Hello, everyone. Welcome to our session on how to build a metric tree, a live billing session that we'll be doing with you, over the next sort of hour or so. Nice to meet you. I'm Ollie Hughes. I'm joined here by Mitra. Mitra is our head of customer success. I am, the CEO of Count, and our goal with the next hour is to talk you through everything you need to know to go away and build highly, clarifying metric trees for your organization. This is the second, webinar we've done in this series. Our first, webinar we did was with MoonPay who talked us through how they went about leading their organization to a clearer view of performance and elevate their data team to being a, kind of a problem solving work, force within MoonPay by using metric trees. This session, we're gonna get into how do you go about actually building, a metric tree. We'd love to hear from you as we go through the session, so please use the q and a section, which you can see there. Rather than the chat, ask questions, as we go. Our goal is to spend about thirty minutes walking you through exactly how to build a metric tree and then have time remaining to answer your specific questions that you can get going straight after this in yourself. So please give us the questions as they come. We'll make sure we cover them off as we're going through. Hopefully that makes sense. Mitra, do you wanna say hello? And we can then get into a bit more of the data and a bit more the plan for the day? Hi. Yes. Thanks, Ollie. Yeah. I head up customer success account. I will be taking you through more of the practical build, and thinking about the different stages that you go through to create a metric tree. I think we'll kick off with Ollie. He'll give you a nice introduction, and then we'll move on to my part and then look forward to your questions at the end. Exactly. Well, so let's kick off. One of the things we're really grateful to you and which we love doing at Caltech is making our our kind of sessions like this much more interactive and make sure we're answering the question that you have coming into these sessions. And one of the things we asked you, before this session, if you if you if you time of a week or so ago, was just to ask you how many dashboards are your team looking at on a daily basis to understand what's going on in your organization. And this was the results that you sent here in the the bottom here. These are the results of the survey from you, I hope, people who are here. It looks like if you look at the results here, what really if you look at this in detail, the key stat is over half of the people in this webinar, over half of our organizations, people are looking at over six dashboards or more, to understand exactly what is going on. In fact, here, eleven plus, people looking over ten dashboards, quarter organizations are looking over eleven dashboards to get to value. This is the problem. How can you possibly understand what's going on in your business? What really matters if you have to look at this many reports to get an answer? And to back up this problem even more to set the scene further, we know that, one in four day leaders believe they have a good understanding of their growth. Well, that means three quarters of businesses don't really understand what's driving revenue in a really clear way and that a lot of a data team's value, a lot of the data team's work is spent, on things which aren't actually moving the needle forward for the organization. Only eleven percent, that means, of the work a data team is doing is actually moving the business forward rather than just maintaining visibility and trying to make sure people are clear. And that's really the goal of this session. Our goal for metric trees, we know metric trees can do so well, is they can turn a dashboard mess like this into something which is clear, which is driving clarity, which helps the organization understand itself, understand what the opportunities to grow are, and elevate everyone's understanding so that you can problem solve the biggest bottlenecks in your business. And this metric tree here is the one we're gonna talk you through building today. So now I'm gonna hand over to Mitra, and we can start walking through how do you go about building a metric tree like this. Thanks, Ollie. I'll take over my the screen share. Great. So we see customers every day making use of metric trees like this, consuming these metric trees, applying them into their improvement cycle, using them to identify problem problems and opportunities, using them as a launching pad to then explore what these opportunities or issues are, and then being able to affect change. Being able to affect change at lower levels will then filter up and actually affect your your key figure at the top, your North Star metric. But I think one of the common questions, or hurdles that I find people come to me with is, like, where do we start? Like, we'd love to get to this point, but, actually, we're looking at a bank canvas. What what route should we take? So today, I'm gonna hopefully show you that just by breaking this process down into smaller steps, it's a very logical and almost enjoyable path. And I think by following this, you get benefits not just you don't just end up with a metric tree, but, you end up through collaboration and through discussion about your important metrics. You just end up in a better place as an organization. The stages that I will touch on are the planning stage. Like, what do we put on the canvas first? How do we how do we just kick off? Once we have our plan in place, we need to move on to data and, like, interrogate that data. Is it fit for purpose? Then we get to the build part, and this is the part people focus on most often when I get questions, but, really, that's the textbook bit. Like, if we plan and work with our data and agree what we want our end goal to be, the build, is is fairly straightforward. And then we'll look at using it. Right? Our end goal is not just to have a metric tree. Our end goal is to drive business improvement through, incorporating the metric tree into our everyday life, into our processes, and embedding it in the organization. So this is a really important step. Part of that is ensuring that we are actually driving change and and