Way of introductions, and Vicki's kind of taking a bit of my thunder, albeit not that significant, I don't think. But my name is Jim Horbery. I'm the solutions program director at Interworks. What that means behind all the technical kind of nomenclature is that I'm essentially in charge of presales consulting into work. So requirements gathering, working with clients to develop solution architecture to find the best fit to help clients evaluate tools, more relevantly for today, I have spent probably getting on for ten years now working around embedded analytics solutions. So in my past life, before becoming a consultant. I was working in ad tech and marketing tech, and I was helping agencies build embedded analytics solutions that reported on customers, and clients, digital performance, think PPC and stuff like that and we would sell that product to those clients, in order to help them make better decisions and then roll it forward to today, and over the last, I think, almost seven years or so, I've been advising a much broader range of organizations about doing very similar things. So standing up embedded analytics solutions, thinking about the way, not just in terms how you architect these solutions, but how you bring these things to market, and and what things you should look out for and so on. So, yeah, as Vicki says, I think I probably contract we obliged at this point to point out, I've written a book on the topic with a friend of mine called Donald Farmer. Donald runs his own consultancy firm in the US. And it's fair to say that some of the topics we touch on today will be covered in this book we have some copies that we're going to be giving away, which is what Ricky alluded to. So we'll be picking some names out of a hat. And approaching you guys that are selected to see if you'd like a copy. However, in terms of introduction, about the content there's a couple of points I wanna make first off. And and that is that at interworks, we have a content management system called curator, some of you who are on the call might have seen some marketing material around this or might be aware of the product that we create. Curator is a, a content management system that integrates directly and a sort of deep, deep integration with tools like tableau, with thought spot, with power bi, to enable you to build embedded analytics solutions. So it is, it is the vehicle, if you like, for us to be able to deliver embedded analytics as part of the the products and services that we offer. As you might imagine, and again, some of you may have attended things. We do an awful lot of activity around curator. We we do blogs. We do demos. We do webinars like this. We do kind of technical deep dives. And One of the things that, has become obvious to us as we've kind of gone through this process is that whilst some, of the clients and organizations that we work with are kind of ready to to dip their toe in the water with embedded analytics There's a significant proportion of organizations that are very, very early stages. So maybe they don't even have, an analytics platform in order to produce things that can be embedded. Maybe they don't know how embedded solutions can be obstructed, what components need to be glued together to make these things work, or maybe then even unsure about how you might position and sell and talk about an embedded analytics product. And so what I wanted to do today was kind of take a little step back away from very specific, very dry kind of technical detail and attempt to answer that question there on the left. What value does embedded deliver, and what do I need to know if I'm thinking about doing it? And I think that that latter part for me as a consultant is super important. We've seen, lots of success, lots of individual organizations go out there and and win with embedded analytics solutions, but we've also seen a lot of organizations that that fail. Because they approach it wrong or they're confronted with a challenge that they didn't understand. And so I thought it might be useful to put some content together around that and specifically on the right there, this is the order that we're going to attack this. I'm going to talk a little bit in a minute about where you might see embedded analytics happening. And I'm gonna I'm gonna very deliberately, set that up as two kinds of possible use case which will hopefully make the explanations later on in the session a little bit more clear. I'm gonna go back to basics a bit and just touch on why you'd even bother doing it I want to talk about visual analytics, which is often the the main thing that you end up embedding, not the only thing, but certainly a primary thing that that that has value. But I wanna talk about the ubiquitous topic of data monetization as well. And again, I'm not gonna go into a deep dive, but I just want to kind of you know, draw a picture about the reasons that, you know, these kind of products can be so successful. And then, and I hope this bit's gonna work. As with a lot of tech chats, I'm gonna use an analogy to help us get familiar with the basic architecture of an embedded analytics solution. Now, I really hope that the analogy works, and it isn't too cringy because I'm gonna hold on to it for a couple of slides, and I'm gonna feed that in to to describing some real world challenges. So I'm gonna sort of hopefully transition between the analogy in the real world, if that makes sense. Last but not least, it would be remiss of me at least not to mention curator that the product I talked about, a few minutes ago. I'm not gonna do a deep dive. I'm not gonna do a demo, but I do want to use some screenshots from, curated demos that we have publicly accessible at interworks. Those demos, I think it would just be a good kind of, bookend to the presentation just for people to see what's possible, and I'll talk a little bit about some of the points that we will have, by that time, we will have covered. So, yes, I hope that makes sense. I'm I'm probably aiming for excuse me, maybe, forty five, fifty minutes of talk, and as Vicky says that, hopefully, there'll be some room for questions, at the end as well, if anybody has anything they want to ask. So Let's get started. First off, where do we see embedded analytics happening? And this is that kind of very distinct two tier use case. Now, some of you who are more familiar