Building the Metrics that Every Product and CS Leader Needs to Drive Growth

Every product and business leader looks for ways to measure customer experience, accelerate adoption, and improve customer retention of their product. To measure the right things requires gathering data from multiple systems so cross-functional teams can automatically gain insights from data to identify growth potential or churn risk. 

Integration platforms (iPaaS) play a key role in addressing the many app integration challenges product teams face in modern SaaS companies. In this session, Matt Graney, VP of Product at Celigo shares his expertise on what metrics are important to track and how Celigo is using this framework to measure customer health and drive growth.

Topics include:

  • Customer experience metrics that matter
  • Integrations that built Celigo’s customer health score
  • Practical tips for getting started with an integration roadmap
Full Webinar Transcript
Hi, everybody. Good morning. Good evening. Good afternoon, wherever you may be throughout the world, and welcome to this presentation. I am Matt Graney, a VP of Product at Celigo. So we’re going to talk today about building the metrics that matter. I think you’ll find that the lessons that we cover today will apply to you no matter what sort of product you’re involved with. In our particular case, though, we had the advantage, of course, of being steeped in the integration world and looking at the role of integration in terms of getting the data where it needs to be so that we can better understand the state of our customers, their overall health, and use that effectively to drive retention and growth of our business. We are based in San Mateo, California, we are series B start-up and let’s get started. So. The first thing that hopefully is nothing new to you is, of course, that product leaders need to be thinking about retention. This isn’t just about building new features. It isn’t just about working on technical debt or whatever it might be. The goal always has to be focusing on retention. And according to this study, it costs four times more to acquire a new customer versus selling to existing. And this really has always been the premise of a business in that it’s good for the buyer and it’s good for the vendor because everything is right-sized. Hopefully, the customer begins by only buying what they need. But there’s this ongoing expectation that if the product is successful, that it will expand. You’re already into an account. You’ve already planted some seeds, laid, hopefully, a solid foundation, and there is a great growth opportunity from there. And compare that to the cost of sales for a new deal, having to prospect, do a lot of marketing, whatever it might be. Obviously, in Celigo’s case, where a B2B software company– but I think some of these same lessons would apply to B2C as well. So it’s clearly going to be a lot easier to sell to existing customers than going out to acquire new ones and ultimately is efficient. And of course, the formula we should always be thinking about is that growth is driven by our retention and our upsale. Retention, of course, takes into account churn that may occur. But the upsell is really where the big opportunity lies. So the question then becomes, you know, how do you know who’s going to churn? How do you know where your opportunities are to upsell? So the upside flywheel, if you will, is critical here, but it has to be powered by data, and we’re talking here about usage data. There are two aspects to this. The first is around the user journey. How quickly can users find what they’re looking for? How can they solve their own problems in their first contact with your product? And then over time, how can their increasing engagement allow them to find additional value? And this we would refer to as adoption. So you could think of the user journey as maybe focusing on an individual user and adoption, talking about the entire account. Imagine an account where you’ve sold your product, but there’s only one user. Right? There’s that main point of contact, the champion who was– maybe you have to go into bat for you to win the business. What happens if that person leaves ? If there’s no champion anymore, you will see usage just plummet, right? So this is the sort of thing that inevitably will drive the churn. The next time the subscription comes up for renewal, someone in accounting or the CFO is going to ask some tough questions and there’ll be no one to stand up for you, right? So it has to be a combination of getting that initial usage and then building the flywheel. So this graphic really captures that. Initial acquisition, quickly finding value and then finding paths to growth that deliver yet more value leading to expansion and so on in this virtuous cycle. That’s why it’s referred to as a flywheel from an old model car, perhaps, of where the power just keeps being applied and momentum builds over time, right? So this is really one of the keys here. And as we’ll see, data is really important in terms of understanding where your customer is in this adoption flywheel. So the way we think about it, it’s legal. And I’d encourage you to consider this as to think about the model here in a classic 2×2 grid where we have on the X-axis, the user journey, from one that’s kind of negative through to idyllic and positive, and adoption on the Y-axis, where we’re talking about limited adoption through to quite broad adoption. So on the user journey side, we’re looking for what are the measures that best align with the way the customer perceives value and their overall sentiment. And from an