improving the business, and we need to be able to monitor the work we do off the back of this metric tree, and be able to see, actually, is it working? Did we get this right? And we want to iterate. We always see this as a living document. We would if you go away next week and hopefully build a metric tree, I hope in a year, it doesn't look like it does now. It needs to grow with your business as as your business evolves. Sounds good to me. What to get through? Yes. Cool. So let's start with our blank canvas. One approach that we often take, and it's not the only approach, is to build backwards. So if we're thinking about a metric tree, the first thing we want to do, and I'll just use a sticky note here on the canvas, is just decide what that North Star metric is. What is that number that we care about? It has to be a number that you want to understand and that you want to be able to improve. This could be at an organizational level. So for our example, we're going to be looking at revenue. We think it's fairly universal across a lot of organizations, but it doesn't need to be across the organization. This could be, at a department level. It could be a project level. Whilst Ollie mentioned before about the the clutter of dashboards, we rarely it's it's not that you just need one metric cheaper organization. We can think strategically, like, about having a small suite of them that covers your organizational metrics, your departmental ones. And then what comes next? It's just breaking this down. Like, we need to decompose this metric. So we can start to do that by just think about, like, what does our revenue what's it made up of? It's new it's made up of new business. It's made up of expansion, and it's made up of churn. And we can keep going like this. So we start to suddenly see the next level of our metric tree. And then we just go again. Like, what makes churn? What how can we break that down? How can we break down new business? And the idea is that the lower you get in your tree, the the more likely it's more likely that you're gonna be able to influence that metric, and that's where you wanna be. You want to be able to have a suite of metrics that you can influence that then the effect of which will ripple up and affect the North Star metric that you care about. That's right. You wanna go from sort of leading metrics, things you can influence, and then as you got the tree, they start becoming the outcome. That's cool. Yep. Absolutely. So here's there. So a couple of things with this. Mine is symmetrical. It's the perfect looking tree. Yours definitely does not need to look like this. Some branches might be longer than others. It really it just needs to reflect your business. A few hints on this is collaborate early. The best thing you can do for to ensure that this is reflective of your of your organization is to get people in who have, more domain expertise than you, make it cross functional. I we rarely build a metric tree that doesn't impact more than one team. And also look for a senior sponsor because if you can get by in at this stage and allow them to influence what this looks like, the likelihood is you'll have an easier time to embed that, later on. And it's fair to say we Rachel, I think it's fair to say that you don't need to make perfect enemy of the good. You're probably not gonna be able to know exactly the right the even the right structure of the tree until you start getting some data in any way. So you just wanna get to a a relatively agreed sensible structure at this point. And as you go through as you get further into the process, you can refine it. You don't need to get stuck at post at no level because, actually, you probably can't have the information to make a better decision easily at this point. You can get stuck in a kind of a a kind of a a kind of a stalemate. So you wanna keep moving at this point, and just get through the process. And as you pointed out earlier, iteration is the key to this and getting buy in as you go. And that's probably true of any report you're building. But if there was a dashboard, you through and scope well, and then it's ready to move. And that that's probably true of every everything you build, it should be following these principles as well. Yeah. Absolutely. Then keep up momentum of of use your best guess, but involve others. Document your thinking so you can return to it. Don't worry too much about like, this is quite a mathematical link. These all add up to the next tier. That doesn't need to be the case as long you just need to think about influence. If you can find some path of influence between your metrics, that is strong enough for you to be able to put this into use and to see an effect. The other thing is don't over complicate it. We have way more numbers than this, but there's real power in being being able to simplify it into what you can actually, affect. So, yeah, keep it simple. And and also with pointing out that this isn't about just doing a tree. We talk about metric trees, but, actually, at count, we talk about metric maps because we think what we're really talking about here is contextualization showing the relationships between metrics. That doesn't have to always be a tree. A top down thing obviously works very well for revenue, but there are many processes and things that you can do apply this to within an organization. And, yeah, these are great examples here. Yeah. Absolutely. Use the extra dimension of of the cameras here to really, really map your metrics out and use the layout that works. So here, for example, this is a customer life cycle map where we have an awareness stage, and then people enter this funnel where they become interested, engaged, and and so on. Initially, I started this out as a long linear line, and then I realized, actually, all these all these points have drop offs. Like, no one no one leaves this cycle. They never become unaware of of of my organization, and they jump between stages. Right? You might be a customer, but you might still be engaged or interested. So I use that layout of this of this loop to be able to then visualize these drop offs as well. Here's a checkout flow. So this is a, like a a website or an app and the the behavior of users when they go to checkout. There's loads of different