VP of Product at Amplitude | Connecting Product Work to Revenue: From Daily Decisions to Bottom Line

Thank you everybody. Excited to be here speaking to this amazing audience. Amazing energy in the room. I was walking around earlier and just felt the momentum from everybody. So I always like to start. Let me see if my clicker is working. I always like to start a talk by telling people a little bit about myself. And you know, Rebecca went through a couple of my career milestones as well, but I want to double click into it for a second. So about a decade ago I used to work for Amazon and I was the product manager of the Kindle Fire tablet. If anybody's ever used one of those, a couple of couple hand raises in the car. Okay, great. I then worked at Twitter on service infrastructure. So my job was making sure Twitter was reliable, fast, cost efficient work globally. So I was actually one of the few PMs in the engineering organization at Twitter. I then worked at Box leading the developer products business. So I actually grew a B2B product from less than 10 million in ARR to over 30 million over the course of a couple years. And lastly, and now the last three and a half years I've been at Amplitude leading our core product team which is focused on analytics. Now I'm telling you all this not to give you a recap of my resume which you can read on LinkedIn. I'm telling you all this to highlight there's a bunch of different types of product work which I'm sure all of you are familiar with. I think we have about 2,000 people in the audience and so I'm sure there are people here who have done all kinds of product work. Whether it is zero to one product work like I did on the Kindle, which is creating a brand new product from scratch, whether it is what is called platform product, which is owning services and infrastructure and the backend that powers vertical product development, whether it is actually scaling a business right where you found product market fit, but you're actually trying to figure out the go to market side of it and grow your revenue base and your retention. Or whether it's actual core product work, where you have a core user base, you're continuing to innovate around features and maintain your lead. And the reason I want to highlight these different types of product work, which I'm sure many of you participate in, is no matter what kind of product work you're doing, ultimately you're going to have to connect it to the bottom line. No matter whether you're doing new product, platform, product, scaling product or core product, your executive team, your leaders, your CEO, your cfo, your investors, your Board are going to ask you, how does that product work that you're doing connect to revenue? It's something I've had to deal with in every job, and I'm sure a lot of you have to deal with it as well. Now, how can I possibly know that all of you have to deal with this problem of connecting product work to revenue? How could I possibly know? Any guesses? Anybody want to shout it out? No, I know because I see the things you post on Twitter. So here's a randomly selected tweet that I saw a couple of months ago that is sort of an example of the types of questions people ask. Usually some PM will come along and say, I've been asked to come up with a North Star metric. My leadership team only cares about revenue. What do I do now? This is usually how the topic of conversation starts. By the way, I did not make up these tweets. These are real tweets that you can go find. So you have to come up with a North Star metric. Leadership cares about revenue. What do you do now? First, there's denial. Why does leadership only care about revenue? Shouldn't they care about usage and adoption and engagement? What is leadership's problem? So first you sort of push back gently. Then somebody says, look, it's really not about revenue. It's about what people are doing with your product. We should talk to our users. We should understand what they're trying to solve, what the use cases are, etc. So this guy really wants to apply product thinking. Then somebody says, but how do we ultimately monetize our product? That's what we should be looking at, right? What aspect of usage is the thing that actually leads to people paying us money? And then people usually come along and say, well, here's a bunch of buffet of menus that you could track. Active users signups, retention. How about customer lifetime value? So now we're in the mode of let's just throw out a bunch of metrics and see if one of those sticks. Then there's frustration. There is no good metric. Why are we trying to come up with one metric? This is a fruitless exercise. Let's stop trying to do this. Why isn't leadership interested in the value? Why is leadership so focused on revenue? So we're back in our denial state. And this is a very long Twitter thread. I actually just picked the best tweets I could find. There's a couple more. Usually somebody comes along and assumes certain things connect to revenue. They're like, maybe usage connects to revenue. Maybe uptime connects to revenue. So you're starting to look for ways to connect the work you're doing to revenue, but you can't definitively prove it. Of course, there's somebody who says, I'm not a fan of a single metric. There's actually a bunch of different metrics we should track. So stop trying to artificially constrain me to one metric. But again, leadership is going to push you for one metric and to show that it connects to revenue. And then there's always somebody, no offense to analysts, who comes along and says, how about a bunch of random stuff? Maybe that's the metric we should track. Let's obfuscate the problem and just confuse everybody. That's probably not going to work either. And then there is the person who says, well, you know what we already track? We already tracked dau, so maybe we should just keep tracking the thing we've always tracked and that connects to revenue. Of course somebody's going