just to show people how to use the token system. It's because it was overnight that it was open and people weren't really sure even how to use escalators for instance. So I think this needs to be administered. Although it seems simple to most of us because we are active users ourselves. But I think bringing usability is something that I think is also another form of data that we will need to eventually connect, collect, filter, siphon and so on and so forth in order to sort of really figure out and think do we really need it. Because want is one thing. I think everyone wants something or the other. I want a Ferrari. Do I need a Ferrari is a different story. Do I need a Ferrari in Bombay is another story. Can I afford a Ferrari is a later question to be asked. And similarly with mobile applications with chatbots and so on and so forth. Call all of these things martech solutions if you may. I think wanting something is very different to justifying the need for it. You need to have the data to substantiate the need. Just because we have or a particular hotel company has 350 restaurants doesn't mean that everyone is going to download your app and start ordering the food online. But I'm fairly certain that this particular company might have sort of run a survey something as simple as a survey and saying let's just go out and ask 3,500 people out of the 35 like members that you have and saying if we were to put an application which allowed you to order food from the comfort of your home would you get it. Similarly if we were to do that that's the approach I would take because that prevents you from plowing that limited capital expenditure you have into something else. And we all know that tech in hospitality or tech in general depreciates much faster than physical assets. The shelf life of an application is only going to be limited to the time that it's actively getting used and then eventually there's time for upgrades, time for rebuilding it, repurposing it, not to speak about the operational expenses or the marketing expenses that one needs to sort of plow in in order to convince users to actually start using the platform. So the spiral continues and I think the appetite is not just for today it needs to be there for the next 5 to 10 years at least. What you put out today is ideally going to create a foundation for you for the next 5 to 10 years. What we are doing in the current scheme of things thankfully we do have time on hand to sort of divert our energies towards finding auxiliary revenues but also towards future proofing ourselves and saying what is it that we really need versus what is it that we actually want. That question needs to be answered before we really start talking about where is the business going. So would you agree if I ask you that when it comes to a mottic solution to hospitality there's a limited solution that you can offer. You cannot totally say market can provide a solution unlike other sectors maybe for example an FMCG or other. Do you think that would you agree with this that yes it can provide solution but maybe for example loyalty engagement connect but it cannot be the alternative it cannot totally solve the problem for the industry. What Anruil I think you got it right in fact I think that was the rhetorical question for me you've answered it yourself. Yes hospitality again is a very touch and feel service I think people like to talk to people people like to meet people and so on so forth. So let's just take an example bots for instance right is that going to replace customer service for royalty or in general it's not in the hospitality space. If it was a very binary product let's say a static device like a mobile phone for instance or a pen or or anything that does not have too many moving parts in it it's easy. I mean I think Martek has a whole lot of plug and play solutions that can solve for it but you know anything like booking a stay to a certain extent yes but beyond a certain point I think we'd have to draw a line because I think questions that come away usually are more people centric and more from the heart than they are necessarily from the mind. Now I think binary systems can sort of usually answer everything that comes with a yes no or a one and zero but anything that has the concept of maybe in there which is I think you need something that can actually reason with the customer right and then if a system was able to reason I think we've all collectively identified and created artificial intelligence as one may call it but in the largest scheme of things no in fact we spoke the other day and even this afternoon the best example is VFSI I heard was was a good adopter of fake. I tried to renew my own auto insurance day before yesterday with a particular insurance company I'm not going to name them. I called their customer support system because there was variance between the premium I had to pay versus the premium they suggested I had to pay. I called them their IVR suggested that I go to the website I went to the website their website said that I can do this easily on their app I went to their app their app eventually introduced me to their chatbot their chatbot couldn't even generate an OTP and it said why don't you call us and that was like I just came through this entire circle and you're sending me back there you know what I'm just I'm getting a little tired and a little frustrated I'm going to sort of go conventional call an agent tell him to find me the best insurance paid that was done in 15 minutes right in the largest scheme of things and and what is important here is not that they didn't have a tool they had a tool I think they had all the good intention of having a tool in the current scheme of things where having people at their call centers may or may not work but the problem is because they didn't sort of have all their ducks in a row or they didn't have a safety net or they didn't have a contingency plan they've just sort of lost let's say a lack worth of insurance premium right they've lost the customer in the largest scheme of things I would look at that as