with this topic will immediately say in back of your minds, well, look, it's not as clear cut as that. You're absolutely right. There are lots of use cases where these two things are blended together. But for the sake of explanation, I'm I'm gonna use this kind of color as well as a mechanism to tag things. So the two, distinct categories, first one, top right, as a standalone product, that little graphic that I've stolen there from the internet is, similar web. Some of you might be familiar with it. Some of you might have used that in your day to day jobs. We see embedded analytics being used in lots of standalone products. And just to be clear about what I mean by that definition, the definition being the product is embedded analytics. So you're coming to a specific thing usually paying for it or there's some mechanism by which you are granted access, but that is the thing that you're going there to consume. And and data and, market research market intelligence companies. I think, similar web talk about themselves as a kind of digital intelligence, service, a great example, and there are hundreds of these kind of companies out there. So they take data, they will add value to that data. They will embed visualizations around those datas and around that data, and they will put that into a product that you pay to access. That's the most obvious, starting point. And I suspect most of us, even if it's sort of unbeknownst to us probably access some sort of standalone embedded analytics product in our day to day lives, but it's not the only category. We also see standalone products being created by companies who want to add value to something that they already do. So, as an example, we worked with a loan servicing company or a group of companies that were involved in loan servicing during the pandemic, when, customers were offered, repayment holidays on mortgages and loans, which some of us may have taken up, that act of deferring a loan or deferring the the capital payment on a loan has a significant impact in in the background of these companies on the overall portfolio of loans, so duration, the the risk category, interest rates, and all this kind of thing, and the company that we worked with kind of, cottoned on to the opportunity and built a stand alone embedded analytics product that visualize the effect of this kind of movement, this churn in in the mortgage and lending markets and they push that out as an added value service to their existing customers to kind of differentiate themselves from any other company doing similar things. So It wasn't something that was paid for directly, but it was very much an added value, like a sort of usP service that they provided. But again, it was a standalone product own branding, own logo, own kind of messaging. Last category, we see their organizational intranets, less often, to be fair, but in organizations where the brand really matters, and, you know, I know it always matters, but especially so in things like, creative industries, advertising agencies and so on. Of a certain size, they usually like to control the way that they're internal analytics are displayed. So think about, headcount, think about benefit schemes, think about kind of top line P and L charts and things like that that might exist within a a an internal intranet, all beautifully branded, all very much the same category of embedded analytics within a standalone product. So that's one category which I'm showing as a a pink title, and I'll come back to the colors in a second as well. The other section here is within the workflow. The distinction I'm attempting to make here is embedded analytics that are being used to add value or augment something that already exists to do a purpose. So it's not a standalone product. It could, in theory, work just fine without the embedded analytics components, but embedded analytics adds value. And most of you will recognize that. That's a screenshot from Salesforce, which kind of the the ubiquitous CRM that most people are familiar with if they don't use it on a day to day basis. Those guys were one of the first organizations to put visual analytics within the workflow. So a customer could work through their day to day role. They could be interacting with entities in Salesforce, and they could be seeing charts and graphs that related to what they were seeing in the same place within the workflow. Similarly, we have, tools here at Interworks, which would fall under the category of kind of, ERP sort of resource planning tools, and in interworks, it's a custom tool, but we augment that with visual analytics, our selves as well. So that would be another example of sort of within the workflow use. And the last example, and I just want to call this out because they were kind of early to it. I did a stint in telecoms a long time ago. And, for call center management, obviously, there are hundreds of different vendors out there that produce kind of call center or, you know, telephony, management tools Those were some of the first tools that saw embedded analytics being added to the tool to add value to it. And I think in hindsight because of the immediacy of what a call center does and the fact that you need to be, not only have up to date information, But you need to be able to act on that information immediately. Open up another group of people to answer phones, transfer calls. These are very immediate and visual analytics is a fantastic way of communicating that. So service delivery would be another category there as well. So, the reason I've broken this up is in these two ways is because as we walk through these slides, some of the things I talk about will apply to one and some will apply to the other, and some do overlap as well. And I thought it might be important just to add a bit of context here. So at this point, I think my colleague is going to launch a poll where we're going to ask you some options about what embedded scenarios are relevant to your organization. It's very useful to understand from our perspective where the customers that we interact with, where they sit. And you may not have a very clear answer. As I said at the beginning of this, some of these are necessarily gonna overlap, but a very simple poll, if you would spend a couple of seconds just clicking some options in there as we go, that would be very much appreciated. So, so