to come along and say, the whole business model is wrong. Why are we even in this business? Why are we worrying about a metric? We should change the business we're in. Obviously you can't do that overnight. Then you start offending the person who asked the question. It's a poorly constructed question. Why are you even asking me this question? By the way? Has anybody ever tried this with their leadership team? Your question is irrelevant. It's not going to fly. And then, by the way, this is me. You can't tell because I'm wearing a wig on Twitter. But eventually is the cold heart practicality, which, which is you don't quite know what your users are doing. You do know what leadership cares about. So in a world where you have these constraints, maybe you come up with a best case scenario and you use it as a starting point and then you go from there. So why am I sharing this Twitter thread with you? Right. To go back to the beginning, I wanted to sort of empathize with the reality that all of you deal with. So let's talk about that reality. The first is, no matter what you do, you're going to be asked to show a connection to revenue. But it doesn't stop there. Problem number one is there's so many different inputs in order to show how the work connects to revenue. There's too many dimensions at play here. That's number one. Number two, as we saw in the Twitter thread, sorry, I can't call it X. I call it Twitter. I used to work at Twitter. The number two is you don't have a complete Picture. You might hear bits and pieces from your power users, from support, tickets, from your sellers, from sales, from support, but nobody has a full picture of what is actually leading to usage, adoption, monetization, renewal, etc. Right? It gets even worse. There is a strong desire in the organization for a single simple metric that ties everything together in a neat bow, right? And people want to make sure it is easily tracked and reported. Because if your metric is so convoluted that an analyst needs to run a query once a month, it's not the kind of thing you can actually track on a regular basis and explore and reason about and make decisions around. And then getting to the other side of this branch, if you have an incomplete picture, you're going to be struggling to figure out what actually correlates to revenue, what actually has a causal impact on revenue generation. And I'm sure a lot of you have dealt with this issue, which is if all is lost, you just start focusing on are we shipping anything at all? And we start focusing on the output versus actually worrying about the outcome, which should be that connection to revenue. So this is sort of the mental decision tree of your asked to own a metric and you're asked to show how it connects to revenue. And this is the problem that all of us at one point or another, whatever kind of product work we're doing, are going to be asked to solve. So let's talk about revenue. As I said earlier, there's two different facets to revenue. One is there's many, many different inputs that feed into it. And two is you don't have a complete picture. What if we leaned into that multiple input aspect of it to actually try to complete the picture for our people want to know is this business a viable business, not just a high growth business? So this idea of strategic levers is very, very important because it is the connection between the KPIs that your leadership team cares about like growth, retention and margins, and the day to day metrics that your team sees. Now I see a lot of people taking pictures. If you follow me online, you've seen versions of this slide many a time. So this is not some secret sauce. The slides will be available later. You can find the same kind of content on LinkedIn, Twitter, et cetera. But if you want to take pictures, I don't want to stop you. So this is how you connect the dots here. And it's really, really important to use this as a translation layer when you're engaging with your leadership team. So if your leadership team comes along and Says, how are we doing on revenue? You then ask, what kind of revenue? If your leadership team says, I care about customer retention specifically, you can say, these are the things we're doing for customer retention. Now, this framework, like most frameworks, the more you use it, the more value it starts to produce over time. For example, you might notice two different teams working towards the same metric. Is that a good thing? Is that a bad thing? Should they join forces? You might notice somebody driving a metric that doesn't really connect to anything. Should that team even exist? You might have a situation where you think driving active usage so that you have more free users to convert to paid is the number one thing you could do as a company. But it turns out that that's like the number 10 thing from an investment perspective. You can then reallocate investments to focus on the things you care about. It's really a device to have strategy and execution conversations from top to bottom. Right? And so just to summarize this in a pyramid, one of my goals for this year was to use this pyramid visual. So I'm forcing it. Here is at the topmost level are the business KPIs that leadership cares about. In the middle are those strategic levers. And at the bottom is the North Star operating metrics that you as a day to day PM deal with. Those business KPIs are lagging. They are ultimately what people care about, they are what is connected to outcome, but they don't shift day over day. Right? If you had a chart for retention and you looked at it every day, I guarantee you it wouldn't change. And so those are two long term horizon to look at regularly. The levers go one click deeper. So the levers are things that move on 3, 6, 9 month horizon. So they're great for organizing your teams, for deciding your priorities, for setting goals. Right. And then lastly, the North Star metrics, assuming the whole ladder connects top to