 So, today's presentation is going to be a little bit about basically the art of data driven marketing. So, I'm going to talk to you guys about, and this is a little different I know from the typical sessions here at Dribblecom. It's my first time, so I apologize for being a rookie at this, but I'll do my best and basically try and help you guys understand a little bit about our approach in data driven marketing. Who am I? I'm originally, my name is Gus Murray. I'm originally from Australia. I'm quite a long way from home, as you can see. I actually live in Copenhagen, Denmark. That's where I work and basically work with FFW. What else can I tell you? I basically focus on digital strategy and in my spare time I like to do ultramarathons. And for those of you that don't know what an ultramarathon is, it's basically races around 250 kilometers in far out regions of the planet in Mongolian deserts and Jordanian outback essentially. And recently I'm actually from a small country town from Walgert. It's called basically in the middle of the outback. So I grew up on a farm, so I'm a very, very long way from basically my farming roots and also from my actually birthplace. So without further ado, any of you guys, I work with FFW for anyone that missed the basically the drys keynote or missed the huge blue stand in the middle of the actual session where FFW, which is the combination of pro people and link reaction and a few other agencies at the same time. Essentially, I won't bore you guys with all the stuff we do, but essentially what we do is and focus on strategy, websites, mobile, social media and optimization. And what we work with and especially what I work with is the digital transformation of some of the clients we actually work with. All right, so to the interesting part, what I want to talk to you guys today about is the shift from basically creating business cards to actually creating business tools. And this is some of the, I think this is a very passionate area for me about how we actually change the way that websites work and what they do for our users and the end users, basically. I think our industry has been guilty of making very pretty brochures online, but a lot of them aren't delivering business results. And that's one of the things that I'm really passionate about and I hope to share with you guys today. Firstly, though, I want to talk a little bit of story about something I read in Wired Magazine. So two companies, one is Planet Labs and one is NASA. You're probably familiar with NASA, I'm guessing. Planet Labs, for those of you that aren't familiar, is basically a bay area based startup. They started around 2012 and they create basically microsatellites, which they call DOVs. And these satellites are really interesting and their approach is especially interesting. For example, NASA, as you guys all know, they launch massive satellites into basically outer space and they do this on a schedule of probably once every 10 to 15 years, especially in the global imagery market. And these are huge satellites. They're massive in both their scale, the scope of the project and the number of people and resources they commit to actually building these. On the other hand, Planet Labs, I don't know if you can see in the background photo there, they create these microsatellites, which are basically the size of a shoebox. And they started in 2012 and since 2012, they've actually launched 70 different satellites into the actual space. So their approach is extremely different. And the way this came about is they asked a simple question, what if you were to put up a simple inexpensive low flying craft instead of a big expensive one? Because what they worked out is that a satellite essentially is made up of the same components of a smartphone. So the same basically things that you have in your iPhone or your Android, which is an accelerometer, a CPU, some storage, a camera, and basically some sort of interface they can update. So they, instead of going the usual route, which I think we can all relate to in terms of building websites and that's the thing, is the requirement driven waterfall approach where you start with the sort of objective and you waterfall all the way down and eventually come out with a pretty website, you know, three or four years or in what we talked about in Drupal 8's case, it's taken four years of development to build this, you know, Drupal 8 basically release. So, and this is some of the challenge that I see faced with our industry. And this is why, you can't probably see it in the background there, but it's a picture of the Mars rover. The satellite and the way that NASA approaches things is very likely to work, but that's how you get a two megapixel camera into the Mars rover. And that's a, you know, a multi-billion dollar project which has technology which is basically outdated by several years now. And that's some of the challenge we face versus the guys at Planet Labs which have a much more iterative approach to actually how they design and build their essentially satellites. And they release, recently their whole aim is to have 150 satellites launched into space of which they'll basically have essentially a daily scanning or imitering of the entire planet which is completely different from the NASA approach which takes around two weeks for you to actually book the satellite, readjust the coordinates, basically have the imagery taken and then you wait another two weeks to have anything else done. Whereas this is on a daily essentially 16 hour time frame. So enough about, and this is where I think the core of what I really think is the shift in the way we're sort of building and approaching our business problems or our clients' problems in terms of building new and more agile development. And these guys in Planet Labs, they call Agile Aerospace which is basically the design, build, test, repeat sort of methodology following on from the lean. So I'm going to talk a little bit about data-driven marketing. If you have questions at all, please feel free to ask. I'm not afraid of getting anything. So what is data-driven marketing? So this is a term that I think we've heard a lot of lately. It's one of the big buzzwords that's followed along from big data and everything else. But this is essentially I think one of the things that we can really harness into how we actually develop products and solutions for our customers going forward. We've seen a lot of sort of things change in the last couple of years. I mean, I'm not going to bore you guys, most of you know this. The cost of processing storage has dramatically reduced. Tool costs are way more inexpensive now than they ever have been before. We have the ability to actually measure stuff that previously in our industry we haven't been very good at or we haven't measured at all. So we have these different things that now create this opportunity for us to actually make basically products and services that our clients can actually use in a way different way than they traditionally have. And this is my whole thing I think about moving from the business card approach to the business tool approach. Like how do we create actually business tools online