adoption point of view, again, looking at breadth here, it’s looking at engagement, stickiness, frequency of use, and so on. So from this grid point of view, you can imagine the worst-case scenario is being in the bottom left. Very negative sentiment from a user point of view and very limited adoption. Again, the next time that subscription comes up for renewal, you’re going to be, probably, in a bit of trouble. If we go to the bottom right, if the user journey is actually quite positive but the adoption is still limited, you remain at risk. This, again, is this example of where there might just be a single champion or a single user, even, at one of your customer accounts. If that person leaves or there’s a reorganization, there’s an M&A activity, you’re not going to have enough champions to go to bat for you. In the top left, we get to a point of stagnation where you might have broad adoption or somewhat broad adoption, but no one likes it. I think we’ve all used products that you sort of hold your nose and no one loves it at all. It kind of does the job. And you know at that point, it’s unlikely to grow any further. And, probably, if given half a chance, if a more attractive competitor comes along, the customer is going to churn out. Of course, where you want to be, as always, with a 2×2 grid is to be in the top right here where you have overall positive user journey, good sentiment, and adoption that is broad. This is where you really get to the point of growth. So in our case here, what does that mean for Seligo? So we think about this. I’m going to focus mainly on the right-hand side here. But just broadly we built a health score that is a combination of these user journey metrics and adoption metrics from a stickiness point of view, which really relates to adoption. There are three aspects to this: value, engagement, and advocacy. And then we can also talk about proficiency. We won’t talk about that too much here. But briefly, proficiency is the ability of how to use it to do the job themselves, right, not to have to rely on your services team or support team or third party services provider, but to actually do the job themselves and handle increasingly complex use cases on their own. So if we just focus on stickiness for a second– again, in our case, we’re an integration platform. So an integration platform allows a disparate set of business applications. Typically, cloud-based applications in the market that we focus in allows those applications to work together. So if your sales team is using a particular CRM, your finance team is using a certain accounting system, maybe marketing is using a different marketing automation system, the chances that those combinations are all designed to work together is pretty slim. So this is where integration comes in because there are many business processes that have to span multiple applications, whether it’s the HR system for onboarding, whether it’s the expense management system, whether it’s the order fulfillment service, whatever it might be. So in our case, it’s about connecting applications and the flow of data. And I just give you that as a bit of context because the metrics that matter for you will be slightly different. But I think you will find there are metrics in this in context of value, engagement, and advocacy that you will be able to imagine for your own product as well. Okay. So on the value side, we’re an integration platform. If that platform isn’t moving data and connecting applications, then it’s not really delivering value. That really is the way that value is experienced through a product like ours. So for that, we want to see if there is a decent amount of data being processed by the integrations? How many active integration flows are there, of course, and how does that relate to what the customer actually bought? What’s the utilization of their subscription? How many different applications or endpoints are they connecting and how many have they built themselves versus built from templates that we might have provided? From an engagement point of view, I think you’ll find these are fairly common for many apps. How many users are logging in and how often? How many are certified or have gone through the learning management system we call our Celigo University? Have they completed the courses and how often are they actually checking the docs before submitting tickets? These are the ways that users show the degree to which they’re engaged. Then finally, advocacy, which I think is going to be broadly applicable to all of you. NPS, Net Promoter Score. How many of ours are net promoters? How many are actually referenceable? What’s the average customer satisfaction score? And then, do we have executive level sponsors? So these are the sorts of things that we use that all go into the mix for our health score. So how did we actually go about building that? Well, it was a journey, right? And for us, it relied on our lead product, but I worked very closely with the head of our customer success organization to take the company on a journey to be much more metrics-based as we thought. Everyone talks about customer 360, but the way to succeed with this is to be metrics-based. So this was a joint initiative, really, because there was information from multiple sources, including products that needed to feed the information that the customer success team uses, which in turn informs the decisions that my team needs to make on the product side. So the first was setting a foundation of basic customer information visibility. So in our case, we use Salesforce as our CRM and NetSuite as our ERP. So this is where all the