that's where we see it. Let's talk a little bit about why you'd bother doing this in the first place. And I think one of the slides that I have used, various versions of this in probably fifty or sixty presentations I've given over the last few years is about visual analytics Now I appreciate that you can and indeed these days more and more non visual things are being embedded around analytics. So I'm referring to things like, chat prompts or NLP kind of asking your data, you know, text based queering and so on and so forth. In the most part, most of those tools, and I'm being deliberately kind of broad brush at the minute, There is still an analytic a visual analytics component to that. When you get your results, those results are visualized, they might not be visualized by somebody, they might be created on the fly, but visual analytics is important. And this is where I I I wanna make the slightly more, abstract point that from a human perspective, we are pattern recognition machines. The the way if you if you don't have any visual impairments the way that your brain and your eyes work means that we are, very good, very quick at evaluating quantity and evaluate the status and category of things without necessarily having to count those things up. And I know it seems like a crazy example, but think about looking at an apple tree, and being asked the question how many apples are ripe is the tree ripe, and you'd look at the apples and you'd see, well, I can see most of them are red and some of them are green. Yep. I'm gonna go with most of the apples are right. You're using your visual system. You're not counting everything, but you're just evaluating the picture as a whole to come to a decision making point. There are countless examples of that stretching all the way back the way that sort of cavemen think and all that stuff. But to bring it back to our sort of, purpose, consider this example we've got here. And I think it's pretty obvious where we're going. When we have a grid of numbers in front of us, that evaluation system that we rely on so much through our day to day life doesn't work. Right? If I was asked here and I'm being confronted, it might be a little bit blurry for for some of you, but We've got weeks, and then we've got sales figures by three different segments here represented in a in a crosstab. If I'm being asked to interrogate this and to come up with some conclusion, I'm gonna work through this, and I'm gonna be maybe keeping a mental total as I run through. I'm not particularly good at that kind of that's quite hard for me. I might be able to spot outliers in this grid, but I'm gonna have to look at this for a minute and maybe click around in it to understand, what what these value mean. Logically, if we start to apply some visual analytics techniques, that job becomes easier. So we just apply basic heat map as an example here. And instantly, I can see that there is an outlier in the corporate section. I can see that consumer generally trends above the other two agrees all the other two segments here. I don't have to add these numbers up anymore and taking this analogy all the way through, obviously. I can then those of the figures themselves, maybe I hide the figures away or I make them available as a tool tip, but clearly I can chart these in a bar graph, this gives us a much better idea of movement. I can see how much better consumer are doing than the other two segments. I can see that outline week of corporate, and I can see that maybe home office are having a bad time, but things might be trending up. And the last example there, and for those of you that know your best practice visual analytics, you might point out at this point where you're using green and blue, which is definitely not a good idea. That is entirely correct. So for those of you with, some level of visual impairment, that might not be a great combo, but it serves to illustrate my point in the I can go a bit further here, right? I can add some average lines in here so I can understand if this performance of the consumer segment is above or below some sort of notional average for targeting purposes. So all this serves as a sort of baseline point, really. Vision analytics is a great way of displaying data. More importantly, it's a great way of allowing people to make conclusions and interpret what the data is showing them. And there is a speed element here, which lends itself very well to being embedded in something else. So, as a logical extension to that, the second reason, that this is, that that this is worth doing is all about that workflow example. So you remember we talked about sales force, the CRM application. Oh, sorry. I didn't do that one there. We talked about Salesforce. That's the screenshot down at the bottom. The reason that Salesforce put embedded analytics in their product is in order to keep you in the product. Know this is logical, but if you think about the practical side of things, if you needed to do some analysis, if you needed to plot a bar chart or trend line and the data was available to you in your CRM, but it wasn't visualized. You have to do some things. You have to export the data. You have to, I don't know, copy the data out, paste it into something else, load that application switch windows. I mean, how how many of us are so used to navigating between fifty or sixty different apps, on their desktop every single day, you you gotta do that. Right? So if you keep the user in the workflow, you don't need to switch focus as a user. And clearly, making a callback to my original point here, This means you can explore and you can interpret this data in the same place, right? So you're not jumping around. You you can see, you can cross reference, you can use that, visual evaluation system, if you like, that I kind of alluded to at the beginning. Logically, then the conclusion is that you can do things faster. You can have a look at the data. You can make decisions about the data, and you can act upon those data those data points and, you know, and potentially trigger other actions from the tool as well. Last point on this, which I will come back to, as we go through here, is it looks good or it should look good. If you are embedding things in applications for any reason, to be fair, not just workflow, but as a standalone product, how good it looks cannot be sort of, you know, underestimated