rather than just pretty brochures? I think this sums up pretty nicely. We must move from a numbers-keeping score to numbers-driving actions. We've all been very good at looking at essentially a nice scoreboard where we get a monthly report of what the traffic is to our website, some of the basic click-through rates and things like that. But essentially it's just vanity metrics. It doesn't do anything for us. Apart from making certain people in the marketing department actually look good because our vanity metrics are all going up. That's essentially all it can do. What we want to do is actually dig deeper than that and actually understand how can we use this data to create actionable insight. And part of this we've seen plenty of times before. I mean, for those of you that know Facebook, I'm pretty sure everyone's familiar with it, with Facebook ads, for example, you have the ability to go into basically, you know, go from a demographic to a sociographic to psychographic information where we can dig down to, you know, different interests, what people, education levels, their relationship status, all these different things that we haven't actually been able to do before. Typically the approach on a website is that one size fits all approach. It doesn't matter who you are, where you came from, what you do, we give you the same front page or basically a variation of a couple of landing pages and that's essentially all you'll get faced with. And essentially, you know, if you look at, there's hundreds of examples out there, basically, you know, where you have individuals which aren't dramatically different from their actually demographic sort of stereotype. And that's, I think, one of the key things. How do we actually improve delivery and improve conversion? And that's one of the key things I think we have mostly in this industry. So one of the ways we look at conversion is with this formula. Essentially, conversion we see it makes up or is made up by one motivation to actually purchase or download or, you know, generate a lead or something like that. We have the value proposition. So what is the value proposition I get? Is it, you know, the quality of the content? Is the speed, the efficiency, the product offering? And then we add, basically, the incentive less the friction. So the incentive might be, you know, reduced cost, it could be a sale, anything like that. And the friction is basically the jumps or the hoops you make me jump through. Essentially, you know, do I have to fill in a form? Do I have to, is it a gated piece of content? Do I have to use, you know, PayPal versus my preferred Mastercard? Things like those things. And then minus, essentially, anxiety. Anxiety can be in the form of, OK, are you going to steal my data? Are you going to, you know, is my credit card information safe and secure with you guys? So those different elements is basically what we see is making up the conversion. So how do we handle that? And I think what you prefer or what your designer prefers doesn't matter if it's not getting your conversions. This is the key thing that what it comes down to. We see a lot of this in the agency-client relationship where, you know, our lead design on projects like, oh, I really like the color red or the pixel over here and the client on the other side is like, well, I want it in blue over here. And basically it comes down to who speaks louder or who's paying the biggest bill, essentially. So, and that's not really a great point of, you know, departure. What we want to do is say, OK, what does the user care about? What does the user actually want? What matters to the user? And we've seen a lot of examples. I mean, Google has been, you know, famous for doing this on their front page and how they eliminated a lot of stuff. And what we can see now is our clients are taking on different form. They're essentially becoming this sort of Dr. Jack or Mr. Hyde sort of element where we have, you know, one part artist, which is about the content, the, you know, assets, the social media sort of aspect. And then the other part, they're part scientist, which is, you know, about the performance tactics, the operations, the analytics, you know, and that's what they're looking at all the time. And we have to play to these sort of elements. So it's making sure that we sort of keep both sides of the client happy. I think we, who, does everyone know Scandinavian Airlines? Or does everyone not know Scandinavian Airlines? So Scandinavian Airlines is a regional airline based out of basically Scandinavia. So this is our major carrier. These are the guys I love to hate. I love using these guys in essentially examples because essentially their view of me is, I think, pretty typical of the way a lot of companies deal with their clients. So this is, I'm one of their, I'm a fairly frequent flyer for SAS or Scandinavian Airlines. So they know me relatively well. I have certain destinations I fly to on a regular basis. There's certain, you know, seating arrangements I have, certain types of flights I take, and certain times of day, everything else. But to them, when they send out information based on my behavior and everything else they've collected, it looks something like this. So I've got a, basically, first of all, an offer to a castle. I've got a nice coffee machine, some cutlery. Then I go down to, you know, I've got watches and pots and pans and wines and candles and women's jewelry because that's what I'm probably interested in. And it keeps going to dressing gowns. What else do we have? Like bed linen, everything else. So their approach is basically, if we throw everything at this single person, then maybe they'll click on something and then in the hope that they'll actually buy one of those things, which is not a great result or approach because now, as you can imagine, when I get an email from Scandinavian Airlines, what do you think I'm going to do with it? Delete it. Like, I'm not going to spend 20 minutes reading through all their offers to me because I don't have the time of day. Then I've also been to Tokyo once before, but then they're actually offering another trip to Tokyo again. So understanding my travel behavior and things like that is not really that well done either. And then they changed language from Danish back to English because they've worked out I'm not actually Danish and I actually speak English. And now they're telling me about how I can go to New York. So their communication style and the way they have personalized is extremely broken. They have so many data or so much data around my behavior and my activity with actually Scandinavian Airlines, but they haven't used any of it. Basically, the only thing they've personalized is my name at the top and my membership number. On the other hand, I'm guessing you guys all know about Amazon. So Amazon are really, really good at this. They've been doing this for a long time and they're one of the best in the industry. I wouldn't pay too much attention to this. I'm really fun at parties, parties contrary to popular belief about my choice in books. But they're really good at understanding my behavior, my purchases, what I would like to buy, other