finance data is going to be captured. So the first thing we did was to automate the synchronization of customer information and opportunity information. So in our case, that meant once an opportunity in Salesforce reached, they closed one site, we would sync that over as a NetSuite sales order. And this sort of began the tight relationship then between the customer record in Salesforce and the customer record in NetSuite. And this is important because it’s almost like the primary key, if you will, if you’re anyone with a database background. This is sort of the main linkage between these systems, and it’s exactly what’s used to spread throughout the rest of these phases. So the next step was to bring as many stakeholders into the company better understanding of the customer health, bringing that visibility. So that meant, okay, now we had the basic relationships in place. How did we understand the way the product was being used? So in our case, we began by just gathering basic product usage statistics and getting them into various places. In our case, NetSuite originally and then onwards to Gainsight CS, which is the cockpit that our customer success team uses, and also then to a data warehouse that just collected a whole bunch of other information. And I think part of gathering information, you don’t always know what’s going to be valuable, in fact. There’s this challenge, I think, that you face because you don’t necessarily know what’s going to be valuable, and yet once you know it, you really wish you had more history of that data, right? Because ultimately, what you want to be able to do with the health score is to say, “Well, okay, I’m collecting this information, when I go back and look at churn, or I go back and look at customers that have expanded the most, what are the telltale signs? What are the metrics that matter? What are the things that best indicate whether this customer is going to expand rapidly or is going to sink like a stone?” Right? So you have to sometimes collect a little bit more than you think because you don’t know for sure yet how these metrics are going to correlate with customer success. All right. So this was the first step of gathering usage information. We were then using that to say, “Well, okay, what are the telltale signs that show our customer’s at risk? So this involved extending the way we engage customers by ensuring that everyone was aligned across support and CSMs, that we were gathering NPS data in-product, whereas previously we did it via email. We found it was a much more honest accounting once we did it directly in the product. I think an old email-based NPS tended to provide a bit more optimistic view, but not really that realistic. People would say, “Yeah, we love the product when you ask them in email.” And we found the sentiment was actually not as positive when we asked them in-product. And that was really information we needed to know. We’re not in the business of trying to kid ourselves with the metrics we collect. We also needed to improve the customer engagement by making sure that CSMs had only one place for the subscription data as well. So what has the customer bought? How much are they using? How are they using the product? All of that information put together, how many support tickets if they raised? When is the subscription up for renewal? Worst case scenario, of course, you’re in the renewal phase and then there’s a major that’s going to make that renewal conversation very, very difficult. And the final thing was to make this much more proactive. I think as we went on, we began by just making sure we had enough information so that when a crisis occurred, we could react. But, of course, in terms of optimizing this and getting to a point of increasing that retention, it’s important to be able to be proactive to use the information we gather to drive the behavior that we want in our CSMs, our customer success teams, and effectively try to head things off before they expand and run out of control. So this is sort of the primary four steps that we went through on our transition to being customer based. And at the end of the day, it looks quite complicated here. But I can talk you through it. The arrows here represent different integrations. Again, I’m sure you all do. You have information in different systems. You can see here, of course, support uses Zendesk, in-app engagements, we use Gainsight CS for customer success. NetSuite has been our traditional single source of truth as a company. Splunk for some product usage data and FinancialForce increasingly for the way our services teams work. So we made sure, first, that customer contact information was synchronized across everything, as we already said, which enabled us to build dashboards for CMS inside Gainsight. And so it meant, actually, we could be much more efficient in our CSMs, right, because they had the information that they needed all in one place. We then focused here on getting subscription data into Gainsight as well. So from the source of truth, which was NetSuite, into Gainsight. And so this enabled our CSMs to be much more proactive about knowing exactly where they were, when’s the right time to be really focused on the customer usage. Of course, you’d love to be able to do that all 12 months of the year. But of course, it’s much more critical during the sort of prime renewal phase. For us, we tend to start the renewal process in about month nine. So we really want to make sure everything is buttoned up and in good shape. We then started gathering NPS data, so taking information from Gainsight PX, which is a tool that enables you to launch surveys