branches. There's loops that add back in. There's really no rules. The import an important point here is these stages aren't the be all and end all either. These links are just as important. So we can show the state that people are in in each of these squares, but we can also show conversion rates on links. And that's really where we get power out of this. Like, we are linking data, and we're showing it in context. We're not just relying on, you know, modular dashboard. It's all about making complex things simple. And, actually, one of the things I we've been finding our customers is sometimes, actually, we don't suggest a metric trick is the first thing to build. But, actually, to go into, you know, because it it because there are you know, you can split the metric two different ways. Instead, maybe there is a a process flow map of the business or a sales funnel or, as you can see, their checkout funnel where, actually, pages of the funnel are kind of very understood. It's very much just determined by the way your product works or the way your website works. And so mapping that kind of thing could be a great first use case to get through this idea of metric mapping because the structure of the metric map is not in discussion as much. Everyone understands the process already. It's just about visualizing it better. Yep. Absolutely. And, actually, we'll show you some examples at the end, which I think most of these have been built up into real examples, and it'll be quite nice for you to see the final version. Okay. And then we wanna add data. Right? So we've got our plan. And I think the important thing to remember is, particularly with the the power of the canvas, we can bring data in alongside our plan. Like, this doesn't have to be a document that lives separately. This is very much the bones of our work, and we want to keep it. The first thing I do is add some more depth to our plan. Right? So we've defined our metrics, but we also need to think about the additional dimensions, that people will want or need to use. So I've split this up. I usually split this up into two. So there's time frames. I think most metrics will relate to some form of time frame. And how will you want to segment the, the information? Like, are there other dimensions we need to build in? We need to make sure that's all in the data before we build. For this example, I have some post it notes here. So for time frames, I actually found there was there was a bit of a complexity here. So ARR with a snapshot, it was like a a point in time. Like, what's our ARR yesterday? What's our ARR last month? Whereas everything down here were events that happened at a single point. So we need we need to define, like, a time frame of these events that we want to look at and aggregate, and we also want to see snapshots in time of our ARR. And segments, I've just pulled out geography industry. And, again, it's a really good one to collaborate early. You know, we've all shared data outputs, and then someone's asked if we can split this by x, and we haven't thought of it. So, again, trying to get buy in and and sign off on these. And then have a think about, like, do we want to display these upfront, or do we just want to allow people to explore them? So, like, my time scales, I guess, we're we're displaying them. We probably will let them explore. But my segments, I don't want to lead with those, but I want to build in functionality that will let people explore. And then we can bring data into the canvas. So here, I have some raw tables. So I have some data relating to companies, and I have data relating to deals, one through this company. And I start using SQL styles just to build out this analysis. So I've done a simple join on this information, and this will actually give me a lot of a a lot of the figures I need. I will be able to aggregate from this. But I also need to build in some snapshots in time of ARR. And for that, I've, had to consider the fields I have and the flags and the time frames and create some logic. So this is a really important place to be able to come in and define your metrics because ARR might be quite straightforward, but you might be dealing with something else like cost of customer acquisition, for example, or customer lifetime value or something that just isn't as well defined. People might be using it differently in your organization, and here's a really good place to get people in, tag them, get them to work with the data, and agree on a definition that you can use across the board. Yeah. This is this is great. This is kind of just the one thing I think is really powerful about this is you as you pointed out here, here's a metric you may have built, a kind of an early version of the a metric a metric for the tree, but you've now of the whole thing. You've got the raw data, the the the analysis, the metric, and the structure, so you can really start to iterate through all these different elements together, and then it all cook as you build out Yep. Absolutely. And this is where you can test it. Like, here, for example, I've just created a big number, and I've put some logic in here, and I've connected my controls that I want to be able to split on. And this is this is what I believe is my current ARR. Right? So I could I'll just bring this down here for ease. So that's great. But, like, is that right? So I've just gone and, like, gone into an another system where I know that we have our kind of single source of truth for ARR, and I've just screenshotted it and thrown it in. And I'm like, does this match up? This is pretty important. Like, there's no point in me making a metric tree with a different number on it. And we can see that it does. We can also look through rule level detail. You know, we can see there's duplicate companies in here. Like, is that okay? Like, we could go and check. We think, yes. Actually, it is okay. These are just different types of deals. But, we can do the checks that we need. We can also check, for example, my snapshots. Maybe I just want to do a quick visual. And here, I would want to check that my monthly snapshot from June, oh, it's one point four million. That doesn't match the one point two million. Actually, when I look into it, that's fine. It's just because I'm defining what the ARR