Host:
Welcome to Insight Exchange, presented by L.E.K. Consulting, a global strategy consultancy that helps business leaders seize competitive advantage and amplify growth. Insight Exchange is our forum dedicated to the free, open, and unbiased exchange of the insights and ideas that are driving business into the future. We exchange insights with the brightest minds of the day, the most daring innovators and the doers who are right now rebuilding the world around us.
Ben Sher:
As any subscription-based business knows, customer churn can be a major break on growth. Not only are customers expensive to replace, a common rule of thumb is that acquiring a new customer costs five times as much as keeping an existing. But new customers can also be harder to cross sell and upsell to. That's why growth-minded businesses place extra emphasis on customer retention and existing account monetization. While many organizations have a good grasp on how they perform on retention or cross-sell and upsell, a key challenge is knowing when a customer is likely to take either kind of action so you can respond accordingly.
Real-time predictive analytics can be the answer to minimizing attrition, maximizing cross-sell and upsell, and ultimately driving significantly improved revenue metrics. In this episode we'll discuss how churn and cross-sell and upsell affect net revenue retention, how to gain foresight on customer behavior, building intervention events to change outcomes, overcoming data silos for greater insight and improving your customer retention and profitability. To provide insights on these topics, let us welcome Dominic Perrett and Nick Barker. Dominic and Nick, please take a moment to introduce yourselves.
Dominic Perrett:
Yeah, absolutely. Thanks Ben, and absolutely delighted to be having this conversation with you both today. I'm Dominic. I'm a partner in L.E.K.'s Technology Practice sit out here in San Francisco. I spend a ton of time focused on topics around commercial excellence and optimizing go-to-market strategies for the B2B SaaS companies. And obviously a huge part of the opportunity for those organizations is around improving their net revenue retention, driving, improvements in cross-sell, upsell within existing accounts, but also minimizing churn and churn and attrition.
This is a topic that we are really excited to talk about. We think there's a ton of opportunity here. We think there's a huge amount of potential to leverage more advanced data and analytics techniques to really drive improvements in these metrics. And yeah, very much looking forward to chatting through today and being able to collaborate with yourself, Ben, and obviously Nick and the data analytics team as well.
Nick Barker:
Awesome. Well, I can jump in now. Thank you so much Ben and Dominic. Great to be with you both today. And hello to everyone listening. My name is Nick Barker. I'm the director of data analytics at L.E.K. in the US. What that means is I oversee our team of data scientists, data engineers, and advanced analytics specialists, and we spend most of our time working with clients, helping them get the most value out of their data.
And that could take many forms from building predictive models, setting up data infrastructure, helping them work through geospatial analytic optimization, or helping them think about and deploy AI. The subject at hand today, subscription optimization is one I have a lot of experience in. I'm excited to give the listeners a bit of a flavor of that today. A huge part of this is helping clients access and drive value from their vast data that they're collecting and being able to leverage advanced data science techniques and machine learning to help green insight. Looking forward to it.
Ben Sher:
Fantastic. Well, thank you both. I think to get started, let's talk a little bit about how churn and cross-sell and upsell affect net revenue retention. And to start out maybe how can a company increase its net revenue retention and improve its long-term success?
Dominic Perrett:
Yeah, absolutely. Happy to dive in on this one, Ben. And maybe before we actually dive into what some of those opportunities look like and how organizations can go about improving net revenue retention, NRR, probably helpful just to spend a couple of minutes upfront here defining what that actually means and what the metric represents and what the various building blocks are for that formula. If we think about a subscription business, any organization, typically SaaS organizations, there are some critical elements of growth for that business. If we start with what we call the starting MRR, monthly recurring revenue, and then we add in what we call expansion related to existing accounts. This might be selling more of the same products and services to existing customers. It may be actually expanding the set of solutions and products that you sell. Really talking about here, driving cross-sell, driving upsell. And then we take off a couple of pieces from that equation as well.
We take off what we call contraction or downsell. You may have accounts where you are seeing churn in licenses or you are seeing a reduction in the features and functionality that are being used and purchased by those customers. And then we have another element that we take off, which is just churn itself. This is customers that we had at the start of a period that we then lose over the course of that period. It's effectively, we take our starting MRR, we add our expansion spend, we remove the contraction of the spend and we remove the churn, and then we divide that by the starting MRR. That is fundamentally the mathematical formula for net revenue retention. But there is so much that is inherently tied up within that very simple formula that can drive growth for a business, but can also be a significant hindrance to growth.