things that other people that have bought similar books to me have. So they really have a different approach to this. It's very, very personalized. And the results speak for themselves. They're extremely successful. So for me, it's looking at the journey, not the destination. Like we have to stop thinking about, okay, how do we sell, sell, sell, more about how do we actually push people or actually get people to go along the journey? Instead of this whole lock-in mentality, how do we get people to lock on because of the value we provide and the personalization we provide? And this, I think, is done through, we have to change our perspective of, the dot-com isn't the only place that people are actually using these days. Like, we have a huge amount of different touch points that we as consumers interact with on a daily basis. So as a brand and a client, we have to understand this sort of network effect and take advantage of it and use this to actually inform us on how to actually better engage with our clients or customers. I think this is an example by Rand Fisken from SEO Mods. So he's made the very, you know, sort of dumbed down version of how we see the web, me make content, humans click, then we buy or then we get money, basically. This is the way we've traditionally seen sort of essentially content and websites, where in actual fact, it looks something like this, like me make content, humans click, if I'm like, they remember, maybe they see more content I make, visit again, me build trust, relationship with humans, when they need me, product them, come back. Like, this is more about how it actually works. So try saying this very fast as a good tongue twister. But this is more how the actual web works. Do you know what I mean? We don't go to websites and then just see it once and then buy the product or service that they're selling. It's often a longer journey that we see. It's more interactions. It's building that relationship that we sort of talk about all the time. And this is the whole element that, you know, as I said before, we often have one spot where people come into our website and they see the front page and it doesn't matter what stage of the customer decision journey they're in, they see the same information. Whereas what we're trying to do now is look at how do we actually break up this customer decision journey and deliver small pieces of information that match where these guys are in that stage. So if they're in the consideration phase, it's gonna be very, very different from someone that's in the actual purchase phase. They don't need to get to know your brand anymore. They know who you are. You know, they've established that relationship and now they're interested in actually going through ahead and purchasing the actual product or service. I think this says, you know, you need to understand who these people are in the business world and where they are in that buying process. If they're very early, you don't want to serve the messages of content low in the funnel. So this is a key part of basically understanding, okay, where are people in the actual customer decision journey? How can we target them and what type of informational needs do they have at those different sections in that journey? So this comes down to actually what is the strategy? What is the action plan that you're gonna go with? Here's a good example from, you guys should be familiar with this one, NFL. So comparing this to the Scandinavian Airlines example, these guys have done a much better job at understanding just a small fraction of what I like. Do you know what I mean? They understand my team, they understand my name and they know it's my birthday. So straight away, they've actually made a win or a small win based on very, very small piece of information that they can essentially get in the public domain. So this is a much easier and personalized approach. So my communication relationship with the NFL is gonna be way different than it will be with Scandinavian Airlines. Another great example, this is from one of your Nordstrom, one of the big retails here. Has anyone seen this example? So basically what Nordstrom do is they basically put this small plastic tag on their most pinned product in the women's basically shoe section. So it's a really clever way of understanding, okay, the offline world and the online world are heavily integrated. So this is a great way for women in the market for purchasing red stilettos to understand that there's some sort of social proof attached to the purchase of those shoes. So if they purchase those shoes, there's a certain degree of safety that they know that when they go home and show off to their girlfriends, they'll be like, okay, that's the great pair of shoes. I also like those ones. So we have to connect these two worlds because we know that a lot of actually transactions actually happening in the offline world. So how do we actually bridge this gap between the offline and what's happening online and actually help consumers understand that and connect our social activity and our basically dot com activity. Another great one is, or this is an example of, sorry, where you can actually integrate some of the different elements of essentially great features like Facebook into websites. Facebook are just launching their Messenger app basically for businesses. So this allows businesses to actually integrate the Facebook Messenger directly into the website. So you can actually have a dialogue with customers straight away in their actual Facebook app. So you're not actually forcing them to go outside, come back to your website, find the help section, find the contact us section. It's basically within their native social environment of Facebook within the Messenger. Another great example is KLM. These guys are great at basically, the way they approach social and digital. For those that haven't traveled with KLM or seen their social channels, the really interesting thing is the cover photo right at the top. I don't know if you can read it, but it basically is the number of minutes that they'll take to essentially respond in terms of customer service. And that update, they update that cover photo on Facebook and Twitter every five minutes, based on the actual availability of their customer service agents. So they're actually integrating and connecting the social world to their actually customer service platform online. And actually integrating that in such a way because they know their users so well that, okay, if you have a flight issue or anything like that on, when you're stuck in Amsterdam or whatever, you go to the Facebook page or the Twitter page to say, where's my bag or where's my flight? And straight away you know that, okay, they're gonna respond to me within five minutes. So it's a really simple way of integrating and actually connecting those two. The other, and I think these guys are really good in terms of, I won't spend too much time, but the way they approach their social and integrate this is extremely interesting, even from a commerce perspective. I think last year they generated about 52 million euros through their social channels. And I think this is the opportunity that as