and engagements directly inside our product. So as the customer’s using it, we gather both usage information, which features are they using, how are they using it? And also when they do respond to the NPS survey, making sure that data makes it into the right place. And this improved our response rate in the CSMs by more than five X. Just by having the information available, we could see the problems that a customer might be having as they evolve. We could identify new opportunities because of the patterns we could see in the product usage. The way they were using the product indicated that there was a new opportunity for us. It might have just been as simple as noticing that they’d used all of their subscription up, so there was an opportunity to upsell. And this just allowed us to react much more quickly than we would have otherwise. In the next phase– ooh, excuse me, sorry. The next phase, we made sure all the Zendesk ticket information made it into Gainsight as well. Again, this just provided better fidelity into our health score and enabled tickets to be proactively escalated to the CSMs at a certain time, right. So in Gainsight, you can have certain programs that you run and certain triggers. And by having the information all in one place, also employing a tiering strategy, which, again, this was in combination between product and customer success team to identify the customers that represent our ideal kind of tier zero customers versus tier one down to tier three different type of opportunities that we might see based on the way the customers are using the product. We then started gathering even more low-level information out of Splunk that we maybe couldn’t get purely through the user interface. So again Gainsight PX is a product that allows you to gather information in the middle of the way a customer is engaging with the product. But our product as an integration platform, it’s middleware, which means a lot of the most exciting things happen when no one’s logged in when an integration is running, maybe on a schedule every hour, or maybe it’s triggered in real-time when a certain event occurs. That might happen in the middle of the night when no one’s even logged into the product. So in our case, this sort of middleware-based data, it was really important for us to get that data out of Splunk, and then route it. Originally, we had it into NetSuite, but we also started popping that into Gainsight as well as into a data warehouse. And then the final step here, we also included information out of FinancialForce, which is a professional services automation tool. And this enabled us to save a significant amount of time in onboarding new customers because we could just greatly streamline it. The CSMs were able to easily step a customer through the onboarding process for starting a new project. And we were also able to get all the project data – what did the sales team actually sell when it came to an implementation project? – and again, have it all in one place. So the net of all this is that it allowed us to get a 360-degree view of the customer to monitor usage along the way and look for things that suggested that we’re at risk. And there are multiple things. It’s not just a high number of tickets. It might be very low usage, right, compared to the subscription that they purchased. It might be any number of things. Maybe there’s only one user that ever logs in. This is, again, one of these telltale signs from the stickiness slide I showed you earlier. We look at the alerting from a support ticketing point of view. We’re looking at NPS scores and sentiment monitoring and the percentage of onboarding success, how are customers moving through this, overall. And then as a whole team from the CSM point of view, it was about reducing churn. So you’ll notice in this, I’m talking about a lot of things that really relate to customer success. But I would suggest to you that as a product leader, it’s not just about your relationship with engineering. Obviously, you need to be really intimately tied with the customer success team. In our case, we were in the middle of growing a CSM team. So in order to onboard a new team, they don’t necessarily know the product. They don’t necessarily know what’s important. It’s a matter of marrying together all this information. What did the customer purchase and why? How are they using it? Who’s using it? How often? How complex? What’s their real capability? Are they able to do the job themselves? And in our case – and I suggest it’s probably the same in yours as well – this information is in different data silos. And you see some of them listed here. There are additional ones that I never mentioned. You don’t see Salesforce on here, which is the cockpit for our sales team. You don’t see HubSpot on here, which is a marketing automation software. So there’s a lot more than this that’s going on. But the key is each of these specialist products bring their own inside, their own piece of data. But you can’t really see the whole picture until you bring the data altogether. And of course, that’s where integration comes into play. So it’s not just about the metrics, it’s about actually getting all that data that drives the metrics into the same place at the same time so it can be properly analyzed, and then, ultimately, acted upon. All right. So how do you actually go about that? And we want to talk about this now in terms of automation. You could gather this information independently and maybe begin by manually putting it together somehow, maybe some giant spreadsheet somewhere. But overall and ultimately, you’re going to want to automate this