in terms of how important that kind of feature is. People are predisposed to spend more time interacting with things where there has been careful consideration of spacing and and typography and styles and color palettes and things like that. That's why so many companies pump so much money into branding. Right? So It's a it's a very it's not always pursued, especially in workflow. We've seen a lot of bad applications of this. But it's looking tightly integrated, so no difference between embedded analytics and the thing that it's embedded in is very important. So, third and and last reason here before we start talking in a little bit more detail about the architecture here, data has real monetary value. Now, most of us that have been close to this kind of topic over the last five, six, seven, eight years are very familiar with, sort of cliche phrases about your data being more valuable than oil. And there have been several, case studies and kind of technical exercises where large organizations have done studies and they've they've done some research and they come back with a conclusion that says, okay, well, we're, we're actually our data is worth more than the generally accepted valuation of the company. Now, those figures are kind of startling, and of course the measurement framework itself might not be universe agreed upon, but the point remains that there is definitely, obvious and not so obvious ways in which you can make your data be worth something. And personally, not necessarily, completely into work centric way of looking at it, but when I talk about monetary value of data or monetizing data or building data products, I often break it down into these three categories, because I think it makes sense because you can sort of explain the synergy between them. The first one, the most obvious one in the center there, is you can visualize data. Simply visualizing your data and taking advantage of some of those things that we've touched upon so far you know, speed to insight the ability to, for people to be curious and to interact with the data. Doing that with your data can provide value. So, so to be very clear, and into works, we work with several organizations who have taken a lot of time and effort to visualize data that they already have access and are already using. And they have made that data publicly accessible, and they are generating revenue from people accessing that data, but it is visualized in a very beautiful, a very effective, very compelling way. And I've attempted to illustrate here. I mean, I've just taken this from Tableau Public, which is a great spot to go to if you're looking for inspiration or indeed getting resources together for a webinar. But to me, this is a fantastic example of great visualization. I'm not really concerned so much about the the technical validity for this example, but if you look at this dashboard or the snippet from this dashboard, it's clean. It's concise. It's clear. There is attention paid to spacing, the way that the elements are laid out. It is ripe for investigation. I wouldn't mind clicking around with this and playing around with some of these filters and so on see how this was behaving. So so that's almost like, you don't have to do it, but it's like a de facto number one thing. Visualizing your data in an appropriate and compelling way can immediately be make it worth money. So we then get into the slightly more complex topics. On the left hand side, we've got enrich with ELT or ETL depending upon your your approach. Now, again, again, thinking about the sort of, the level that we wanna keep at here, I'm not gonna talk about the processes of integrating data, but let's just take an example. Let's imagine that we have, agricultural data. Maybe we are visualizing seed and livestock pricing, right? Maybe that's the product that we're trying to to monetize. And we've collected all of this data from our own transactions, from our own database about prices, bulk prices, you know, price rises, price decreases, that kind of stuff. But if we take an external data source, maybe something like weather, forecast patterns, long term trends, and so on and so forth, and we integrate that data with our own data, so internal and external data together, maybe we could offer some more advanced visualizations that we could forecast the price over long term based on weather patterns and weather fluctuations. Now, I don't know anything about the agricultural, agriculture vertical, probably obvious from that statement. But I can imagine that there certainly would be value wrapped up in that. If I was gonna be able to forecast, seed prices for the next six months or even into the next season, that might be of significant value for me. So think about this rather than doing one or the other. Think about building up this kind of, you know, the the the suite of added value fix. We're visualizing this data in a neat way. We're also enriching it. We're getting external data sources in there as well to do things that other people may not be able to do. And so to take that to the, the next level, and I think every webinar has to have some mention of this these days, but DSL, so data science, machine learning, those processes can be used to augment the data. So we're not necessarily integrating other data sources here. What we're doing with DSML is we're looking at the data we already have, and we're saying, are there any learnings or insights that are locked up in this data set that I don't realize and that I might be able to surface with some with some machine learning models. And and one classic example, would be something like, health care, where it's possible to build machine learning models based on existing people and existing clinical conditions, and you are able to build models that can predict the likelihood for people to develop clinical conditions in future. So to put that in layman's terms, given my demographics and my height and my weight and all these kind of things, it's possible to predict the likelihood that I will present with a back problem or a vision problem in later life. If you can imagine the value of that within the health care industry, being able to predict cost is important. And one of the more interesting aspects and potentially more valuable, if you can predict the likelihood of an individual presenting for some sort of clinical