essentially marketeers and helping clients understand is understanding what we can actually do on social and driving that back to actually the main platform. How can we use this information on our main platform? Here they ask the simple question, do you like the physical boarding pass or do you like the digital boarding pass? By understanding the users and which users have responded in which way can help them serve their customers better when they go to their website. Look at what you can actually get, for example, you have the email interaction, you have website activity, you have purchase history, and then you have the people information as well. Basically all this information is available in the public domain where you can actually customize and tailor essentially the experience that these customers are having on the website based on this information alone. So through their email campaign, where do they go, what do they click on, how do we actually tailor that and use that information to better serve these guys? And we've all seen this, I mean, it still amazes me in 2015 that a lot of companies aren't using social as a login essentially. I mean, Scandinavian Airlines, who I really love talking about, they still don't do this. They don't use essentially social plug-ins to actually help complete forms or anything like that. And that's just one aspect of it. The information you can gather by actually doing this is extremely interesting. You probably can't see it on the form here, but essentially what they're asking for on both LinkedIn and Facebook is access to your connections, access to your, especially in LinkedIn, your professional connections. They can also tell that, okay, if someone logs in with LinkedIn on their website, they're most likely a business traveler. If you log in with Facebook, you're most likely a leisure traveler. So those two experiences are very different based on what you log in with, essentially what platform you log in with. Then they can customize the basic products and services they offer you based on that login credentials. And then when you connect the other accounts, they can create that sort of single customer view of the actual customer, which creates a much more rich and dynamic experience. And this basically helps, and this is one of the things we talk about with strategy framework. So what assets do you have? What are the goals? What is the hypothesis? How do we test it? And then how do we learn and repeat? How do we actually make sure that what we're actually looking for, we can actually test in a way that actually isn't gonna take us four years and a huge multi-billion dollar process, that we can do this on a regular, lean and sort of agile methodology. So this comes down to the objective targets and KPIs. Does anyone know? I'm guessing everyone who hasn't been hiding under a rock for the last couple of years knows Psy and the Gangnam Style, basically song. Many of you may not know is that actually Psy actually broke YouTube. Before Psy's actually Gangnam Style hit, the number of views that you could have on YouTube was only in the hundreds of millions. But Psy actually broke that and created two billion actually views and it's still counting and going up. Which helped basically meant that YouTube had to basically calibrate the whole view counter. The challenge we see with a lot of clients is that when we talk about essentially setting objectives and targets, everyone is focused on the total views. They're focused on, okay, Psy got two billion so therefore I should be able to get at least 100 million. But that's the wrong approach because what we want to look at is, okay, what is the rate of change rather than the total number? So how can we actually look at this and say, okay, what affected the change? And unfortunately I wasn't working with Psy on his song but if we look at the steep growth curve, like what were the contributing factors that actually helped make that go viral essentially rather than worrying about the total number which is essentially meaningless and we're never going to actually hit it. How do we actually understand what is the rate of change affecting the growth and how can we target that and how can we hone in on that? So the key thing here is not to decide what to do but what to do first. What we try and do is look at, okay, how do we create a data framework? Because as you guys know, there's a massive amount of data out there at the moment. So how do we actually make sense of that and how do we actually use the data to create a framework that helps us make decisions more intelligently? So we use what we call the core methodology. So this is about the customer experience, the opportunity, the revenue and the expenditure. So these are the four things that we go through to actually determine how we use the different data sets. So customer experience, you can see here is basically the different elements or the potential to actually, you know, positively impact on the actual customer experience. But the opportunity is the amount of users affected basically. So we can actually start going through and actually eliminating different options based on this. And then the potential for basically generating essentially revenue, what is the impact going to be? And then the expenditure in terms of time and essentially cost. So what is it gonna cost? If we have to completely do a new re-platform, then it's gonna be a long time and gonna have a huge cost. But if we actually do a small AB test or a split test, that has different implications and different costs. Then I think the next thing we've seen, which is I think still really badly done in our industry is implementation of analytics that actually measure stuff. I think we've seen too long that actually people are just inserting the Google code or the Google script and then going, okay, it's in there. We don't have to worry about like, how do we actually implement this stuff to actually make sense? For example, one of the ways that we approach some of the problems is through these three tools. These are basically, for those of you who don't know them, relatively cheap and free in the case of Google Analytics. So what we do with Google Analytics is to identify the problem sections of a website. So we go through it, maybe sections that have high bounce rates, it may be sections that we can see the user is not either scrolling past the fold, maybe where there's challenges in terms of downloading. We can look at different, basically parameters here. Then we use Qualeroo, which is basically a behavioral-based pop-up questionnaire. So what we can do is based on user behavior, we ask them different questions. So to identify, okay, why didn't you purchase? Why didn't you go back and add that to your cart? What stopped you from buying, reading, whatever it is? So that gives us another level of information. And then we use Optimizely to test the hypothesis. So those three tools individually aren't that interesting, but when you combine them together, they become extremely powerful as a way of understanding how the user is using your basically digital assets and how you can actually change and adapt them to get better