to ensure that the information flows into the right spot at the right time. So the first thing is that repeatability is key. You want to be able to get this data reliably from point A to point B. So often, there will be pre built integrations that you can use. So some of the apps that I’ve mentioned here do have out-of-the-box integrations that have been built by the vendors themselves to serve those use cases. And they can really help to get started. But you also need to think about whether those integrations are going to do the job for you, right, because they’re often designed for a sort of a happy path and don’t necessarily take into account the nuances of your business. But regardless, it’s important to build an integration roadmap. So to think about this from a set of information silos, think about your integration roadmap in terms of how it’s going to provide benefits, the impact and the reach, the costs and the complexity, and ultimately have sort of a prioritized roadmap for your integrations and automation’s here around building these metrics. You should also beware the lure of doing it all yourself. And when I say do it yourself here, I’m talking about building an integration. I mean, as product leaders, of course, we work every day with engineering teams and potentially with IT teams as well. You will find, of course, that the technical folks on your team will say, “Look, we can do this. If data needs to get from point A to point B, no worries. We’ll run a script, and we’ll take care of it.” And that is technically possible, of course, but it can be a bit complicated because you have to think about error handling. How are you dealing with reporting? And it’s not really the core competency of your business to invest all of this time and effort into integration. And more than that, there’s a high opportunity cost. The other thing that we find, I mean, we do this every day, of course. We find that the business needs change and APIs are always changing. There’s always new capabilities or maybe deprecated capabilities and APIs. And so if you’re building all this stuff yourself or your internal teams, you’ve got to know that you’re signing up for a long-term commitment here as well. Excuse me. So just the parting words here is that in order to build the metrics that matter, it’s a matter of taking product usage data and a number of other factors and bringing them all together to help identify the growth and the potential churn in your customer base. The best way to head this off is to understand what are the things that represent a successful customer. One of the most important things here, of course, is to empower customers to be self-reliant and to become deeply engaged with the product. Again, the worst-case scenario is to have what you think is a happy customer that’s really represented by a single user. And when that user leads the company or there’s some M&A, there’s not enough power in the organization, not enough footholds to be able to stick in the fight and to own the account moving forward. And that inevitably leads to churn. So you need to use data. As I said, you probably end up collecting more than you actually need, but you don’t know upfront what are the telltale signs, what are exactly the right elements or pieces of data that best correlate to churn. So you have to be prepared to collect a little bit more. But once you have, you can use that data to understand where to make improvements in the product, of course, how to drive adoption, reducing barriers to adoption, as well as making sure that the customer success team has everything that they need to do their jobs and to do it efficiently. So they’re not hunting for information in five different systems, bring it all together in one place so that the customer success team can do their job and do it at scale. Okay so that’s the end of the prepared content here, and I know we have– let’s see if we have any questions here. Now’s a good time if you do have any questions. I see one here from John, and I’m not sure if everyone can see it. They might be able to. But John asks, when building out a measuring system for the first time at a startup, how do you know which metrics you need to focus on? Would you recommend any to prioritize? Yeah. So, John, this is a great question. You’re not always going to know. Let me just go back to this earlier slide. I think you will end up having to focus somewhere here, right? You have to think about value engagement and advocacy. This is where I’d recommend. So value. It sounds obvious, but it’s common sense. But as I said, common sense is not always that common. The key here is like how do your customers experience value? It might not even be exactly the way you price the product, for example, right? Really, how does the customer experience value? So in our case, we know that a customer experiences value when effectively, they’re able to automate a whole business process or a part of a business process, right? So that was fairly obvious to us. But even over time, this has evolved, right? So, for example, we realized, and I haven’t touched on it here, but we realized over time that it’s not just about is the data moving? It’s actually a little bit more than that. It’s not only – is the data moving, but who did the work? Was it our services team, was a third-party services team, or the best case here is the customer is able to do it themselves? So this is one where they’re getting value, but the value is coming from their own proficiency with the product. That is exactly the right spot for us. So I think the biggest question for you is really to focus on value. You