the opportunity cost because it's not the investment that that I'm so worried about at this time but what worries me more is loss of revenue now I can't sort of expose my loyalty customers people who really like my hotels they swear by it they love everything about it it has taken us quite a lot of effort time monies to acquire this customer now I'm not necessarily going to risk losing that customer to a machine because the machine couldn't speak a particular language or the machine couldn't reason with it or the machine simply wasn't available right so with having said that I would much rather save that 20 or 30 thousand rupee per license and plow it towards another person right because I think a person would be able to reason reason with me and saying you know what we are sorry even even something as simple as giving me a real apology would would have would have resolved it but it doesn't so in many things Royal to answer your question no not really I think hospitality will continue to have people although I mean there are a few companies who who do a good job at it married for one does a good job with their boss but that's largely for employment and and yeah I think that that representation is good perfect so so we are all waiting for this normal to come back and of course it will come back we don't know how soon how late but it will be there once it comes back what will be carried from this phase to that new normal I mean what would be the addition in the offerings that you had before COVID once we go back to the normal stage what will stick to us what would be adapted by the hospitality prayers especially for example you're coming no that's a that's again a good question and this is something that my you know our managing director Mr. Bhattakeswani strongly talks about because he's probably the best version of a human Reddit you will ever find out there he pre-selects a good amount of readings for us and sends it to us everything from from you know pre-pandemic to post-pandemic and what how we will sustain going forward there's a whole lot of cost-saving measures that we as a company have taken through time you know these are not necessarily related to cost savings by downsizing the human resource of it it's we're not talking about that at all we're talking about basic expenses that we would otherwise take for granted right lemon tree as a part of its sustainability initiative does a whole lot of things as well right well I mean you'll always see these little things out there with regards to how we conserve water how we conserve electricity how we conserve fuel how we conserve energy energy as a whole a lot of good learnings have come in here do we again this is again want versus a need piece we would want all our customers to sort of have 21 degrees but the ministry of health otherwise tells us that you know what 26 degrees is what you have to set it at so some of these these learnings have come in as a result of a lot of trial and error and and the necessity itself so in the largest scheme of things speaking about tech we realize that we do not need to spray and pray anymore right I think Kashyap and and Viraj spoke about emails still being a big and a and a very good source of communication in the west in India unfortunately it is more a spray and pray mechanism which we don't believe in we believe in doing tactical stuff which I don't need to send out 1.2 million emails to 1.2 million 12 like 12 like members that I have you know I might be able to sort of do that with 400 000 for instance a lot of these little little things we sort of looking at people's conversion ratios open rates so on so forth all sorts of manageable data that we can find given that there is a scarcity of data in fact if I can just share my screen I'm currently running a bivariate test on three three of three versions of my website because I want to figure out which version works best but do I have enough data to sort of tell me that option a is better than option b versus option c it currently does it right and I only sort of did this now because I didn't have slides to share but because I wanted this to be a more open conversation but yeah in the largest scheme of things we're trying to do this do as much as we can given the current measures we've taken to sort of really control costs but like I said we're also sort of building this strong CRN system in the background just so that we are future ready so when it's time for all of us to travel again and I hope that that's really soon largely with a lot of selfish interest for the industry as well I think you receive communication which is more suited for you as opposed to just a simple spray and spray mechanism that we're more accustomed to so I think we'll retain a whole lot of these cost-cutting measures they're not cost-cutting actually they just call them efficient measures for for lack of a better word right right great I mean of course I think it's going to be more specific more tailor made the communication now I mean it's going to be the offerings are going to be very very different than it was earlier I have time for one final question if I have to ask you personally as the leader as somebody who is a tech evangelist who understands technology or from the hospitality sector what what have been your learnings from this phase and what would you as a person you know like to suggest advice to people who are in the hospitality tech side I'm I'm I have a strong belief with regards to having hope right I think it's it's important that all of us continue to have hope you can't give up on anything you can't give up on yourself you can't give up on the business that you're in because I mean I'm saying this from as as a hospitality professional who's probably hit the hardest right personally and in general I think all all all my peers and my colleagues in in this space would sort of agree that we are very passionate about what we do we love it we love it you know we we like the concept of what what hotels are about what resorts are about the whole idea of having to get on an aircraft and