If you're losing customers, if your customers are utilizing your platform to a lesser extent, that's obviously going to be a drag on your growth. And what we find across organizations is that actually a lot of decision makers have a pretty good sense of how they're tracking on those various metrics. They may even understand why they're tracking to a certain extent on those metrics. What is it that's causing churn at a high level or where are they able to see cross-sell and upsell success? But oftentimes what we find is that organizations are not very well positioned to derive a deeper understanding of insights associated with NRR. Are they able to predict which accounts are those that are most likely to churn? Are they able to identify where there's the greatest upsell and cross-sell opportunity? And where we think there's a real opportunity here, and we've obviously assisted certain clients with these questions and actually built predictive analytics engines for them, is utilizing their data to a better extent to actually identify those areas of improvement.
We'll obviously dive deeper into that today, and Nick will go into some of the specifics there around the predictive analytics. But there's real opportunity to leverage inherent knowledge that exists within the organization and combine that with really rich data sets that organizations are oftentimes sitting on and not necessarily utilizing to the greatest extent. And it's that combination of using the knowledge that the business has and that the business is able to bring to the table with that data to really drive improvements on that metric. Not just tracking churn for example, or having a relatively good sense of maybe what the next customer or the next two, three customers are that are likely to churn. But what events do we see have high correlations with churn maybe six, 12 months down the line, and how can we intervene in those events? And similarly with cross-sell and upsell, what customer behavior do we believe translates to greater opportunities to drive increased NRR within existing accounts?
We think there's a really exciting opportunity for a lot of different organizations and we think there's real potential here to move the needle on what is fundamentally a critical lever for growth.
Ben Sher:
Yeah, that makes a lot of sense. Nick, I'm a little curious, how can organizations gain predictive insights into customer behavior? If you could talk us through the mechanics of that a little bit.
Nick Barker:
Certainly, certainly. Yeah, it's a great questions. It's one that we get asked often by our clients. I think once organizations start to have the building blocks in place in terms of their data and infrastructure, in my opinion at least, being able to get predictive and prescriptive insights out of their data should really be the goal for those firms. I think as Dom mentioned, many decision makers may have an inherent sense of who may be likely to churn or who may be good candidates for upsell or cross-sell, but we're now in a position where we can really use data and advanced techniques to get much more precise and much more targeted on that.
Organizations that have rich historical data, they're starting to get the data infrastructure in place, can really start to utilize advanced techniques to predict this behavior. If we maybe go a little bit more into the mechanics of how this works, you can imagine for the subscription businesses, when they look at their past transaction data, their past subscriber data, they can use that to identify patterns and trends. For instance, if they look at all their past customers that may have churned, they can begin to look at what characteristics did they have in common, or what behaviors did they start to observe for those churning events? And then they start to understand future behavior. And in a similar way, if you're looking for customers that may be ripe for cross-sell, you can start to look at, historically when we've done cross-sell, what characteristics did those customers have in common?
What behavior did they do that meant they were a good target? And you can start to play that forward to make predictions. We can obviously do that, I can talk about it, but the real trick here is utilizing the machine learning algorithms to detect that and then learn those patterns of what they need to look out for. The algorithm will be able to detect going forward if a customer appears to be displaying similar behavior to what it's seen in the past, it can then flag it to a stakeholder, whether it's a marketing team, a sales team, to make an intervention, then the company can react on that.
One, maybe to bring this to life a little bit, one recent example is we were working with an information services company, and they had many different customers that could log into its portal to find out information. And what it observed by looking at past churners was that its customers would begin to log in less frequently. When they did log in, they would spend less time and maybe there was less people from that org were logging in. And that seemed to be very closely correlated with churn. Once we had that learning, we were able to look at all their current customers and play that through the algorithm, and it would give you a set list of companies that you maybe need to talk to, "Why are they interacting less with the portal?" And by using the algorithm there, we were able to give quite concrete next steps to engage with some of the customers. And ultimately you may be able to keep them on the platform.
And that's like talking a little bit more about the mechanics. We can then play that forward and start to build segmentations. The algorithm may be able to detect certain groups that look similar, whether it's similar interests, similar behavior, similar preferences, and you can begin to put them into more discreet actionable groups that then you can deploy different types of marketing approaches with them. Maybe there's different sales techniques and just get a lot more proactive where the accounts that maybe you're looking to prevent from churning, or looking to upsell, or to cross sell.