problem, you can potentially get there before they present that and prescribe some sort of alternative non clinical method go and get more exercise, etcetera, etcetera, which is usually at the root of these things, but you get the idea. These are methods that can be kind of stacked on top of one another to really add solid value to your proposition as an embedded analytics, you know, solution. So these aren't the only reasons, but we need to get into the architecture a little bit, and this is where I'll start talking about my analogy. And I really do fingers crossed hope this sticks for the purposes of this webinar. So as you can see, my analogy here for talking about architecture and a sort of infrastructure for this, is a closed door, and I have named these in a very specific way, but I hope this of make sense as we walk through it. Our clothes store, needs somewhere to store its clothes. Obviously enough, it has a warehouse and the warehouse might be, it might be connected to the store directly. It might be off in an industrial estate somewhere, but the key is that there is a secure storage location where all of the products that we sell at our store, live. And they are organized and categorized in an appropriate way. Look at those neat shelves. Within the store itself, we also have this guy, who is our federated access point. Now you might think, okay, yeah, that's just the shop guy, but in in our analogy, he's all about federating access. And the reason I picked that so specifically is because he actually does two things in, in my example. He's the guy you need to talk to if you want to transact in this short in this store, so you need to bring in your jeans and you need to give him your credit card and he needs to take the tax. He he's the only method that you're gonna be able to interact with to leave the store with something. But let's take that a little step further and imagine that it's a high end store, this is also the guy that you could present your VIP cards to, and he would say, okay, Jim, you are a member of the VIP club. You're allowed in the special section of the store where we're gonna show you this new collection that we've just got in. It's not available to the general public, but you're a VIP member you get access to this stuff before everybody else. So maybe I have paid for a sort of premium level of access. I don't actually know. I don't think I've ever been in the close shop but does personally, but I'm sure they exist. So bear with me on this. Third element there, whatever happens whether I'm regular member of the public or member of the VIP club, I'm gonna see at some point a curated display. And the point here is that somebody in this store has gone to the trouble of organizing these elements, organizing this new collection in a way that makes it very compelling for me to buy. I can see what other accessories or what goes with what somebody is trying to tell me a story with these curated displays. And no doubt in in a big store, there's gonna be lots of different curated displays. Last but not least, obviously enough, all of this sits in the store, and we're calling that the storefront, right? That's the presentation layer, and I think you can probably see where I'm going with this. Key things to remember with this, We have to have somewhere to put all these things that I've described, but the store itself usually has some sort of theme, right? It might be an eco sustainable materials store where all the clothes have made out of sustainable cotton and what have you. It might be a super high end spoke design a shop with only the very most expensive things in there. And let's take it a step further. It might also have some non clothing things in there like homeware or whatever. So you might be able to find bargain homeware with your bargain shoes and vice versa. You get the idea. But the point being, everything in that storefront is deliberately associated and related to each other thing in order to make it a compelling agreement. You know as soon as you open the door, what kind of shop this is, you know, before you even get to the door, what you're going to expect. So I hope the analogy is sticking because we're now going to start overlaying this to the real world components and bringing it back to our embedded analytics example. And I think maybe hopefully obvious. The secure storage, well, that's our database. If we're building an embedded analytics solution, we need to have somewhere to store our data. Now it could be within the product itself. So in a workflow example, like Salesforce CRM, well Salesforce are hosting your data for you. That's where your your, your, your product is being stored. If you like. It doesn't have to be like that as long as those boxes can be ticked, that it is secure and that the data is correctly organized in there. That that's a good start. Right? Then we have our federated Access guy. Now for this, we're imagining, software that can control who has access to our system. So think about, a login to the system. Somebody putting their username and password. But here's the little nuance that I was trying to introduce with the VIP idea. We also need to store information about whether or not Jim, me, has access to the VIP stuff or not. So this federated access layer, which we'll talk about in a second, also has to store information, not just allowing me access, but it has to store information about what I'm allowed to see, what plan, what enterprise level, what skew I might belong to, and again, we'll explain that a bit further. Obviously enough, I'm making a link between the curated display and that being a dash board. So there will be n number of dashboards or visual analytics objects that we're actually embedding in this, solution. And I think the key thing that I would, you know, that that the key thing that I'd start to highlight at this point is that the dashboards cannot be there just for the sake of being there. They have to provide some level of actionable value, right? Actionable insight, what they have to provide the user, the visitor with some value that they didn't previously have before. But key point, they are curated. Somebody has spent a lot of time trying to think about how best to fit those kind of elements together with one another. And then last but not least, obviously, we have all of this tightly integrated within