 Hi everybody. My name is John V. I'm actually pretty excited to be here and take you all through my experiences of building analytics products both for business intelligence users as well as cloud monitoring applications. So without much further ado, let's jump into my presentation. So I, I'm a product manager at follow alter networks at the moment and I build the autonomous digital experience management product along with the cloud monitoring product for Prisma access. And typically this is to help ID admins and users figure out, you know, issues in their cloud monitoring in the cloud infrastructure, looking for specific issues around application user experience as well as auto remediation and things like that. A little bit about me would be I, so I've been a product manager for quite some time now I started at, I actually started as a developer software engineer at SAP and worked for the CRM product with the building engine with CRM and right after that I joined Salesforce as a solution engineer worked on the pre sales organization and, you know, worked pretty closely with customers figuring out their niche use cases and product fits and demos, etc. Which was a great learning experience for me because I was so close to the customers and understanding their wants and needs made me made me think about how that could influence the product. And which is where I pivoted to a product manager role internally within Salesforce and I moved to Einstein analytics at that time which is now Tablo CRM, and I worked as a product manager on the analytics templates product, which typically helps business intelligence users build their sales analytics, service analytics, market analytics, all of that quickly, right from the templates that are available out of the box, ensuring fast time to value. Also, ensuring that they can follow some of the best practices that we put together after, you know, numerous customer sessions, etc. So that was a that was what I did at Tablo. And currently, I am a senior product manager at Paul Walter networks. And I work on the product autonomous digital experience management. A little bit of learnings on the way have been predominantly around building analytic applications and my also my major in college was data analytics and data science so I feel it was a natural progression for me to, you know, do to start to work on products that facilitate other fellow fellow learners such as myself in data analytics and data science fields. So, I, as you can see I have a big, you know, a varied experience all the way from software engineer to pre sales to product manager so I think I that definitely helped build my skill set over time, and definitely helped me view a product in multiple ways to kind of assess how the users are using it how the sales is pitching it how the builders are building it so it was a very unique perspective that I developed by across the Swedish experience had. So, I want to take you through a little bit of my learnings from the previous two products that I have worked. For sales force I pretty much worked on a business intelligence tool so the kind of users that that I focused as while building the business intelligence and analytic platform was predominantly sales service and marketing users. Now, these are some users that want to generate dashboards and generate reports in a timely manner. They want to get drive insights out of the data available quickly, and they want to stay informed about metrics that move the needle for them that they care about. And they want it in a place where they do their everyday business. So, which is where contextual analytics becomes really important. So, that was something that sales service and marketing users cared about. Now, there's still a team that utilizes the analytics platform to build these dashboards and reports for the said end users. That's the large organization of business analysts data and data science. So for them, what they cared about was, you know, predominantly the tools that's required to develop these dashboards and reports quickly enable ad hoc analysis. So if there's any questions that they were encountered with or they needed to provide answers, we should provide a toolkit for them that makes ad hoc analysis, relatively simple and easy to build. And they also have to deal with a large variety of data sets, the data sources being disparate, they, you know, they had to bring data from multiple sources, prep the data before that was visualized. So we had to ensure that the analytics platform also provided an ease of use and data management and prep tools that enabled business analysts and data analysts and like typically dashboard builders to quickly get their, get the dashboards and reports out or quickly get their, you know, get up to speed with some of the dashboards that they needed to build. They also cared about the fact that the platform needed to be or the product needed to be fully customizable and extensive fully customizable because oftentimes the dashboards out of the box may not suffice all the requirements and typically they also tend to use data sets that are outside of a single product or platform to build that holistic 360 degree view for their end users, which is where the extensibility of the platform was very important. Also customizability, right, and, you know, for some reasons that they for some users that they cater to one half of the view was very important versus the other for that we needed to build a fully customizable platform that lets users choose a variety of visualizations to present their data in a variety of filtration mechanisms as well as publishing mechanisms. So all of this needed to be cater to with me analytics platform. And finally, there's another group of users who mainly work on generating these predictive insights for their end users. And they typically cared about the ML capabilities and running the model and hypothesis testing and figuring out if that model is working well, how the accurate their predictions are, etc. So that's another tool, a plethora of features that we had to build with me analytics platform for that specifically cater to data scientists. Right from Salesforce to the product that I own and build right now is the cloud infrastructure monitoring and analytics platform. And typically the users here are network and security admins ID admins and the SRE group. They typically care about fires and how to fight them. So they want to understand or they want to be notified about change of events or any incidents that occur right there. And they typically also look forward for