performance out of them. And then it comes down to basically optimizing. So using sort of a lean methodology of build, measure, learn. So how do we actually, you know, building something, putting it out there, putting a measurement framework around it so understanding what to measure and then learning from it and then going through that process again. And not making this a huge cumbersome thing, but actually doing something small. The number of times I've worked with clients where, you know, we've had agencies that have come in and said, okay, we're gonna redo the whole platform. We're gonna, you know, launch date and we have that very hard fixed time. Then they, you know, switch the new place to the website on and the new one crashes. The old one breaks down and they have a huge drama. What we wanna see now is actually where we actually designate the new website to a certain percentage of the traffic. I mean, Facebook has been doing this for years. They test new features on their users in different markets that are representative of larger markets. And we're still not doing this in most of the way we approach basically development and the launching of new websites. Now I think this is a huge opportunity for us to actually use the data to learn and become more intelligent about how the users are behaving, what they're doing, what they're not doing and then incorporating that into the design. And so we don't have a fully fledged, you know, 100% working website, but we actually have something we can build and learn actually design on and top off. And this is just one of the ways, I mean, this is just the way that we approach it, you know, through defining the strategy and then basically executing on the back of it. So understanding, defining the problem, creating the vision of the plan, then exploring, using the build, measure, learn framework to actually iterate on the back of it and then basically create the solution. There's hundreds out there, it's not to say this is the best one, but it's just having that basically process of going through and helping people understand, okay, it doesn't have to be 100% finished when you actually deliver on it. We can actually deliver something that actually is 80% finished and then iterate on the back of it. As long as the core components are there, we can then start learning and understanding what we can actually use. All right, so enough with all the theory, I'll go into use some cases now. So one of the things that, you know, what drives earned media, traditional or social, this is often a common like challenge we see, okay, how do we actually generate like traffic or views or shares or downloads or purchases? Is it by our earned, basically approach or our owned or social? One platform, Kiva, which is a micro financing platform. One of the things they did was basically do an analysis of okay, understanding where all their traffic was coming and what traffic was actually doing on their site in terms of generating sales and repeat sales. So you can see traditional earned media was generating around 900 new sales, around 400 repeat sales versus a blog post was around 90, new sales, 63 repeat sales, then they had a community forum post which was 99 new and 48 repeat sales. So based on this, this was basically a monthly different element. It was pretty obvious that traditional earned media was generating the majority of their actual sales. It was the one that was driving the business. But what they didn't, what they discovered by actually digging into the data was that actually what appeared to be true on the surface was actually not true in reality. So what we looked at was the blog post had a huge influence on traditional media coverage. So most journalists, I hope there's none in here in the moment, like a relatively time poor. And so they don't have the time to go out and develop all these new stories. So if you can actually give them content that is interesting to their readers, they'll use it, spit it out and change it a little bit and push it out into their audience. So by using some of the blog posts, they could actually generate more articles based on that. What was driving the blog posts was the community posts, which basically were the key drivers for developing the blogs. So actually the thing that drove the most sales was actually the community forum. And the community forums, as we see them, some of the first things to get killed in new web projects because we don't see the value in them. They create a lot of noise, but we don't see the actual value in actually generating on the back end, especially for a lot of brands and commercially. So here by just digging down into the data, it was very apparent that by supporting the forum, it would drive more blog posts and therefore on media and that helped them understand, like how do we actually generate more sales? So what's one of the biggest predictors of human behavior? Anyone want to stab at the answer to that one? What's that? Past behavior. Past behavior, yeah, somewhat, that's true. But weather, I mean in California, you guys are blessed by having like amazing, sunny weather basically all the time. In Scandinavia, sunny weather is a scarcity rather than something we don't take for granted because there's not much of it. So it has huge impacts on our human behavior. I mean, and I think it's sort of true, probably more of the East Coast than the West Coast where you guys have sun all the time. But one of the things that we looked at is how can companies more easily and more intelligently take advantage of this publicly available information? We had the whole snowmageddon happening on the East Coast this year and last year I think they had a pretty tough. So how can companies, I mean, for all the females in the audience, what happens when the humidity level in the air goes up? What happens to your hair? It basically messes with your hairdo. So there's a lot of opportunities where companies now can actually take advantage of this publicly available information, use it on their website and actually promote different elements to the target audience, basically live while the weather is actually happening. We saw it basically on the East Coast during the snowmageddon. There was a brand that was actually taking advantage of essentially this information by offering winter jackets. They have a special sale on winter jackets every time basically the snow reached a certain amount of noise on Twitter and those different social platforms. Then they use their website to actually deliver on this. So by integrating what's actually happening on the social platforms back into your .com and using publicly available information from the weather bureaus, like I'm using this in more intelligent ways than just having standardized approach to actually how we deliver content. For altering behavior, we've seen this a lot in the whole gamification. And some of the best guys are doing this are basically the VIP or the actual loyalty programs. This has been happening for a long time. We've moved from the loyalty program where you had like the physical stamp with the coffee store to the actual more digital one where someone like JetBlue is actually