is on the first of the month. I'm not defining it on yesterday. So maybe there, I put, like, a red Post it note to be like, you know, caveat this. This is important to know that there's a discrepancy. I'm okay with it, but I wanna keep that decision somewhere. So maybe I maybe I put that in a frame, and I and I document it well. And I think sorry. If I can just add one thing, Richard. I think it's helped to recognize that during this whole process, as you show the screenshot, actually bringing in the wider company into the building of this process as early as this, not only with Post it notes, but also to help them understand and look through the row level detail, understand the metrics are different. It's all buying to the trust, all the buy in, helping people feel part of the process is one of the best ways you can get to better answer, but also get buy in and understanding about what we're doing here. It's an amazing cleansing experience, if that makes sense, as you're kinda remodeling and rethinking through the business and checking it as you're building. And that's that's true of any asset, but particularly in case can you try to define, in this case, an ARR metric tree? That's gonna get a lot of people interested. So, getting them into this and checking with you, getting the logic right, showing logic before it gets locked down into into a data model is a really, really powerful thing regardless if it's a metric or not, basically. Absolutely. And from a selfish point of view, it gives you so much more confidence when you do push that out there that this isn't just you. It's not your decisions, and and everything is gonna rest on your shoulders when it goes wrong. Like, this is a this is a collective decision, and you can logically show why you've come to this point, and you can retrace your steps as well. Okay. So we can do some tests. Like, I would maybe just put all this in another frame and just keep the the decision making going. And then I would just build up my figures, and I would apply them to my tree. I will just show you one I did earlier. And what you will end up with is just like a rough outline of, like, big numbers. I wanted to show, a time series as well. And this is, again, an opportunity for you to be like, does it make sense? Are these all, like, worthwhile being here? Are there are there any, anomalies we need to look at? For example, here, I I realized that whilst whilst our kind of new business expansion, contraction, and churn all affect this number, There is also, like, an existing ARR that that really needs to be added in to make sense of that number. So I decided at this point, actually, I'm gonna add in another metric. It's not gonna be a whole card, but I just want full clarity on that. So you can evolve your plan as you go. That like Ollie said before, just keep it moving and see what it looks like. And at this point, what you've got is you've got your raw data. You've got your, like, analysis, like, business logic in SQL maybe or whatever it or low code if if you if you have all the whole report from raw data to this final, you know, MVP output is all there and easy to iterate, to walk through, to check. Back back just another version of the same thing we've discussed before. Like, now you've got the whole thing. Nothing's been necessarily locked down yet. You can you can got all that to play around with, get right before it goes into production, and then it's harder to change. So it's that that's one of the big powers of this approach of this of the cameras, I guess, is that you can do all steps in one place. Yeah. Absolutely. And this might be an easier discussion for some audiences. Like, this might be the point that you want to bring in, some senior because they're, you know, they're gonna fully appreciate it in this format rather than in the raw SQL form. Yeah. Actually, that's what MoonPay said in the last webinar we did was exactly that. They built it internally as a data team first and then launched the MVP to the business to get them to see go, wow. Like, how cool is it all to see in one place, and then they productionized, which is cool. Nice. Cool. Okay. So we've got our plan. We've collaborated. We have made decisions. We've tested our data, and now we need to we want to put this into production. And we have a decision that we can make here. We can either build this directly from our database, and this will be essentially, like, a just a tidied up version of what I've just shown you. Or we have the option to actually build this model into our semantic layer and create a catalog. The database version is, like, quick and easy as a first draft. Maybe that's an option you want to go with. But long term, it's not accessible or reusable. You know, it's sitting as canvas, in in this form. The benefits of modeling this out into the semantic layer is that it's essential source of truth. It's locked down. You've already gone through the deliberations of, like, how we define these metrics. Metrics. Like, why not put that additional layer, of of security on top and be like, let's let's put this in a catalog. Others can come and use it, and we're not gonna worry about the calculations that they're using to get to these metrics. The joins are already predefined, and the aggregations are standardized. So taking a step out to actually put this in the catalog can reap a lot of benefits, which we will touch on as we as we go on. And then we want to complete the build. So as I said, the earlier stages are really the work. Like, by this point, you have a plan. You have defined metrics. You've already tested the data. You've either created your catalog or you're gonna work from your SQL cells. So the the way that we would typically do this is I would create a card. So this looks sort of similar to the one that I showed the rough outline of. It's tidied up a bit. It is just a series of visuals and text fields, and these are all built from my catalog that I've created rather than SQL cells. So, the SQL's all happening under the hood. This is just a local environment. Once I had a finished card, I would simply oh, excuse me. I would simply, look to copy that, and duplicate it, and then I could just update the filters and, and change this. So but it's fairly straightforward. I just go