look, look to they typically also look to identifying the root causes of those events and incidents if that which one needed to be solved. Right so which makes monitoring issues in real time and real time aspect of the platform really really important for these users. And normally detection catching all of these issues early on in time to prevent or at least mitigate the impact is very important to the users to the current users, as well as in a lot of situations where it comes to infrastructure and network. Things are tied very closely to the amount of bandwidth they could use amount of licenses that are distributed across an org so based on their usage it's very important for them to predict what kind of capacity that will be required going forward so that that's where the capacity as planning aspect becomes really important. To ensure this, you know you want to make sure that you're catching telemetry building this rich repository of data with logs and metrics and telemetry across different vantage points within the infrastructure, bringing them together, correlating some of these metrics and incidents to provide an easily consumable insight for your end users, making their analysis to be simple and giving them more time to respond and remediate these problems rather than, you know, having them spend time on assessing the root cause assessing the issues and going back a historical trail and figuring out what's happened. Sometimes that is definitely important where you're figuring out where they do want to figure out the retrospective the incidents and events through historical analysis but oftentimes it is very proactive so the real time and the predictive nature of a platform becomes very very important in such a scenario. So there's a couple of points or a couple of themes that were similar across the analytics products that not only have I built but also studied, you know, studied while building out the competitive analysis viewpoints etc. So, across all of these products there's some themes and those predominates focus on the first one being they focus on unlocking insights that drive action ability and decision making. We want a rich feature set that enables visual storytelling, either through a do it yourself mechanism or build that within the analytics you're building itself visual storytelling is very important when it comes to consumption of these analytics themselves. For a analysis to be complete. Ideally we need them to be descriptive, diagnostic descriptive and predictive. I'll take us through what each of these mean and how they played out in the success of the previous products in the upcoming slides but I wanted to call them out quickly here. It's also important for users, not just to deliver dashboards and reports right they need these insights, where they can view them. They need to see these automated insights where they can actually take action and where, you know, in a timely manner that they can act quickly. That's insights where they're, you know, providing insights where they are actually needed for the end users. Then comes extensibility, which is APIs and customizations and, you know, furthermore is automated insights. So let's talk a little bit about the actionable it's the whole agenda of building a analytics dashboard report is to answer a few questions. Typically, users tend to work backwards where they identify the questions that need to be answered. They understand the metrics that the person the end user that they're building it for and user that's consuming these analytics are very, you know, they care about these metrics. So to that end, we need to focus on a design that makes correlation of these metrics easy. And we ensure that the design also accounts for this. Eventually, as an outcome of the analysis or, you know, the outcome of the deep dive in data that the end user does is to take action to make smarter decisions. Oftentimes, you know, customers who appreciate the product actually love that you can embed the actual ability. You do an analysis work through a flow and finally end that in an action that influences some a better outcome for the end users. So to that point, embedding actions within these visualizations is really, really important to ensure that the analysis they have done needs to actions. So if there were cornerstones of design or pillars that you guiding principles to build these analytics solutions, driving action ability or providing actionable insights that drive a better outcome is the number one guiding principle that, you know, you need to follow while building the analytics products or providing if you are building the analytics within the product integrated within the product, make sure that there is provision for it. Or if you are building a platform itself, make sure that there are there's a toolkit that's available that enables users to embed this action ability within their analytics platform within their dashboard. The next important that I've noticed in like building a successful analytics platform or product is visual storytelling. So it's not just important that you combine graphs, plots and tables in a single view, but it's also important to focus on the design elements to add a compelling narrative to the analytics that's built. Oftentimes you notice that analytics is not just for consumption, but it's also to present to a larger group of audience and some of the findings, which is when visual storytelling becomes really powerful and really strong. You're able to drive your point through a narrative that you have formed out of the dashboard that, you know, you could build on through your product. When you walk through a story of a user trying to perform a specific analysis, make sure your design validates that story so that way you're able to present a story accurately for instance. For instance, if you are trying to help a sales leader identify where their quarter will land, make sure that you start from the metrics that are important. Where they are today, the reasons for that and what can be improved this now this design makes for a strong narrative and also engages the audience when you're when you're presenting your insights. Also, when you're building the platform, make sure you provide the necessary toolkit such as a library of visualizations, recommendations if possible, some standard design patterns that make building this visually appealing, beautiful dashboards and reports easy. Make it easy for users to navigate between the different widgets, correlating these different widgets or placing these different widgets that make sense that need