using this as a way of basically upselling different seats. They know that have these seats available in the planes by actually rewarding their users with these different elements to actually use their points and upgrade them. They also collect information about their users. So by them using their points, it gives away certain information about what they do, their preferences, their number of pets, how they like to actually sit or travel. And that information they can then use and then reserve to those people based on different marketing techniques and basically presenting that information. Sephora are also really, really good at this. So Sephora have different, basically levels of loyalty program. I'm not sure how many of you guys are members, but what they do is basically, they give away free products as part of their membership. So they give away on their website, free samples or free actual products. The interesting part is that they allow the members to actually pick which products they would like to take home for free. The really interesting part about that is that that gives them an indication of okay, if they're choosing the lip gloss or the facial moisturizer or whatever it is, they can then use that information to better target those users when they next come back to their store. They know what offers to send them, they know what products that they like, they know what brands they have preference towards so they can completely tailor and customize their basically web experience. So it's changing the way they're actually doing their business by simply giving away a small item on their web store. So for the cost of lip gloss, they're getting data that they have to do focus groups and a lot of testing on basically for free, essentially for nothing. So it's a very, very simple way that brands have the ability to actually tap into this and just by activating these simple data points can completely customize and personalize the experience that these users are having on their website. And by doing this, they're locking in that loyalty. They're becoming way more loyal than they have before and they're building that relationship with their users. So how do you get people to share a brand that most users prefer to keep secret? So how many of you guys actually work in B2B businesses just a quick show of hands? So in B2B, I've worked with a lot of shipping and oil companies and places like that which have extremely uninteresting and unsexy basically content or areas of business. But what they don't realize is they're sitting on usually gold in terms of the data that they can actually create which is really interesting for the users. So for those of you that don't know either Pornhub or Fifty Shades of Grey, I won't go into the details, but Fifty Shades of Grey Pornhub is a really interesting product for several reasons. I won't go into those details. But what they have a challenge with is that no one wants to talk about porn in the public media space. No one, they're not gonna put New York Times isn't gonna allow Pornhub to advertise on their platform. So what they figured out is that actually if we translate data into infographics and make it shareable, then people will be willingly share the basis of the brand all over the web. So what they did, I don't know how many of you guys know Pornhub have this, basically part of their website called analytics. It's Pornhub.com slash analytics. Don't ask me why I know this, but I'm into this. But what they have is they basically created these essentially infographics about the behavior of people during the Super Bowl and also of Grey basically when they released the movie. So in terms of the Super Bowl, you can see the different usage. So I'm not sure if you can read it, but United States is the blue line. So obviously when the Super Bowl was running, most of the men and women for that matter were preoccupied with the actual football going on and less occupied with the porn going on. But obviously when the football game finished, back went the essentially porn viewing habits. And you can see that Canada is in the middle. So obviously they care a little bit about the NFL, but they're still strongly influenced by their porn. Whereas in Europe, there was a very minimal dip. So there's a few people that cared, but most people went about watching their porn still. So understanding these different things and this infographic was published basically in all the major media sources, like Mashable, TechCrunch, like some of the great media publishing places on the web were pushing this out. And in the top right hand corner, you had U-Porn and the Porn Hub basically. They were basically promoting their brand through data. And I think this is a, not suggesting that we go out and make pornographic infographics, but the case is still interesting. If you have a B2B company, you have massive data sets sitting underneath you that you can actually take advantage of and utilize and push into the public space to take advantage and increase the exposure of your brand. And a lot of this is basically sitting in the user behavior and activity on your own website right now. And that information and data is basically being wasted or sitting in a database somewhere, not being utilized. And the same with the movie, like this is in terms of BDSW searches and the difference between male and female. So you can see the light purple one is female searches. So that's the pre-launch of the movie versus post-launch of the movie. So you can see the increase in female searches of BDSMW, like actually skyrocketed after the actual movie launch. So this is an easy way to actually, get that content out into the web. Another one is actually, how do we actually understand user behavior offline? We've seen a lot of use of attribution modeling. We have all these different other tools that we have for online, but we know that majority of transactions and actually goods are actually being purchased offline. There's a lot of things in terms of, we've seen loyalty cards, which I just ran through, but also iBeacon. Does everyone know iBeacon or heard of iBeacon technology? Roughly. So iBeacon technology is basically these small physical installations you put on the wall. They're very small. They have sort of RFID tags inside, or sorry, light Bluetooth tags inside them. And every time you walk past with a mobile phone that has an app that is connected to them, a message can pop up and actually engage with the person. So whether it tells you about a special deal with a store or there's different elements it can actually promote to you. But it's a push-based sort of technology. And I'll show you an example of something that's happened in Finland recently. E-commerce is killing brick and mortar. Realty investors spawn to disagrees. If e-stores are winning because of their efficient use of data, why not take their best practices and use them in the real world? Introducing physical cookie. This simple key chain begins the new era of loyalty programs. No apps, no cards, and absolutely no sign-ups. Simply put it in your pocket and you're good to go. It works just like cookies online. When the person carrying it enters the