in. I'll do one for you. Let's see. We were gonna look at expansion. So we build up like this, and then we add in our connectors. And, essentially, we end up with our final tree that looks like this. There it is. This is this is exactly the same as the MVP one. It just looks substantially more clear. It's it's it's formatted well. It it looks great. It looks like a finished article. And it's running on the count catalog and then the the semantic layer catalog too. Yeah. Absolutely. So, I mean, there are design elements you can bear in mind. Like, I've used I've used colors to kind of group the branches. I didn't have to, but it helps to segment them down here. I've used some conditional formatting for, looking at changes from previous periods. I mentioned that I wanted to add that extra number in. I've done this here, just formatted it in a different way. And then maybe the easiest way that I could look at this is if I go into presentation view and just go to the right tab, we can see my metric tree here. So we'll see that I've made use of this this overview panel, just to be able to introduce this document to whoever's looking at it and also to, put my filters in as well. So people can now come in, they understand what it is, and they can decide to change, any of these that they wish, and it will update. I also mentioned that we had I had a caveat. So, like, my total trend line and the ARR from yesterday doesn't totally match up, and I'm comfortable with that. I know why. It's just a methodology difference. And I've just put a note that I'm gonna address this in the in the next update. So it's nice like, it's okay to present this as a living document, I think. It's nice to highlight the restrictions as well. And I think that gains that level of tail gains trust as well. And then actually using our metric tree then. So we know from what we said before, the idea is that we have a number that we care about, but we don't have clarity on actually what we can do to to change this number. By looking down the tree, I'm looking at the trend, looking at the, the change percentage as well. We have an idea further down of what's doing well and what's not doing well, that might be affecting this number. So we can see, for example, between May and June, there's been a bit of an upturn in ARR after, after a period. And we can see down in these corners, we have a lot of red. So we can see that we have a lot of churn that's just started happening, and we have a bit of contraction. Initial deals seem a bit steadier. There's a little bit less in the in the latest months, and expansion is due now. So there's a lot happening actually. This this upturn is actually looks like a huge amount of expansion, but then brought down a lot by churn that's going on at the same time. So once we've looked at this, we've discussed this, this could be, you know, lots of people in a cross functional meter departmental meeting. We can look into it even further in situ. Right? So this is all based on a catalog. So if I were to click on my chart, I have the ability to explore this cell. And this is because it's being built on, my predefined model on my catalog. So here I can see I just alright. This means that you can now slice and dice this chart in any dimension, any time frame, any anything you wish just to dig into what's going on because all powered by, the logicalness of the semantic layer, which means you can just pick any metric dimension and split this chart out. Yeah. Absolutely. So I'm doing this in, in a new single cell view. I'm not I'm not in a canvas. I can't pan around. But I am able to, yeah, do what I like with it, and it's not gonna affect the underlying, report. So I, I've just split by geography, and I can see, okay. There's like, UK is predominantly driving this. I could also look at that as a table if I wanted, to see the underlying data. Let's stick with the visual. And if I wanted to save this, I could save it as a canvas. So let me just save it back there. I've just changed this visual. That's okay. And then what that's gonna do is push this into canvas, where I can do further work on it. So, if I get back to where it was, so it might be now that maybe I am a local user. Maybe I've just discovered this. I've explored my own, and I want someone else to go and, like, dig deeper. I could tag them, to come to have a look. We have a lot of success with, I'd say, low code users that then have the ability to identify issues or trends and tag others that maybe want to look further into the SQL, like do or bring in additional data that's not in your model. So, you can build a fairly simple catalog, and it still can open up, to a lot more. I think I saw Ollie darting around in this canvas. Yes. I'm I'm into this canvas now because you've you've obviously you've taken out you've seen a metric which looks strange. You've exploded it, explored it a bit, and then you've said, I wanna save this to a new canvas, which is now gonna be more of a kind of working environment to problem solve what's going on. And I've joined you in here, and I'm now bringing in other metrics and exploring the data myself. I've actually brought in something here about expansion to see if the, you know, what the geographies which are driving churn. I need to say geographies which drive the expansion. So let me just run that query here, and you can see actually that the expansion is coming for a lot from the UK, which isn't really where we're getting our our our drop as much. Actually, that's not true. The UK is dropping too. So it's not that we're changing a geography split here. We've got churn and expansion happening in the same region, so that's not necessarily the answer we're looking for. But at least we're digging into the numbers together. We're cloud we're working in a real time and collaborative environment. And maybe to finish this off, I'm now gonna make bring in a template, which is just a kind of a a slide in this case, a slide presentation, and I can then drop in that chart to say, expansion, region isn't a cause of journal, whatever we've come to realize. So I can tell the story that the metric tree points to and help us get to a decision point, a presentable outcome, all built off that semantic layer. So all the metrics we're looking at, all