to be interpreted together in a single view, and also provide a wide variety of two kits that is necessary for drilling slicing and pricing the data as we popularly refer to this as so that way you can that way you're you're equipping your end users with the right set of features to build a visually compelling dashboard and reports. Let's talk about complete analytics now. So in both the products that I've worked on the business intelligence and the digital experience management product, I have noticed that you build analytics analytics come a complete circle when you are addressing what is happening. So this is the first level of analysis that most end users would do which is descriptive right what is the exploratory data analysis the pretty basic ones of what is happening and you help your end users see and interpret this data. You have drawn data set and you're just trying to build some roll up aggregate and measures and metrics and trying to compute the percentages to see what has changed over time and this is this is rather descriptive this is the basic. The basic element of analytics. You further strengthen that with the diagnostic analytics that is when you're figuring out. I'm seeing my data but why is that happening right like why what is the cause and can I see a pattern within that so that's that's that's the next step of understanding of presenting the analytics. You also want to understand, given, you know, given the general richness of data that you have, can you leverage some of the AI ML capabilities to understand what could happen. You know how you want to ensure that there is also the outcome that you predict also has some level of accuracy so that way, your end users are looking at what's happening, why is that happening and what could happen. Right so. And based on that, you need to finally give a list of recommendations or a list of steps or actions that the user could take, which is the prescriptive part of the analytics. Now, providing these four elements makes for a complete analysis. For instance, if you're looking at the revenue numbers of an organization, you want to look at where my revenue standing. Why is my revenue the way it is so what are the deals that are contributing to it, etc. If can I hit my goal this quarter or not. And finally, if I am not able to hit my goals, what should I do to ensure that I can hit my goals this is an example of one of the previous. You know, workflows that I designed and it was really validated by end users on the importance of this analytics. So, and you should also keep in mind to provide insights where they are needed. Oftentimes, your end users perhaps care about one or two metrics which are very important for them, which they want to be notified about all the time. So, to provide that this two mechanisms that you could use one is contextual analytics that is you embed your analytics into your solutions. So, for instance, if you have the end user, you know, taking a bunch of actions on your opportunity items or your records, you want to ensure that you're embedding the analytics right where where they can also, you know, back up there, you know, perform some sort of analysis, and then take right action. Right there so you're providing them a holistic view of the problem with the data to back up and also in equipping them with the necessary mechanisms that they can act on. When it comes to notification alerts, it becomes very important, especially in situations where you have to fight fires constantly that you be notified proactively about issues that might cause some sort of a service disruption, or if you're not hitting your goal, these are the metrics that users want to be notified about. So, either you build that in the product, you build a notification and alerting system where you can notify users of events and metrics as soon as they happen right within the product or equip your users to build them themselves. Some sort of customization capability or building a developer toolkit if you may to provide these alerts and notifications to their end users. This enables end users decide and also you by this way you're also enabling your end users to decide which metrics and changes they really care about and use their own discretion to build notification alerts around them. It's not just it's one step to provide these notifications and alerts and also another step to make consumption of them easy. So, you know, you often what's what was proved in the past minor experience was to make the consumption mechanism really really easy and reliable. Use popularly used tools such as Slack, email messages to get these notifications upfront to customers so they're aware of the data changes or events as and when they occur in a place that's very easy for them to access. And last but not the least, the extensibility of a platform becomes very important when there's disparate data sources. So, when you want to build a large scale data analysis tool, you it never, it never happens in isolation you off users tend to ingest data from many sources to build a 360 degree view, so that capability to export data from additional sources. I mean import data from additional sources or leverage the API's and connectors to get data from third party systems analyze them in your own platform or make the or build a rich set of API's that users can consume to get the data out of your system becomes very important and really necessary for end users. And follow that out by customization users always require the ability to make customizations to the dashboard you presented for instance, oftentimes the time of time of the analysis that you're performing becomes important so you need to be able to provide the customizations of that they're looking for, or, you know, if the end user feels like visualization could be represented in a different form. That is something you could offer up as a customization technique. If there's any specific changes to the underlying query that the user wants to do that's another customization. So, a rich set of customizations rich set of publishing methods that you could extend to your end users will definitely move the needle in terms of adoption and engagement. That's some of the insights that I gathered. I learned during my time as a product manager for analytics products and it's an ongoing journey. You know, I learn about it every day. I hope the webinar was useful for the audiences who attended and feel free to leave me any comments or questions. I'm happy to answer that offline. I hope you had as much fun listening to this as I had presented it. Thank you.