go to another city and sort of stay in a particular hotel have stories about that there's a lot that goes into it as well but you know what what most companies and I'll tell you that what what my team and the marketing team at lemon tree hotels is currently strongly working towards is largely come life coming full circle and us still trying you know we're trying to sort of fulfill maslow's basic hygiene factors right we're sort of still fulfilling the lower pyramid of maslow's theory of hierarchy where we still have to convince people that you know it's it's safe we have to feed the customer's mind as opposed to his heart and his soul right aspirational marketing is something that a lot of brands within the country luxury brands do this very well you know they they they they try and and sort of send that message more to your heart and your soul but I think what's become more important is sort of keeping it real making making sure that we are feeding enough information that is good enough for the mind to make a decision with regards to his choice of brands products hotels airlines whichever but in the largest scheme of things I think we find ourselves just conveying that message across we have our own safety and hygiene program called rest assured where we partnered with one of the largest sanitization players on the planet because we know that people will need that and that's still ironically still a hygiene factor even if it's it's as simple as that so I think we I'm done telling people to sort of stay safe in fact the only thing that I want to wish everyone under the on the planet today is to start traveling safe because I think we have vested interest in everyone traveling again and hopefully traveling safely and and making the right choices and still taking care of themselves thank you for this wonderful conversation and unfortunately like we do in hospitality I can't shake hands with you but yes waiting for that look forward to it and thanks again for your time likewise thank you very much for having me thank you Prasad thank you real we move on to the next session which is the last session for the day we have with us allen fell sharp he is the founder of deep analysis and co-author of practical artificial intelligence and enterprise playbook allen has over 25 years of experience in the IT industry working with a wide variety of end user organizations like FedEx the Mayo Clinic and Alice state and went us from Oracle and IBM to startups around the world allen was formerly a partner at the real story group consulting director at Indian services firm with broke he was also a research director at 451 and VP for not America at an industry analyst firm open he is regularly quoted in the press including the world's free journal and the Guardian and has appeared on the BBC CNBC ABC as an expert guest welcome allen thank you very much I'm happy to be here I am actually having a few problems with zoom right at the moment but don't worry we'll get it figured and we'll get started with or without slides so video going on here yeah so hi there so whether I have slides or not don't worry about it we'll get going so it's been really really interesting listening to the sessions this morning and I think they blend together rather well we talked a lot about AI a topic that's very familiar to me and in fact I'm the co-author with cash up who spoke earlier of the AI playbook and I think there was some great stuff there but rounding sort of out here what I wanted to talk about today was really the human element and the practical elements of customer experience digital marketing etc because I think we live in a world where it's very very possible and very likely actually to think that it's all about tech and it's definitely not and the last session there and present we talked very eloquently about how technology only plays a part so if we go forward here and as I say I'll play in the background here and try to get my slides going but our focus typically is always on the concept of delighting customers right customer experience it's all about making it wonderful for the customer and we want everybody happy and that all makes perfect sense that's all a wonderful thing but here's the thing your most unhappy customer is likely to be the one who gives you the most trouble they're going to have a lot more impact on your business than anybody else and here's the thing if we focus on always delighting our customers giving the best possible experience nurturing leads nurturing prospects keeping going that's good but you're going to have very few loud evangelists for your product or service and even those who are really happy with you sell them are going to bother to leave a review or even possibly recommend you they're not going to go out their way to do that some will and that's a wonderful thing and you should definitely do your best to to look after them but those who you really annoy will likely go out of their way to post a bad review to not recommend you to tell people to avoid you so when we talk about customer experience management when we talk about you know the whole concept of digital marketing and enhancing a customer experience we have to remember that experiences are both good and bad or indifferent right so the reason i'm bringing this up now is not to be just sort of miserable and to give you all a hard time the reason i'm sort of bringing this up now is to really express how important it is at this time with so much political upheaval with you know COVID-19 everything's going on not many people are happy many people are looking for targets for their anger and their frustration so handling your customer experience today is different that's for sure i think we can all recognize that and moving forward you need to have some kind of strategy for how to deal with yes this current crazy situation that we're living in but also to become more adaptable as we move forward and that's really where things like AI which has been talked about a lot this morning and is a topic very close to my heart can play an important role but i think