running off the same rules, all have the same definitions. So our exploration and our metric tree are all working together through one exploratory workflow. Cool. Thank you. And then, I guess, just just a couple more points on on embedding this and and making sure it's used. We've touched on our on some of the features. Right? So we've talked about we have our overview to allow people to dig into the data. We've seen the explore from feature, which is great if you've built from the catalog. We also have the ability to set alerts. So we could set an alert on this entire frame. It could be at a a regular cadence to the people that you want, to put this in front of. So that could be monthly or quarterly to a a set of stakeholders. What that would look like here, I have one. This is just a Slack update, or it could have got it via email, and it sends out an image of the of the current data, the current tree with a link, to this report. So that can be really powerful, to get it in front of people. Alerts also work based on thresholds as well. So it might be that, you know, it's great that we came in here and saw a problem. There's a few problems. But, maybe we want to automate that and not have to monitor it manually every time, so we can set up an alert based on data reaching a certain threshold. So we could do that on each individual cell, and we would get an alert to say, you know, it's past the threshold that you had set. Maybe you should have a look at it. The other the other thing to consider then is so you find an issue, and then what comes next? Like, that could be very ad hoc. That could be, let's quickly go and dive in like Ollie and I just did, or it could like, let's formalize an initiative to actually dig into this problem, explore it, do our exploratory analysis in our Canvas, and then we could even copy this metric tree in to try and monitor that change. Right? So maybe we've explored, we've made recommendations, we've made a decision to change something in our business, and then we want to keep tabs on actually, has that changed? Does it affect this number? Does it affect any of these numbers? So I think keeping that within separate initiative canvases is a really powerful way to monitor your metric tree. Like, you want your tree to be fairly locked down, but, essentially, you could either have activity going on around in the same canvas, which is fine, or you could keep a project of of of initiatives coming off from this metric tree. Thank you, Mitra. This is this is great. I I think we should move on and show a bit more more examples and get to some questions if that's okay. But, hopefully, that's been helpful. You've gone through in a classic, if you know that if you're in the UK, a blue Peter, here's what I made earlier, snapshot of how to build out. And I think, hopefully, that's given you a sense of confidence that it is possible, that it is an evolution of journey. And, actually, a lot of the principles we're applying here about planning, bringing in stakeholders early, iterating MVP where you can change the modeling, see the data, and the end result in one place works for any any report you could build in count, be a metric tree, even a dashboard. So we just think there's some very sensible principles here that I hope you can take to any of your work that you're doing. But before we move on and answer q and a's, I we wanted to show you more of more examples of what metric maps can be to help you see that this isn't just about metric trees and it's not just about revenue. This principle of getting more clarity to your business in the key processes and areas of business works and drives improvement so you can explore the data. So we wanted to show you some of the things that our customers have built, and just quickly walk them through for you so you can get inspiration and see this is not just a kind of a a top level problem only. So let me share my screen. And while, maybe you can throw in some questions if you have them. So just give you a few minutes to think of some questions for our q and a. I'll go through some more of these examples, that we've got. I'm gonna start actually with, finance one. You know that finance teams particularly love metric trees for letting the growth model be clear, but it's also powerful for showing the process of cash flow, the ins and outs of cash going into your business, what are the key buckets, what are the key phases, how does this all fit together, what's the equation there. This is one we've seen people really getting excited by within finance teams. Here's another example. This is actually one that Mitra showed earlier when it came to the Post it note scoping. Here's a customer life cycle where you've got about fifty or sixty metrics all contextualized in relation to each other, showing you your top level channel activity, the conversions through, to different stages of customer, life cycle, be that converted customer to all the way through to advocate. You can see the drop off rates. You can see the row level detail of every single customer, every single stage, all in one very clear system view, which, you know, you can't get this it's obviously not a simple process, but it's as clear as it can be. Another example while you're throwing in some more questions is one which we were inspired by Duolingo. Duolingo have done some amazing work on, customer retention and understanding that the ecosystem of customer retention, how they lay that out, as you can see here, going through, different stages of that. Another example is, similar to the checkout when we showed an onboarding funnel. So you can see the stages of user journey through the through onboarding. And then finally, we've got one here which is back to another version of the metric tree. So, hopefully, that that's helping to see that too. Right. Any other questions, Mitra, that we missed to cover off, or should we jump to some q and a? No. I think we can go q and a. Brilliant. Okay. Maybe I'll let you share your screen in case we wanna have people, jump in Yeah. With anything that you you covered off. Let's let's start here. So Shola asked a question early on, which was, when does it get too granular? I guess the metric tree. How does how does it get too granular? How do you know when it to stop