shopping center, the system begins gathering behavioral data, learns from it, and then shows targeted offers and messages. In other words, it makes the retail space react to the customer's shopping habits and interests. Sponda was the first in the world to launch this system at one of Helsinki's busiest shopping centers, City Center, and branded it as the VIP Key Loyalty Program. Over 14,000 people were given these VIP keys and instantly they became part of a loyalty program. While conventional loyalty programs recognize their members only at the cashier, we do that the second our loyal customer walks in the door. With physical cookie, Sponda created an entirely new way to combine real-time marketing and a loyalty program, making the shopping experience smarter than ever before. So I think this is a really interesting approach. This is used in a retail or a mall environment, but you think about this from an educational perspective. You think about this from whole different way of user sets in terms of companies and different aspects. There's a whole way we can actually start integrating this technology into the way we actually serve essentially clients and customers and users through these different technology points. I mean, this is just the beginning. Lastly, I'll tell you this case, you can't see the image, but what we're starting to do now, we're working with clients now. We're actually based on the data that we're doing or receiving from the users is determining how we actually develop the site. What I was talking about before about working with data and actually launching sites. We have a client called Mouse Glyne and they're focused on two different markets. One is the United Emirates and the other one is UK. So we're actually creating a website which is completely unbuilt basically. It's a small percentage of the real build and we keep developing that based on what the user experience is on that actually first few land pages and we're building out the website based on the data that we're getting in from those users. And this is only a small percentage of their global operations obviously. So it's not their main.com but it's a sub-segment of that.com which is based on IP address and based on these different rules and behaviors. So this is just a small indication of what we're starting to do with this data to actually help us decide, okay, how do we develop, who do we develop for and what type of information can we put into our site experience to create that personalized experience. I'll leave you with this quote by basically Amazon. So this is, if you live in Chicago, Amazon doesn't have to know it's snowing but they know your neighbor just bought a shovel and eight pounds of salt. So with that, I'd like to say thank you. Any questions? I would say the first tool with, if you use something like LinkedIn, obviously you can complete, depends on how you want to use it but if you're using it to complete a form or a login tool to get access to a special area, you can then ask the users to what do you call it, like complete their essentially profile. So once they've logged in with something like LinkedIn, you can then go and say, okay, we've got this piece of information, is this correct, is this how we should communicate with you and then you can validate the actual email address if you need to. But often, I mean, it's not so much the email is interesting obviously with LinkedIn but a lot of the other information is the more interesting part, like who are they connected to? What is their job title? Where are they working? Those different elements to sort of help target them because the email communication is just one aspect but the other information is, I would say, much more valuable. It's a great question. I mean, and I think there's no real good answer. I mean, I have a lot of data on, okay, and that's a little bit about that friction I was talking about earlier, like that friction of like, I really want to log in using my social accounts and I know I'm going to give up some information. What are my, how are they going to use that? And I think the best way is to test it. I mean, you can log in with your social accounts, you can also log in with your email or something like that and also just test that friction and then do a split test where you say, okay, let's push 50% of the users will use social login. The other 50% will actually let them log in for free or without giving up information and that way you can actually test. A lot of the data I've seen depends on the value proposition you give the users. So obviously if it's a site you don't know, you have no idea about what they do or you can't trust them, then the friction will be much higher than a site you do know because you know the value proposition will be much higher because if they use something like discuss, for example, the commenting software or plugging, then you're familiar with that. So your friction will be much less than some other. But I think the best way is to test it because you and I could guess for a long time but I think it's the coming down to actually testing it out and seeing what works and split testing and that sort of thing. Yeah Chris? Yeah. I think there's a, so the question is if you don't have the resources of someone like Amazon, for example, how do you actually do some sort of email personalization? And I think the easy part is just by simply doing small scale sort of steps, basically the NFL one is a classic example. So that individual is logged in with either Facebook or some sort of social network. They've completed their, basically profile, they've given away like their birthday, their team preference, and that's pretty much it and their first name and they've been able to personalize based on that. So even if you have nothing else than just understanding the analytics behind, okay when we push a email to our people that are on our Facebook community, we can see they behave you like this. They typically click on the first link or they click on articles related to XYZ and then they have this behavior on site and then we can use that information to better target them through our email campaign. And there's a lot of cheap effective sort of marketing automation tools out there that can help with that as well. But very simply, even something like MailChimp has effective ways of dealing with this information but just little things like, okay, what did they click on? What titles worked in the email? Like what information did they sort of go to on the website? What behavior did they have? Just using that alone, you can get a lot, let alone if you have their first name and birthday. It's pretty easy. Any other questions? Yeah, Zach? So one of the image I showed before with the data-driven rapid development, that client, what we're doing is basically we're building a new site for them. And we're going through the process of exactly that. How do we actually incorporate that sort of build measure learn into the building process? And what we've talked to them about is actually instead of the starting with the end goal of this very pretty website that our design team came up with at the beginning of the project, sold it in, we