our host application within our host website. And again, I just wanna make the point a successful embedded analytics solution will integrate these things together, this storage, this federated access layer, and these dashboards in such a way as you are not aware that you're interacting with anything separate at all. It will present you with a unified experience. Okay. So Hopefully, that's obvious. And and we'll start moving away from this analogy a little bit when we talk about the considerations now. At this stage, I want to point out that I think in one sense, this is almost the most valuable part of it for me we have, as I've mentioned at the beginning, we've helped countless organizations that interworks build these things. We've helped customers finished something they've started themselves. We've helped customers, whiteboard from from from a zero position, what their journey might look like embedded analytics, and we've we've built those things from scratch to deliver for our clients. I think one of the most interesting and useful aspects of having gone through that process over and over again is that we know the if you like the top ten or the top twenty considerations that often catch these organizations out. And I just wanna walk through some of these things. Again, don't I'm not going to talk about these points individually to to exhausting detail. Some of these are obvious, I would think. Some of them aren't And I think that's why they're worth talking about because they're kind of gotchas that you need to know. Starting from the left, if we think about the data, well, we need to know where it as we've established, we need to have it in a database. We need to be sure that it is ready for analysis. Now that phrase, is it ready for analysis could encompass and does a whole other, suite of webinars about what it means to get data ready for analytics, but suffice to say the number of times that we've encountered organized sations where they believe the data is ready and it isn't is very, very high. The the other thing to bear in mind with getting your data ready for analysis and getting your data ready for embedded analytics is that that work you need to do data engineering, architecture, orchestration, all those kind of things. It's not very shiny work. It's not work that you can show off to, the, the c suite and, you know, and, and demonstrate to great accolade and all the rest of it. It's not visual. So often because of that reason, it's not necessarily given the focus that it requires at the beginning of this process. So I would I would sort of insert that kind of flag there to think about. It needs to be ready for analysis. It's also worth right at the start of this, and I'd refer back to my monetized piece thinking very hard about what we can do to this to to make it worth money. Can it be enriched? Are there easy ways that we could augment this data or determine something from it that might just elevate it above whatever similar competitors are doing. Sometimes, even though those topics are kind of they might sound complex and and are highly technical. Sometimes there are some really easy wins hidden away in data that that just a bit of sort of due diligence might reveal. So it's it's definitely worth thinking about that at the beginning. Last point, performance. We are very familiar with the world of dashboard performance across tools like Tableau, and how to make Tableau go fast. One of the universal truths about that is that it's almost always down to the underlying database, wherever that data is, if it's a file, if it's an extract, if it's a cloud data warehouse, the speed that queries can be executed and return results to the dashboard and allow the dashboard to be loaded for that end user, is critical. And I'd say whilst these considerations apply to both standalone and to, workflow products, I'd say if you are building a standalone embedded analytics solution, and you're not focused on performance, then you're setting yourself up for failure to some tank users, as we all know, we wouldn't do it. You will not wait thirty five seconds for a dashboard to load, especially if you just paid seven hundred and fifty dollars a month to access that service. So the database needs to be fast. It needs to be scalable. Moving along, the federated access part of this, there's a, a nuance here that isn't immediately obvious. We need a way to grant users access to this solution. Fine. That's well understood. It's gotta be behind a login. It's probably got some level of secure data. But we also need a method of knowing whether or not Jim, me, is a member of that VIP club. Do what what am I supposed, what level of access am I supposed to have once I, I get into this system? For that reason, even though it seems a bit incongruous to have it in this section, knowing what your commercial model is going to be knowing how you're going to monetize that data, thinking about what the plans might look like or what you're going to give people or customers or organizations that pay extra money actually needs to feed into the process when you consider where you're storing this information. So that federated access layer potentially has information in there about the plan that that user has bought. So those things are sort of inextricably linked. You can't decouple those those aspects. Then we move over to the curated display part of this, visualizations dashboards that provide actual insight. And I think this is where the core things about visual analytics come into play for us. First of all, and a real schoolboy error does the platform you want to use for your visual analytics support embedding? Tableau, thought spot, power bi, mainly all the kind of, you know, the the leading vendors out there do, not all of them do. And we have seen accounts where people have just made an assumption that their BI tool of choice can be embedded in their SharePoint or in their sort of custom application and they find out it just can't. And that's the end of that. So it's really worth just doing that that that check right at the beginning. Can we actually embed this content? And then the questions of what should we embed and what are the appropriate visualizations to use. Sometimes there is this onus to just put everything in there to to fire everything against the wall and make sure that we've we've sort of demonstrated