bottom, are how you organize and prioritize and stack rank things locally. So you can say, I have three product ideas. This idea is going to lead to the most active user growth. So I'm going to do that because I believe active user growth ultimately correlates to revenue growth. And my leadership team has said that's my number one priority. Right? So this is a framework to help you sort of organize and do that. Now I want to go back to this decision tree. If you remember our Twitter thread, it was I have to connect to revenue, I have multiple inputs, I have an incomplete picture. I want a single metric, I want to track and report it, does it actually correlate and cause a little impact? And I want to focus on outcomes over impact. So I want to share two parting thoughts here. The first is like with any good framework, there are best practices. The best practices for this framework I shared with you are you want to make sure you use the ladder to tell a story from top to bottom or from bottom to top. If the story doesn't make sense, you need to go back to the drawing board. And the first time you introduce a concept like this, you will find maybe leadership hasn't thought about what levers they want to push. Maybe a team doesn't really have a handle on a metric. Maybe we never really proved the causal impact of something. And so the first part of this is use the ladder to your benefit. The second part is those levers in the middle. They force you to make strategic prioritization decisions. They are a way for you to go to your leadership and say, what matters more right now? Retention or new revenue growth. And your leadership team should be able to answer those questions if they've done that thinking, if they struggle, you might actually go to them with a buffet and say, I have a product team, I have one designer, seven engineers, me, researcher, scientist, whatever, and say we could do A, which we think will lead to this much impact on the levers, or we could do B. What sounds like a better story to you? So use that to sort of on focus, force a decision from your leadership team and to talk through trade offs. Right? In the interest of simplicity, you are definitely going to leave out aspects of a metric. But that's really, really okay to do. Because the most important thing is generally that it passes the sniff test, that everybody is able to understand and reason about the metric, that everybody in your organization can easily track it, report it, dissect it, understand the drivers of it, and come up with new hypotheses for how to move it. Right? I would rather all of you have a metric that is generally agreed upon and generally connects to the right dots and improve on it over time versus analysis. Paralysis of trying to come up with the one composite perfect, end all, be all metric. It just doesn't exist. And it's very, very important that your learning rituals surface that. So in your learning rituals you should be talking about what is the work we're doing, how do we know it connects to the metric? Show that the metric is moving up or down and iterate, iterate, iterate. And at every logical juncture, whether it's twice a year, annual planning Whatever the cadence is, you should have a conversation of is this metric no longer serving us right now I mentioned best practices. There are also some worst practices or anti patterns and you saw this in the Twitter thread. One is to just adopt revenue as your metric. It's a very bad idea for a product team to say revenue is our metric. I've been there, I made that mistake myself. You will then become very emotionally attuned to the ups and downs of your sales cycle and it can become very demoralizing for your product team because when you win big deals you're an amazing product team and when you lose a customer, you're a terrible product team. But we all know that's true. Not there's a lot that goes into actually booking the revenue or keeping the revenue beyond just the product, right? So don't fall for the direct attribution trap. The second is what I call peanut buttering, which is you go to your leadership team and say well what matters more? Growth, retention or margins? And they say all of it. All of it matters. Actually all of it doesn't matter. You have to choose because somewhere out there is a competitor who is choosing, who is making a hard trade off decision and a hard priority call because. And you have to be prepared for that situation. Third is what I call anecdata, which is you hear from one customer that this feature is the reason they bought or upsold and then that becomes the driving force. Anecdata is not the same as data, right? I'm not saying qualitative data is not useful, but even qualitative data needs to be reconciled. I mentioned earlier you might have stray teams, teams that actually don't contribute to any metric or any outcome for the business and they should be consolidated and collapsed. We also talked about composite metrics which is trying to jam everything into a metric to tell a complete story and that's just not going to work. And then lastly, I think the number one thing when it comes to a metric is you want it to be legible. You want everybody to look at a metric and say, oh, I think I know what this is tracking, I know how to reason about it and I have some ideas on how to explore it further. So that's sort of everything I had in terms of best practices and anti patterns. I want to leave you with a couple of actionable things because I am at about that time. The first is I love to connect with people so you can always find me online. We can have a conversation if you enjoyed what I shared today. I have a newsletter where I share more of these types of ideas and I even teach classes about it. And so that's a URL where you can connect with me on LinkedIn. I'm iPscribe on Twitter. My newsletter on substack is called Run the Business, and I teach a course on maven. That is my time, and I will hand it over to Rebecca again.