it was cash up who actually pointed out that well it's only as good as your data and the data it's getting now or any AI or machine learning system is getting now is somewhat skewed maybe even very very skewed so what we can't do is lose touch with the customer i'm sorry allen allen sorry just one your video is completely off your screen is blank can you just okay i will refresh it is that working no we cannot see the screen is completely black brilliant that's wonderful well isn't that great well you know you're not missing much there is that is that any better yes yes we're back okay i just switched it on and off again twice so but that's a that's a very technical thing maybe you can try it now maybe you can try a screen sharing now the books okay let's at least try i'm sorry about this is that coming up no no i'm sorry well it's not happening i'm afraid on this end is the video coming back up yeah you are back okay okay well i've got a nice new shirt on so that's what we'll have to do i'm so sorry there's been some problems with zoom this morning the sound's been dropping in and out throughout for me but we'll we'll do the best we can with the time we've got so anyway the the long story short you need to start thinking about a post pandemic strategy and learning from this and figuring out what you're going to do in future we talk about the new norm we've no idea what that that will look like and we don't know what what twists and terms are coming around the corner so the thing is um you know i'm a technologist um i'm an industry analyst a technology market analyst um whatever we want to call me um but i also trained as a psychotherapist so i'm a sort of an odd breed here and if there's one thing i've learned is that um technologists certainly hardcore technologists really don't like to talk to people very much um that's sort of part of the dna here and i probably fall into that but talking to people is really key to figuring out where to go and what to do um i think if we take a step back here one of the things which i've been recommending to people for many years and and it seems very obvious but i have to say very few people do it is to take a sort of 360 degree view of your customers so if you're creating a journey map for example which we talked about this morning a number of times um yes you can do that with data and that that can certainly inform it um and you can talk to your employees and they will give you some kind of input but what you really really need to do is to talk to your customers um and yes surveys are good they're a great start but do talk to people actually get their input and the reason i say this is that in in a number of instances i've worked on personally um we've had situations where and i'll give you a real world example here there was a company i worked with who was really quite successful um they were set up to move forward spectacularly i think with some new investment and everything was going swimmingly well they were absolutely convinced that their customers loved them um actually their customers did love them um because they had the best product slash service in the market we undertook a 360 degree sort of you know um analysis we talked to the employees of this company and we've got a good understanding of why their product was so good and so better than everybody else's and then we talked to some of their customers and here's the thing not a single customer said that this company had the best product had the best service they didn't say that and where the company was absolutely convinced that they were winning against their competitors it turned out that hardly any of their customers had even heard of their competitors the reason people kept going back to this company was because of their customer service their customer service was outstanding they loved this company it was so easy to deal with yes they had apps yes you could go online yes you could do this but whatever they had it nailed when they took a call they handled it brilliantly and an interesting thing they were really good at admitting when they couldn't do something when they were wrong i mean they were just a very open honest company and that was the reason people were going to them i get that as an example because every company is a bit like that even my company i think people you know come to us because we have the best research and you know whatever what our customers actually think of us well we have some idea because we have actually bought this talk ourselves and you know i won't go into it i'm not trying to sell anything here but the bottom line is it was an idea opener for us as to why people were actually coming to us and why people wanted to use us and it was very very informative and we put that into our marketing the thing is um when it comes to being honest when it comes to being open when it comes to being sort of you know responsive and analytical about what one is doing in your company today very few companies are up for that um there was a slide i would have loved to have shown you um but i can't from Forbes which did um just a couple of years ago did a pretty large survey and it was asking people how many people in you know how often in your company do you deal with challenges and how well do you deal with challenges when when problems come up in your company how well do you deal with it and the results were absolutely damning um the bottom line is maybe 20 percent 25 percent of companies actually actively and you know proactively dealt with challenges most companies like to deny it most companies go into point finger pointing etc so here's the thing is we move forward into a digital marketing world and you know i say move forward because not everybody has got this figured out and i think that's an important thing to take away from this conference and from everything that's going on in the world is if you think all of your competitors have got this figured they haven't it's not true um you can look at any company in the world um you know my work and i think the people on on the the conference