that we can put enough stuff in this as possible. Some thought and some real careful consideration should be applied there. Your users will not, in the most part, be happy interpreting very esoteric or very complex charts and graphs that might look amazing but deliver limited insight. So moving across the the the storefront. Last time, I'm gonna mention this analogy. Can we tie this all together? Is there a method? Do we have control over the way this is displayed? Can we create a kind of seamless experience for our customers or for our end users where they don't know. And similar to that, it doesn't end there. What happens Next, once somebody has visualized the dashboard and they've interacted with it, is there a way that we can extend that journey for the customer? So a real classic example that we've seen a few times is I can look at a dashboard. I can see where we've seen this with facilities management. I see there's a problem in a particular building. I can click something next to my dashboard. That will create a ticket that automatically gets fired into our ticket tracking system. With all the pertinent details. So we're thinking very hard about closing that loop. So, so those are the sort of real world considerations. And I think fair to sort of, as I say, completely dispense with the analogy now and just talk about toolings briefly. So what we're saying here, is for each of these elements, that I've walked through, what would we choose? Now, just for the sake of argument, let's imagine that we're building our own version of similar web. Right? We're gonna build a market research hub, but we're gonna build it out of these best in breed components. This is what I choose, and this is in fact exactly the stack that several of our organizations that we work with use. On the left hand side, we've got Snowflake. Snowflake data warehouse, super scalable, super robust, cost efficient, can control the the compute usage in Snowflake, which allows you to control your cost. Snowflake themselves portray themselves as a data cloud. They have a whole host of sort of additional features on the platform that would actually allow you to do some of that enrichment and augmentation of data that I referred to a few slides ago. So Snowflake is a fantastic stick choice as our underlying storage mechanism. Some of you might be familiar with Okta. Octa are essentially an identity management form, but suffice to say they they have a method of both providing that fine grain control around what a user should and shouldn't be able to add, but actually as a an additional thing, and this is what Okta is often used for singularly is for single sign on. So if I'm coming into this solution as an end user, I want to log in. That's fair, but if I move to a page that contains a visualization, it's not really appropriate that I'm asked to log in again. Now that might seem obvious. I'm, you know, I'm accessing content something else now, but for an end user, it's not a great experience. So Octa allows you to tie all that together. Once you're authenticated into the platform, octa knows who you are, and it effectively authenticates for you when you access or request data from the warehouse and so on and so forth. So Oxford is a great fit for this. I'll make the point about Tableau specifically Tableau cloud. If you are building, data visualizations where look and feel and the user experience matters more so with standalone products like this, maybe than workflow, but I think it applies to both. Tableau is the king. It it that the methods available to Tableau developers, Tableau creators, enabling you to build infographic, type, three interactive, extensible dashboards, is a real, a a real US p when it comes to that middle section of, you know, make visualizations because they can have value, accrued to them or associated with them. So I would make the cloud, the case that Tableau Cloud and Tableau would be my choice for a standalone application all day long. And then finally, again, we need a storefront for all this and curate our product, our content management system that I mentioned at the at the start of this would be a great fit. And just to reiterate here as well, one of the the things about curator, and certainly one of the things that I would advise people thinking about these kind of solutions, if they were going at it alone or they were building or buying something else, think about what happens in future. Curator is extensible and you can build contents from Tableau, from thought spot, from power bi, you can integrate all of those types of content together if you want alongside more traditional CMS content, blog posts, news articles, case studies. Think about the analogy again. Apologies, last last callback to the analogy, but in that storefront, We had homeware and we had those curated displays of clothes. Case studies augmented with embedded analytics would be a fantastic example of a good standalone to my opinion. So, so that's our our market research hub. Before we wrap up, I just wanted to work for a couple of screenshots. And and these are just illustrations, really, because I think it helps kind of, sort of reiterate in your mind what these products might look like for an end user. So all of these demos, they're just screenshots taken from into work curator demo page. Just a couple of different examples to talk about what this, you know, what is what this might feel like from an end user perspective. This one is, an HR and people analytics solution. And you might imagine that actually this could be like a workflow solution. This could be branded as such that it is actually part of an existing employee management hub or something like I mentioned at the beginning of the call. And you can see, and I'm kind of building up the case here, we have We've used Tableau for this example, but we're trying quite hard to make it not obviously Tableau. We've got different action mechanisms above the dashboard. They're not from Tableau. We're just talking to Tableau in curator. We've got these very tight integrations in there. Well, we can go a bit further than that. Interburger is one of our most popular demos, and this is where we've really tried to replicate the look and feel of one system with another still Tableau in this example to present something. This, again, could be a workflow application that is tightly