said, okay, what are the ambitions? Like where do we want to, like what's the key thing for you guys? So as a B2B, it's about lead generation. So that's our end game. And then what we're doing is basically we're having a team of basically design and developers working on this sprint, like four week cycle of basically agile or scrum. And then going through and actually designing specific elements based on what the data is showing us. So if people are going into a certain part of the website, we then say, okay, 80% of all people are going into the help section, for example. Okay, that's the next part we need to build out. Yeah, a lot. So it's not as, and of course it's not as easy as it sounds. It also takes organizational change in terms of how you approach, because I think a lot of us in the agency world are very much waterfall-driven. Like it's like, okay, design, then it goes to specifications, then it goes to the, down the sort of step, and then a year later, six months after due delivery date and a million over budget, we'll have your product. And I think what we're starting to do now is say, how do we actually do this in a faster way that the customer, and getting their attention away from the pixels, and focusing more on, okay, what is it that we're trying to achieve? What is the numbers of it? So rather than say, okay, stop focusing on the, whether it's blue enough or it's the light Pantone or whatever it is, getting them to focus on, okay, what is that we're trying to achieve in terms of lead generation or filling the cart or whatever it is, and then using that. But it is a massive change in terms of the organization. How you think about these projects and getting people to actually develop and design on the back of that. Because that also changes the way you think about this whole waterfall approach, basically. I don't know if that answered the majority of it. I think it's, I mean, I don't know if I can give you a suggestion. I think what we try and do is basically have the team involved at the beginning. Like we try and include the design, someone from the tech, basically side of it, designed at the beginning so that they have an understanding of, okay, this is what the client is after, because typically I think it's, you know, in a typical organization, you have the account management or the sales guy who make the sale, they come back, they brief the design team that goes down in the system. Whereas what we're trying to do is say, okay, these guys all need to be at the beginning so they understand what the client is focused on, what the challenge is, and then they can basically go into their different area and start using that knowledge to inform them. And then we have the analysts which are providing the data saying, okay, this is what the numbers are saying. We can see that, you know, people aren't engaging with this button that you created or the menu structure doesn't work. We have to rethink how that information is prioritized or whatever it is. So I think a lot of it is just having that transparency and then involved at the beginning. I think that's, and it's also changing the way we sell projects as well. Rather than making that huge sale, it's sort of saying, okay, how do we actually go in with a different approach and sell on an iterative basis rather than the one size fits all approach. Any other questions? Yeah. So we work with Audubon, which is the bird basically in society. So a classic example of working with donations as our conversion point rather than lead generation or whatever. So we use the same approach exactly, but our end goal is in this case donations. So how can we actually build that donationship element? So there it's basically just to understand, okay, what type of users typically donate to the most? What information do they look at before? How do we take them through that user journey basically? And where do we target the specific information at what stage? And that's, I think that part of that personalization is it through the email marketing, it's through the social channels, it's through their owned assets on dot com and things like that. And then using those together to basically help drive them through that basically funnel. So that's sort of the very light version of it, but we're, exactly. So it's just understanding what do these guys look for on the web? What sites do they go to? What are they engaging with on our social channels, on Pinterest, on some of the more visual platforms? And then how do we use that information to inform the way we design and develop content on our site? And then we can use that information to serve that to them rather than say, okay, as a new customer coming into the website every single time, actually we know you're a big donor. So what information do we give you and how do we personalize that treatment to you as a large donor versus someone that's never heard of us and come through some sort of campaign that we're running concurrently? I think, yeah, so the first two weeks would be the design process and then the last two weeks is the development process, basically. But it depends a little bit on the size of the project basically, but that's typically the way we've broken it down. First two weeks doing the design iterations and then the second two weeks doing the implementation side of it. Yeah, I would say the testing and learning is goes underneath because the data, we set up the data framework or the measurement framework before we actually go through this process. So all the different elements we wanna test are built into our measurement framework. And then the learning is basically on the back of every sprint, we have a review system where we actually go through the data, identify, okay, what worked, what didn't, what is the new behavior teaching us and then we adapt and go into the next phase essentially from that. So the data part is or the measure part goes underneath everything that's constantly sort of running through our analytics, our different analytics platforms. And then the measure and learn or the, sorry, the build and learn is in those basically weekly chunks essentially. And then the learning is essentially a, there's the formal learning where we go through review the different elements, go through the data and understand it and then there's informal learning where it's constantly being tested. So the analysts and the actual guys committing the tests are actually understanding the actual use of behavior so they can then go and inform the designers and help them understand, okay, how do we change different things and test it that way? I don't know if that's a very vague level, but yeah, it's very, I mean, the biggest challenge we have is the client, usually like working with the client and getting them to buy into the process and then getting them involved in that process as well. So a lot of it, if we, depends if the project has a very short deadline then we obviously have to change the way we work based on that deadline. Whereas if it's a 12 month timeframe, it allows us extra room to commit to different elements of measurement or learning or whatever it is. Any other questions? No, well, thank you anyway for coming.