today who've been speaking typically deal with larger organizations that's supposed to be the benchmark that's supposed to be the people who are setting um you know the the mark for us all to to reach um they've got it all figured out they know what they're doing they analyze their customer data not true not true at all i think they're all trying but that's a different thing i mean if we take one thing just as an example from the technology standpoint just about everybody has a crm system today now that can be as basic as some kind of repository where you put your customers names addresses email addresses etc and of course it can be much more sophisticated than that you can have a lot of interesting information about your customers transactions and their likes and their dislikes and you know what they're actually doing and how they interact with you but regardless of how sophisticated or not um that is very very rich data that should be informing and working with your digital marketing efforts even in the largest organizations that's actually seldom the case in most large organizations is a serious disconnect and if we take a step back and out into the the broader world if you like and i'll just give it this is an example because it's very relevant today if we take the world of oil and gas for example you know tons of money very very important to the world's economy you'd think they would have a lot of this figured out they don't and they can't predict demand they can't tie that to production which is key to what we're all doing whether that's in the hospitality industry whether it's in the healthcare industry you know we want to have some kind of predictive insight as to what's coming so that we can adjust our business to adapt um they don't they can't they can't do it um now is the data there sure um can they actually leverage it and make use of it no and that's most of us right we have the answer somewhere and to go back to what kasha was talking about earlier and our very first speaker uh jim um saying very very eloquently you know we've got to start using data much much better than we have in the past AI machine learning automation are all coming in uh even at the smallest level and that's something i want to stress even if you're a 10 person organization versus being a 10 000 or 100 000 personal personal organization the technology is available there for you today it's very low cost technology anybody can have a crm system an email marketing system some kind of campaign system if you like anybody can do social media um those things are there for you anybody can have them but anybody can do them badly too so whether you want to really invest your money heavily um and you know go all out or whether you want to do it on a very very very tight budget the principles are going to be the same what you can't do is separate your customer experience into a digital and into a non-digital world that's got to be tied together and i think this is where the big challenges come up if we take that a step further we have a big challenge within organizations again whether this is small organization or whether it's a large one what we have is the fact that people are motivated by different things right people are motivated by different things um whether you are um a CMO whether you are um somebody who's actually in the call center um whoever you are in an organization you have different um different perspectives different requirements so how do you actually set cx customer experience priorities yeah and i think this is just sort of the practical stuff i want to share with you because again i don't care how small you are as a company how big you are as a company these principles are the same for everybody right you know so there's this basic things here you know yes a website yes multi-channel right now multi-channel sounds very impressive but at the end of the day typically we're talking email tech social chat mobile etc right now you may leverage some of those you may leverage all of those CRM and whether it's a very simple system or whether it's a complex one everybody needs that right and that needs to be augmented and really cared for most people don't augment it most people don't care for it very well so you need to really treat customer data very very importantly and and invest in that as best you can but you also need to tie that back to your organization what are their goals what are their strategies what are you really trying to achieve here so start first with sort of a voice of the customer yeah so yes customer journeys very very important um you know that's a session for another day really uh how to map those but voice of the customer think about that is it you telling them what they want is it them telling you what they want right so you need to be thinking about the voice of the customer how to yes make them happy but be how to not annoy them and i'm going to go into that in a little bit more detail in a second so always give very high priority to customer data and to analyze that and again analytics can be at any range of the the budget spectrum from you know very advanced machine learning to frankly doing it manually the one thing that a lot of people really don't take into account is processes this was sort of the mantra of 20 years ago we had to analyze a map uh business processes and tasks and everybody knew how to every business person at least knew how to draw up a flow chart and to make sense of it and say ah this is how business works of course it wasn't actually how the businesses worked it was how they wanted them to work because we know that there's many workarounds and realities human beings aren't binary um they don't say yes or no they say yeah i think so um so again you know we need to start understanding our business processes but not to nail it down not to nail down every activity in a perfect way because that that's frankly not going to happen what you need to understand is what are the regular tasks what are the repeatable tasks what happens the same way pretty much every single time those are the