get intimate with it, sit with it for a while. It’s not something that just jumps out and hits you. You really have to think through it long and hard. How does the customer experience value? When you talk about engagement and advocacy, I think that things I’ve listed here are likely to be pretty common. You’re going to have some sort of training. You’re going to have documentation. People are going to be logging in. It’s the number of users logging in and how often. There’s probably something related to utilization as well. They purchased X, but what percentage of X was actually using, especially as that goes through the subscription period, if you get to month 9 of a 12 month subscription and they’re stuck at 10% utilization, then you’re going to have a tough time renewing them, right? So to keep an eye on that as well. So hopefully, I guess my message is, yeah, there’s a bunch of obvious things around engagement and advocacy, but the one that you probably want to spend the most time kind of meditating on almost is around value. So Shawna asks, “Have we noticed any change in churn over the past few months given the pandemic?” So, of course, your mileage will vary. In our case as a business, that is in the integration space with a focus on cloud-based apps and in the mid-market. We happen to have a very strong practice around e-commerce. And so where we might have seen softness in certain parts of the business, the e-commerce business has actually done extremely well. And so we’re very fortunate to say we’re actually fully on track for a whole full-year plan, even, absolutely pandemic. So we’ve been a bit fortunate. But I do think a lot of what I talked about here has made us aware of what was going on with our customers. We could see their usage change. Obviously, if they were sort of a brick and mortar kind of place or they were dependent on the physical presence, we could see usage begin to drop off, right, whereas other things we could see begin to spike. And it actually influenced our roadmap. We had to focus a little bit on a couple of minor features to improve scalability. For example, we just survived Black Friday and Cyber Monday, for example, massive amounts of traffic through our platform, and we’ve been preparing for that for a long time. We realized early on, of course, that there was a lot of commerce that was going to move to the virtual world here. Tom asks, ”How do you track other aspects besides stickiness? What metrics would you make sense while looking at the user journey?” Yeah, I didn’t get too far into that in this case because I think it varies quite a bit from product to product. In our case, though, it’s about how long it takes them to get their feet under them. And that means using certain features in the product. So games like PX that I mentioned, this allows us to see which features are customers using. And there’s a certain funnel, if you will, that we’re looking to drive them through. In our platform’s case, are they able to establish a connection to application A and application B? Are they able to get data out of application A, transform it, get it into application B, run it, schedule it so it’s running automatically? That for us is kind of the user journey. I didn’t want to get too far into that because it gets very much specific to our product, but we are absolutely tracking those things as well. Stephanie asks, “How do you recommend a company that needs to reset their 360 approach with some systems and definitely not connect them as well as we could? Feels overwhelming to identify where to start.” Yeah, so that’s a tough question without all the context, of course. But I think maybe I’ll just take us back to perhaps to this slide, right? I would begin by just making sure you’ve got the basics of visibility. Are all the systems speaking the same language? Like in this sense, it really meant that you could basically track the same customer and indeed the same user through every system we had. Even if you couldn’t see everything they did, you at least had, as I said, almost like the primary key that spanned it all. So I would begin there actually, just to make sure that you’ve got the basics in place. And then if you had to choose even within that task, even if that feels too big, as product leaders, of course, we’re always looking to prioritize, I would focus on early on in the customer journey. So in this case for us, between Salesforce and NetSuite, that means when a prospect becomes a customer, just getting that right and then kind of expanding out from there, okay? Now that they are a customer, how do we handle on-boarding? If you do the professional services, what does the onboarding project look like step by step? And I think what we ended up doing is laying all that out first, and then over time, layer upon layer, make it more and more complex. And so the diagram I have here, that was not built overnight. I’ve been with Slingo nearly four years, and four years ago, we didn’t have gang sites PX or financial force. We had Splunk, NetSuite, and Zendesk. And there were probably some other things that have changed along the way. So Rome wasn’t built in a day. Think about it as the classic building a product, your MVP here, minimal viable product for metrics, it’s what that looks like for you? I can’t say for sure in your case, of course. But expect that it’s going to be pretty rudimentary, the classic scooter or rollerskate. And eventually, it’s going to be a bicycle, a motorbike, and then a car. So think about it in layers just like you would tackling any other product problem. And Veronica asked how do you think about the user