And then I think the final thing I'd mentioned that I think is one of the really exciting things about these machine learning algorithms is once you have built them and deployed them, they are continuously monitoring all the new information. They are refining the existing algorithm and getting more accurate and more refined over time. Ultimately the predictions are going to get more precise, and hopefully you'll be able to make better decisions going forward, which I think is really exciting and valuable.
Ben Sher:
Great. I think we talked a lot about the identification of these different customers. To shift gears a little bit, let's talk a little bit about how you can build intervention events. And Dominic, I'm a bit curious how organizations can define and embed mitigation steps into their employees workflows.
Dominic Perrett:
Yep, absolutely, Ben. And look, you're hitting at a critical point and a dynamic which is fundamentally very important to actually derive success from these tools. Where we've seen less of a success, even if you have fantastic data, and even if you are able to analyze this data in its most effective manner, if you don't have buy-in across the organization, and if you are not embedding those recommendations into employee workflows, then fundamentally you have this best in class tool that is sitting on a shelf and isn't really driving change within your organization.
Nick was obviously talking about a fantastic set of analyses and how you can look at historical events to understand what do we believe is going to drive cross-sell, upsell opportunities, where do we think there's churn risk? But then the next step is, okay, well what do we actually do with that information and how are we actually going to drive change? What we aim to do with our own clients and where we see organizations really achieving success here is driving buy-in across the organization and where you have teams that are touching the clients, right? Customer success is obviously a critical stakeholder here, really working with them and understanding early on, what are the types of interventions that you are making with your clients? How do you get involved? Let's say you have a client where you know there's some feature or some capability which is going to drive huge value for them, but maybe you're not necessarily seeing usage today.
How do you communicate that to them? Or if you are seeing behavior which suggests that actually you know what this customer, maybe they're sitting in a good tier, in a good, better, best pricing tier, but you think there's features and functionality and capabilities within a better or a best tier that you think is really going to drive value for them, how do you initiate that conversation? How do you start driving that upsell journey for that particular customer? What we do is we obviously host workshops with the various stakeholders. Sales will be involved, obviously customer success, and understand the interventions that are happening today and then obviously to some of the dimensions that Nick was talking through. The more that we can understand which of those intervention steps actually drive change and actually have driven historical success, obviously those are the ones that we want to start coalescing around.
We actually start building a playbook for intervention and a playbook for mitigation events and steps such that we can actually use this data to then start to drive change. And there are actually tools that you can build, third party applications that you can configure and customize that allow you to actually bake this into the day-to-day workflow, say of the customer success team or the sales team, such that it flags. There is a customer here that we believe is at risk of churn, this is why. But most importantly, this is what the machine and the engine is suggesting as the next step. And obviously there's always going to be a human in the middle here. This isn't a hundred percent automated. It obviously it can be, but we think the greatest success is with human intervention. And that then results in some intervention event.
And then to Nick's point, obviously the beauty is that all of that information then gets fed back into the algorithm and is then used to inform the next version of that intervention. Was it successful? How successful was that event? Did it prevent a churn event or did we actually see greater upsell, cross-sell as a result of that suggested intervention or not? And obviously the machine learns and gets smarter with each of those events. There's obviously the baseline model, which suggests initial interventions. And then over time the model improves, gets smarter, gets more targeted, understands to a better extent how different customer segments are behaving and how they will react to certain intervention events. And obviously the value just continues to grow and improve over time.
And there's also ways that you can obviously build in experiments to help train the model to a greater extent. Nick and I did that pretty extensively with a project last year where we're actually designing bespoke experiments that allowed the model to learn to a greater extent. And obviously that just improves the overall outcomes further down the line.
Ben Sher:
Great. It sounds like you can do quite a lot of cool things if you have all of the right data. I think the one thing that we often hear from our clients is this issue though around data silos. They have a lot of data within one system, whether it be ERP or CRM, but it's hard to be able to stitch together to get greater insights. Nick, I'm curious from your perspective, what are the steps that organizations can take to overcome data silos in order to gain greater insights into customer behavior?
Nick Barker:
I'm really happy you brought this one up. This is, I think a very relevant question to today. And one you come up against a lot. Organizations are so rich with data now, it's being collected across all manner of systems from ERPs and CRMs, as you mentioned, to HR systems. It's stored in separate data warehouses. There's BI systems, and that is then compounded by the fact that these are often overseen by different functions across the org. As you say, they do become quite siloed. And while it's great that companies are doing a better job of collecting the data, to really get the value out of that additional data being collected, we think you need to really be able to combine it. And that is really critical as we think about the subscription optimization piece of work. You really need to be able to combine the data from multiple different platforms to create a single unified view that you can then use for the analysis.