areas where you can apply some automation and again an automation can be a simple little set of rules or a macro if you want to use that term or it can be complex business process management you know again AI machine learning intelligent process management etc but those are the ones you focus on right you know the things which are going to happen the same way every time never try and get beyond that because you never will right it's a corny phrase but you know computers at the end of the day even machine learning which is incredibly adaptable which is incredibly powerful um still at the end of the day they are binary they're zeros and ones yes and no black and white and people are not like that and they're never going to be like that right so automate the things which happen the same way every time and really put your your efforts into managing the exceptions right they're the ones that are going to hit you as you move forward so you know focus on getting that together in your organization and sort of your organizational readiness and there was a great example from present just just previously um on his experience with an insurance company and in one of my slides which i'd love you to see but of course you can't for some reason um i had an example which again it's a real world thing and i it's from healthcare actually but again these these this kind of example i'm going to give you here is pretty common across any industry now healthcare in the us is not like healthcare elsewhere i'm based in boston i've lived in the us nearly 20 years but you may be able to tell from my accent i'm originally from britain and you know whereas we see that as a public service as most writers do here is business and very very big business at that um you know if we take an example in healthcare and this is a this was a real world example this this company had really put a lot of money into building a wonderful website an absolutely fantastic mobile app i mean honestly um this this nail this is this is a award winning what they did however didn't work and doesn't work okay because what they have is an incredible digital experience i won't name the companies they've worked with but think of the biggest digital marketing companies and in fact one of the biggest digital marketing companies actually uses this as one of their case studies apparently it's wonderful and it is until you actually use it and it is until something actually goes wrong and here's the thing do you know what goes wrong paperwork the app works great the website works great and it says yeah you need to fill in this thing and send it afterwards and that all works perfectly but what's happening at the back end the paperwork gets stuck in the same old routines it's always been stuck in and what's interesting here is the frustration you would have had with a manual process of the past so let's roll the clocks back 10 years um you want to have some procedure or whatever you fill in some forms in the US and your insurance company approves it and etc etc you knew there were people dealing with paperwork and you sort of had patience for it frustration up to a point but you were relatively cognizant that there was a miserable person having to deal with your paperwork so you gave it a week or two what they did is they built such an amazing web experience that nobody had any patience they wanted it to happen instantly absolutely instantly hey I filled this in where's my result what why why can't I move forward this is this kind of situation which is actually repeated in a lot of digital marketing situations where the digital marketing has gone out there it's done an amazing job however the organization as a whole has not actually embraced everything it's set now false expectations we've gone from customers who are very patient excuse the pardon in healthcare but customers who are very patient and actually put up with things people who are happy to give a call and they say oh yeah we're working on it it will get you to in the next couple of days people are perfectly accepting of that situation to a bunch of customers now who are not accepting of it at all so be very careful it's great to spend a fortune on digital marketing you can bring in the best people in the world you can get your message out there but be careful of the expectations you're setting everybody doesn't need to be delighted everybody needs to be satisfied and there's a big difference and if you go from five to 10% of people who are unhappy with you to 20% that is going to have a massive impact not an extra 10% impact that's going to have tenfold right you can finish your company that way pretty quickly or you can seriously damage it at a very very least so again um you know start your CX initiatives by first understanding your customers in your own organization but never ever shy away from the good the bad and the ugly and make sure you've got a human element tied into it and with that I'll leave a few minutes and wrap up and I apologize for the problem with the slides I everything's clicking on this end but it's clearly not coming through on yours thank you Alan for this wonderful uh points that you've raised very enlightening and great listening to you we do have a lot of questions but in the interest of time maybe we would take them offline with you we'll send it to you I just want to remind our viewers that tomorrow the day three of the series would be the theme for the day three is let's talk customer experience I think that's what people want to now they're obsessed with they want to understand what's happening in the customer experience space so tomorrow is day three and Alan if you get time do join us we will be glad if you could be there listening to our conversation and so with this we come to an end of day two of the series which is a precursor of the big event that would happen in September hopefully we'll find the vaccine by then thank you everyone for joining us today it has been wonderful it has been a wonderful session and thank you Alan once again see you tomorrow all of you thanks again thank you