journey versus the account journey in terms of tracking adoption engagement metrics? Do you have different journeys for both or just focus on one versus the other? Yeah, I think it has to be a combination of both. Obviously, there are things that are happening at the account level in terms of are they paying their bills on time, right, versus I have a happy individual user that’s able to do whatever they’re looking at. So we’re looking at both. I think for us, it is sort of a combination. We even run into issues, and maybe you’ve seen the same, where you have what you think is a happy customer and you have two users from the same account give very different NPS scores, right. At least now, we’re in a situation where that rings an alarm bell for us and we go, hang on, someone gave us a nine. Someone gave us a four at the same company around the same period of the day, right, or around the same time period. So that’s a good example of where we’re looking at both. We’re not just sort of patting ourselves on the back and thinking the whole account is in good shape. You really need to be looking at a bit of both. So, yes, we definitely– from a product usage point of view like Gainsight PX, is very much about individual user capability. We don’t do too much in there to look at the overall– what the account is doing. For that, we can leverage other methods. Something that’s not shown on here, our own product is built on top of Mongo. So we have other integrations that get the true state of the account out of Mongo and also into a data warehouse that is sort of staged to then feed it into a bunch of other places. So I left that off this diagram for complexity’s sake. But short to say, we are looking at both pretty closely. And Jason asks how do you decide when a metric has reached saturation in terms of usefulness and retire it and replace it with something new? Yeah, that’s a good question. Data is fairly cheap. That’s the first thing, right. So, as I said, we also want the history just in case we need information. So for sure, we’re collecting more than we need. I think one of the key things, and it can be difficult to do, it’s really to do that correlation analysis, right. And you might need to get some data science help to make it happen. But to be able to look at– churn customer data, make sure you’ve got the key metrics for those customers, then look at doing some sort of correlation like which of the metrics best predict whether or not a customer is going to sink or swim. And there might, in the process of doing that, you may find that some are no longer really relevant and retire them. For us, we’re I’ll say somewhat inspired by work that’s been done at Amplitude, focusing a bit these days on what we call a North Star metric. And if you haven’t looked that up, I encourage you to take a look. It’s a great concept just in terms of focusing your effort, focusing, in fact, the whole company, not just product. It might even be the services team or support team, even technical writers, focusing the whole organizational one metric. Now, of course, that’s just the very tip of the iceberg and there’s a ton of supporting metrics. But in the last few months, I found it really fruitful to begin at the top, to think what does it really mean for us and what’s one metric that I can be confident in focusing on, say, for the next two years as the North Star. Exactly how we get to the North Star, we don’t know yet. But we know we can set the compass heading to that. And then from there, work backwards, right. So I think at first, we were just gathering a whole bunch of stuff, not quite sure why, you could say. Do a bit of correlation to figure out which of the things that most drive churn and use that to inform our actions. And now, we’re sort of flipping it around a little bit while still having this as a foundation to say, hang on, what is that one metric that we all should be thinking about? And I think in future, that’s even going to drive the way we think about our OKRs, our objectives, and key results. So maybe a long-winded answer, but don’t be afraid of retiring metrics if they’ve outlived their useful life. But hopefully, that’s given you some insight. I think we have addressed all the questions. I’m really pleased that there were so many. It’s always a bit weird in a format like this. You never know what the engagement’s like, but I appreciate it. Also welcome the chance to put on a button-down shirt for the first time in many months. So with that, I think we’re about done. But I thank you all for your time and attention. You can connect with me on LinkedIn. I think you can see in my speaker bio I’m easy to find. Matt Graney at Celigo and look forward to hearing from you. And it’s a small world in the product space, of course. So I hope to bump into you sometime maybe at a real life conference. Until then, thank you.

About the speakers

Matt Graney

Matt Graney is a seasoned product management leader with over 15 years experience in the discipline across B2B software enterprises and startups. At Celigo, Matt is responsible for the company’s overall product vision, strategy and roadmap. Prior to joining Celigo, Matt held senior product management roles at Axway, an integration middleware vendor, where he was responsible for the global portfolio strategy. Before that, Matt led product management for strategic products at Borland Software and Telelogic (now part of IBM Rational). Matt began his career as a software engineer in the defense and telecommunications industries in Australia. Matt holds a B.E. in Computer Systems Engineering from the University of Adelaide, Australia.