going down the tree down the funnel? What would you say to that, Mitra? How do you stop making the disturbance explode all the way down hundreds of levels? Yeah. I think it's a a really fair question. I think there's, I talked about simplifying earlier. They're sort of simplifying across, and then they're stopping at a simple point as you're going down. I think, really, there's no rules. I think you will naturally get to some metrics that you will think we we can influence these. Like, we already influence these. So we're we're sitting at a level that we understand, like, if this is the needle we need to move, we're comfortable knowing how to do that. And I think that's probably the assessment that you you would need to make. I think if you keep going and keep going, you will suddenly realize that you were just getting into such naughty detail that you'll almost end up being like, this group of ten metrics in the bottom corner is what I need to move, but, actually, I understand these as just the one above it. So I think just contextualizing what you do in your organization to affect these will help you on that. I agree. That's it. You'd be also surprised at how powerful even a high level metric tree like this is only two levels. There's even even the clarity of, like, what's driving performance at this level of detail is surprisingly powerful. So I wouldn't even go that deep to start off with, but, obviously, you can you wanna go ideally, go as deep as Mitrice you can as the lever that you can pull, but even starting off with this level of clarity is a really powerful place to to to start as well. So don't, again, don't let, perfect be enemy enemy of the good in terms of getting a perfect tree before you launch it. This a question from Isabelle, which I think is quite quite a good one, just about the product like, the overview. How do you get filled on the side rather than being hidden in the canvas? Can you just explain the overview feature we've got in the product? Absolutely. So, when you create filters, they initially live in in your canvas somewhere. So these are the filters that relate to my metric tree. I could put these I can make them bigger and frame them and put them on my metric tree, which would be one way of doing it. The benefit of having them in this overview panel, though, is that they come with me wherever I go. Right? I can zoom in to here, and users can still access these controls. So the overview panel, initially, you might be used to seeing the data panel. The overview panel can just be clicked on here. And this is a rich text area, and you can add to it by either using a forward slash or an at. So forward slash is what we need. And here I can either add text fields. I could add checklists. But at the bottom, there's an option for control. And once I put my control in, it will ask me which one I want to select, and I would select it and it will appear here. And then I can drag these about. I can do whatever I like with them. But that's how you get them into the overview panel. You can also link, which is really nice. So if you wanted to, if you wanted to, I don't know, jump between frames on this canvas, then maybe I just do an at sign and say revenue metric tree, and then this is now a link. And when I click on it, I'm going to jump I mean, that wasn't a big jump. Let me jump further. I'm gonna jump between my Canvas. So it's it's it's a panel that helps introduce your your, Canvas, but also help users navigate and and use it. That's great. And it's relatively new. I would say, like, if you're looking at that and you've used Kite and you're like, I've never seen this before, it's, like, fairly new, so don't worry. Evolving ever getting better. Wonderful. Thank you. Tosin is asking, is the Canvas that you we've walked through available as resource and template? I think so. We doesn't those seem reason why not. I think that'd be great. There's also lots of other resources and examples on our website. We've actually have our another training Canvas talking through every step of this. So a version of this total will definitely be available. We'll make sure it's available in the in the off I can they wrap up materials we'll send to you afterwards. Thank you for that. I'm glad it was useful to walk through. And then there's also been a question here about oh, sorry. One thing very quickly. Just as well, if you go into templates and count, you will also see, under some getting started guides, how to build a metric tree. And if you wanna have a look at that, that gives, we've probably gone into more detail here, but this, you can add to a canvas, and it will talk you through, some of these stages. Yeah. And you might find that useful, but we will make this available as well. Thank you. That's it. That's all good. Definitely help you. Don't worry. And then the other let me the question I've got here, is then a very important question is this is a big jump from dashboards. How can I wean my leadership on their familiar reports? I'm guessing I'm asking how do I sell these internally? That's a great question. Like, maybe the question because that that is ultimately the outcome is that you want the organization to get this, understand it, and then be clearer in their own thinking because of this metric tree. What I would say before I hand over to Dimitry is just, which can be a kind of helpful step is you can actually also build dashboards and count. That may be not the the ideal use case, but everything we've done into this process is it makes means that count is a dashboarding tool, but then you can then lean into all this, kind of new way of visualizing your business as well. So we aren't just a metric tree tool. We leave in, giving you all the all the tools you need to express the business clearly. If that is a dashboard, so be it. So you can absolutely do that. And just a metric tree buy in, what I would say, each way could you get maybe embellished with more of the customers you've spoken to is start as we've done it. Talk them through, the the metric process that you've laid out, the planning section, or as we mentioned before with MoonPay in the previous webinar, we've done on the just build one and then go, isn't this better? Interesting. A great example here, actually. Maybe you could talk