integrated, very careful attention to typography. Remember me making the case about how important it is that things look good. That's why we're we've gone to this extent. And it offers a compelling kind of unified example to the end users as well. This looks like an analytic reporting hub that somebody spent an awful lot of time building. Last but not least, again, just a random screenshot, really, you can take it a lot further. Here's a total sales dash port that looks like somebody's designed it from scratch. This is actually an interactive dashboard that people can drill down into. But just goes to show that you can take this idea of integrating embedded analytics alongside other content. You can take it much further that that that it might obviously appear. You can really knit those two things together to give your end users a really compelling experience. So, I think we're pretty much on time. I'm just gonna sum up with a couple of thoughts here at the end. I hope we've been able to at least shed some light on some of these questions and to call back to that original point about the value that embedded analytics can deliver and and what kind of key considerations you need. Few points just to sign off with, clearly, this is an Intuitworks webinar. If you need help, please just ask us. There is probably not a single specs of any part of this that I've talked about today that we haven't dealt with in a hands on technical nature. So even if it's just guidance, informal guidance, some advice, just reach out. We'll route you to the right person and we'll get your your answer as as quickly as we can. Last three points, when you're doing this, if you're thinking about it, be realistic and flexible when planning. Do not build huge monolithic system specifications with all the must have nice to have and so on. The nature of these tools, especially because of the vendors behind is that they are modular. You need to run experiments, think proof of value, proof of concept. That is a great way of doing it. The monolithic projects, will just start to get paralyzed by decision making in in from our view in terms of the consulting perspective. Think about the pivotal technical decisions. I've mentioned four tools today, which I think are best of breed. They might not be best of breed for your particular case, but you should be thinking about what those vendors look like and getting together a group of vendors for evaluation right from the start. Do not leave it. To the last minute on the assumption that all of the tools are are are similar or can, you know, fulfill your capacity in the same way. Last but not least, especially for monetized products. Remember that commercial strategy, the the basic membership, the VI membership. What is gonna distinguish those users? What are you gonna do in the background that makes it worth it for those users and how you can scale it? That's what you should also be thinking about at the beginning. So, yeah, I think we have roughly five minutes left. I think there is one question in the chat. There is indeed. So any examples of visualizations, driving value, or money. So I think you did a lot of examples throughout a presentation, but I don't know if there is a specific one that comes to mind, if we think about anonymizing one of our customers who said, great. This is revolutionized and and driven value. I certainly can. So from a health care perspective, We did a lot of work to support a client who was interested in understanding bed occupancy, within their hospital. Now that made a real difference for them, because they were able to predict what capacity was going to be, and that enabled them to make sure that they have the right staffing levels. So that in itself was real value for the organization. I don't know if you've got one from a financial perspective where it's driven money there, Jim. Well, I I have, maybe not to the same extent, but a very specific example that I just remember. It's more of an anecdote, really, is, We did some work with a company who essentially track, legal professionals and effectively rank legal professionals. I didn't even know that this existed, right? But there is there was a directory of high up legal exec atives and, and, you know, lawyers and so on. And it charges the companies that they worked with and the value that they brought to that organization, and it literally was. It was almost like a Michelin Guide. You'd buy a physical book and you'd look through this stuff. One of the organizations that we worked with pivoted that into an embedded analytics solution, and I remember specifically we put a, a scatter plot in front of them. They wanted to go, you know, quite technical with their visualizations. But the the scatter plot, if you can imagine, had a time series dimension. So you could see the reputation and the billable value of these lawyers and, you know, sort of, financial professionals. And over time, and literally with Tableau, this would animate. You'd see the positions of these individuals change and there would be like a little comet trail to indicate where they'd come from. It it's difficult for me not to sort of evangelize, but they were literally blown away had never seen anything like this that was taking effectively the content of this entire directory and exploding it onto their screens in a way that was immediately understandable. That, that in itself, it was part of a suite of dashboards, but that allowed them to monetize that service to an extent they hadn't been able to get close to with the directory, which was effectively we issue three updates to our directory once a year. That's it. You pay. It was a whole different ballgame. So we I will have to check to see whether or not we've got any more specifics because I understand that people might be interested in, okay, you went from X to why, what did that actually look like? But I can do some poking around around that and see if we've got some examples we might be able to share. Wonderful. Thank you. So what I'm thinking is we will give you two minutes back. As I said before, we have recorded this session. We will be getting a copy of this webinar out to you. Please feel free to share that with colleagues. Thank you so much for your time today, Jim. It's much appreciated. And we hope to see you on future webinars. Thanks for your time, everybody. Take care.