Being able to bring in everything that the sales team and the customer success team are seeing, and it's being collected in the CRM to maybe clickstream data, how your customers are actually interacting with your platform, being able to mesh that all together is where you can get the real value. And yeah, to do this piece of work, you really have to start with a extensive data engineering process that starts with accessing the different data and the different systems connecting into the different APIs. Then building those data pipelines, bringing it all together into one unified place. From then, that's where your data scientists can train the model from.
I think one other thing that I would add is that we can also enrich this internal data by bringing in external data sets to give new angles that maybe the company wouldn't currently be able to see. If you think about some of the government agencies that are collecting public data, you have social media scraping, clickstream data, even financial data, all these different types of data sets can be brought in to enrich that internal data. And when that's all brought together, you can really start to see a much more meaningful insights and hopefully get to better answers.
We strive to integrate all those different data sources together into a seamless model. That's the first step. And then we can begin to train the machine learning algorithms to uncover those patterns and drive the insights. And then once it's from there, it's continuous monitoring, analyzing, making sure all the different data sets continue to be flowing in through those pipelines, then you're off to the races.
Ben Sher:
Fantastic. Well, I think for the final topic that we want to discuss a little bit, it's around improving your customer retention and profitability, and more specifically talking a little bit about some of the key benefits of focusing on customer retention.
Nick Barker:
Yeah, yeah, absolutely. And this topic is obviously top of mind for a lot of people, especially in the current macroeconomic situation and just the uncertainty that we're seeing in the market. I think the last 12, 18, 24 months were obviously characterized by a huge number of growth tailwinds across very many different industry verticals and organizations were fundamentally able to lean upon customer acquisition as a major driver of top line growth. And I think what's really come into focus over the last three, six months is actually there needs to be a greater shift of focus and a greater emphasis on frankly, the metrics that all feed into net revenue retention. Driving increased spend within existing accounts and reducing customer attrition to the greatest extent possible.
And Ben, you obviously flagged a great data point at the top of the call in terms of the benefits of improving customer economics with the customers that you already have relative to going out and winning new accounts. And there's obviously the macro dynamics, but there's also just frankly, the increased cost of acquiring new customers. We saw a great proliferation of tools and organizations getting smarter in terms of their use of SEO, their use of ad spend, et cetera. And fundamentally, the tides have just been raised across the board in terms of the degree of effort that organizations are putting behind customer acquisition. That's just becoming an increasingly difficult avenue for growth.
We talked about some of those drivers around upsell, around cross-sell, around reducing churn. And what we've actually seen, we ran a pretty interesting regression of changes in NRR relative to changes in valuation multiples. And what you see is, I think for every percentage point increase in NRR, you actually get a 0.5x increase in the multiples. Clearly a metric of interest for investors and fundamentally something which drives growth for an organization.
But I think the other piece to keep in mind here is we have a lot of clients that come to us and actually are sitting on quite rich data. And maybe to Nick's point, that data isn't always necessarily aggregated and combined and cleansed in the best manner possible, which is obviously a first step. Even the tools such as subscription optimization, but actually going down the path on something like subscription optimization is actually a way to get organizations excited and aware of the predictive power of their data.
And we actually see organizations, if you think about their own maturity as it relates to their use of data, using this as an initial way to start leveraging their data, and then it unlocks a greater number of use cases further down the line. If you think about pricing for example, and utilizing predictive analytics around driving realtime pricing optimization, or optimizing prices across different channels across different customer segments, and utilizing the data that you have inherently as an organization test and learn and try to identify that optimal price point across different customer groups.
It's both obviously a huge inherent benefit just in terms of what it's focused on, but then it's also got these secondary benefits around getting various stakeholders within an organization excited about the potential for data, and starting to think about some of the next steps and where else they can be applying tools such as this.
To close the conversation, we'd like to thank our expert guests Nick and Dominic for their discussion regarding subscription optimization. We're happy to provide more detailed discussions on request, and we invite you to connect with us to learn more about how L.E.K. Consulting has extensive experience in providing strategic support to subscription-based and growth focused businesses and investors tackling the issues of subscription optimization. This includes building intervention events, overcoming data silos, and improving customer retention.
Dominic Perrett:
Thank you, our listeners for joining us today at the Insight Exchange, presented by L.E.K. Consulting. Links to resources mentioned in this podcast can be found in the show notes. Please subscribe or follow for future episodes wherever you listen to your podcasts. Also, we encourage you to submit your suggestions for future insights online at lek.com.