through this this the a b experiment meter. It's pretty powerful way to show, maybe you could even, you know, give them the choice like this. Yeah. I I mean, yeah, maybe I don't know if you would want to go to the extent of building both, but, actually, it was very quick to build both of these. So this is essentially the same metrics in a traditional dashboard layout. And then here, we've split this into a metric tree view of the key numbers, and then we've grouped the associated metrics together to give context. And I think it's fairly hard to argue that this isn't a more useful view than one where we're just pushing all these into modular, like, silos. So that can be a really good way. I think, like, just educating people what I think it's I think it's quite clear, but just explaining, like, this is the number we care about, and this is how these are the influences that and we can break these down, and then we only have to start caring about the ones at the bottom. Like, that's a really powerful story to someone who's trying to make decisions about, you know, how to drive value in the organization. But also getting like I mentioned before, like, trying to get senior sponsors in early, it's important for them to influence your metric maps because they have the best domain knowledge of the, you know, the direction your business is going in. So I I would I would not shy away from that, or just build one one day and, you know, steer it. Mitch, if you could leave up your screen right now, because I think actually Bruno's got a question, which I think is kinda related, which is about data cuts. Okay. The question is how what's the best way to sort of help people understand the impact of different dimensions and filters? You reckon like, he's pointing out here that he recognizes that, obviously, we can filter the metric tree as a whole and segment it as as a filter to the the board and the overview section, or we could just split the metric tree down by those dimensions as the hierarchy. But, like, what's the best way to showcase the underlying segmentation? I think, actually, all one of the ways we would suggest this, Bruno, there's there's no right or wrong. I would just wanna say annoying knowingly. But one way is, as Mitra's proven here, which is to say, I'm gonna give you the top level metric, but I'm gonna give you the dimensional splits as a call out card either within the card itself on the side so the influence of different plays here. You can see from on the bar chart next to this, on the the number of active users, the first top purple card there. But on the left, we've then got that broken out by different dimensions so you can see what's really driving that. That's a very natural way where you're sort of blending met a dashboard of use of which are often just dimensional splits of a metric and the structure. That's one of the benefits of the canvas is it makes you got that flexibility to go as deep as you want. So that's that's one part of the answer I would I would I would suggest. Anything else, Mitra, on how to better show the cuts and then and the influence of different segments segments in one place? Just I think that's a good explanation. I think, the other the other way is, like, this user breakdown could be a mini metric tree in itself. So this could sit at the side and just be, you know, we could reform at that, and it could live. These cuts could live within your main metric tree canvas. Or if there were if it's important to have full clarity on these, you could create different versions of your metric tree to suit different meetings, different departments. Like, it's all about how you would use us. But, yeah, I agree. Like, putting them in filters is great, but it relies on you getting to that point. So if you believe that the people who need to see that aren't going to take the time to do that, I would suggest breaking those out, visually, even if they sit in another frame would would be very powerful. Mhmm. Yeah. There's no there's no, restriction to say you have to have a metric card like we have here with a number, a percentage change, and a and a time series. We have customers we're working with right now, actually, where the metric that we're building for them has three or four different time series in within one card. It's like a mini dashboard in the metric tree, and that just gives you instant cuts and slices in the way that you'd wanna see as well. Like, the the the key is the is the contextualization of the of the metric by putting them into this kind of metric map view. What you put in those those metric views, you can you can change. You can put their own metric cards. And you and it might be that you start simple and then you add to it. Like, you realize that in every every time you're in a meeting talking about this, every you're filtering on all these different things. And and if that's the case, then take that next step. Like, in include those cuts in your visual. This is great. Well, I think we're at the end of some of the questions. Thank you for those questions. They've been really on point in terms of, like, a way to make this come to life. I hope this has been a useful session for you. We'd love to hear your feedback. And whether you have any other questions on this, we'd love to make sure we're constantly giving people resources to make this leap into a much more clearer way of seeing your business and using metric maps as a way of doing that. You obviously can get going with count. There's a free for a free account, right now to start playing around with this in the canvas, start mapping out your metrics, get your team involved, and start to build out these these, these metric maps and start to explore the data for this purpose, along with all the other things that you can use count for, like data exploration and even dashboarding. We'll be sending out more resources after this, including maybe this canvas or versions of this canvas to help you have that sort of step by step guide to building out your first metric tree. Mitra, thanks so much for going through all the what you've learned with other customers and and tell anyone the women how to go about this. Yeah. We hope you've had a good session, and we will let you know let us know how it goes. Thank you. Bye.