<< All Episodes

Episode 02: The 3 Pillars of Healthy Growth

 

Apple Google Spotify Overcast RSS

Episode Summary

In this episode, your hosts discuss their 3 pillars for healthy, sustainable growth: Path Design, Performance Valuation, and Super-Outcomes.

After introducing each one in detail, we explore how the pillars can be combined to create compound user/revenue growth, without having to resort to engagement hacks.

๐Ÿ”‘ Prominent Concepts

  • "What does Healthy Growth look like?" (0:29)
  • An Overview of the 3 Pillars (6:43)
  • Pillar 1: Path Design (15:26)
  • Pillar 2: Performance Valuation (30:19)
  • Pillar 3: Super-Outcomes (50:52)

๐Ÿ’ก Major Takeaways

  • Don't think of your product as an "object" that users buy access to. Think of it as a sequence of actions. Users are buying your preferable sequence of actions to an outcome they desire. (6:55)
  • Your users make paths for themselves by integrating your product into their lives to achieve an outcome they care about. This happens whether you focus on the outcomes they care about or not. (18:53)
  • Net churn metrics conflate starting churn (the number of people you lose in the first few months of their paying you) and sustaining churn (the number of people you lose further down into the customer lifecycle). Starting churn is always greater than sustaining churn. (42:51)
  • Not all outcomes are equal. It's more difficult to design and measure an outcome like "a stronger presence on social media" than it is a more discrete outcome like "has scheduled tweets for the next two weeks." (54:47)

๐Ÿ“š References

No direct references this episode, but here are some related reads:


๐Ÿ”ค Transcript

Samuel (00:00:24):
Hi, I'm Samuel from UserOnboard.

Yohann (00:00:26):
And I'm Yohann also from UserOnboard.

Samuel (00:00:29):
And today our episode is all about our three pillars for healthy growth. And we will get to the three pillars in painstaking detail later in the show, but to kick things off we wanted to start by defining what healthy growth means to us. And to me personally, healthy growth is growth without the hacks, that you're not trying to come up with engagement hacks for just mashing more people into your product over and over again, or certainly not trying to get people addicted or creating habit loops around products that aren't actually delivering on their end of the value proposition or things along those lines. Anything that subverts the user's best interest for the sake of driving engagement, I would generally call unhealthy growth, in part because it's unhealthy for the users, and also because I also truly believe that it's unhealthy for the business. If you're wanting to take a long-term and sustainable approach to your own growth. Would you add anything to that Yohann?

Yohann (00:01:46):
The only thing I would add is this perspective of means and ends. If you're thinking of growth as an end that you achieve no matter the cost or no matter the means, I think that's an unhealthy way to approach it. The means matter, the how matters, and you want to achieve growth in a way that is not just sustainable, but also accounts for how it comes about.

Samuel (00:02:16):
Can you share a little bit more on that?

Yohann (00:02:18):
I can give you an example there. Say you want to lose weight. If that's your goal, and that's the end, and you're saying, "I'm going to achieve this end no matter the cost. I'm just going to stop myself or just eat under the number of calories that I need per day in order to achieve this," it's not very healthy and it's not very sustainable. You want to think about the end in terms of sustainable means that don't deplete...That don't reduce you down to nothing. In software terms, reduce you down to nothing would correlate to reduce your product down to hacks that engage users but don't help them make progress.

Samuel (00:03:07):
We're not operating casinos or slot machines.

Yohann (00:03:13):
Right. You don't want small term wins, you want the long-term wins. You don't want this to just work for the next month, you want to extend that growth out to a longer period of time. And even if that means making a few sacrifices in the short term, that's okay because you know that you're building something more sustainable.

Samuel (00:03:37):
Yeah. And that's not just like hippy dippy kumbaya, it wouldn't it be great if apps were better for users kind of a philosophy. That's very much a business perspective as well because a lot of times you're looking at the what's called CAC Payback Period, CAC an acronym for cost of acquiring a customer, how long it takes for your users to pay 100 bucks a month or whatever your subscription revenue or otherwise user generated revenue would be. It's going to take a while for them to pay off how much it costs to get them as a new revenue generating user. So you don't want to just be trying to trick people to come back on day two because you've noticed that people who come back the day after they first log in, tend to correlate with better conversion rates. Yeah, superficially that's true. But you've got to be thinking about how do I deliver value over a long term and become part of long-term processes in their life so that my product continues to be relevant to them. Because if not, they're not going to use it, they're not going to get value from it and they're going to turn.

Yohann (00:04:49):
Right. And the systems you have in place to measure how that's happening โ€”this might be a slight tangentโ€” but when you have systems of measurement where you're tracking monthly targets and you're saying, "I need to meet my target this month," you're willing to do things to sacrifice next month's target in order to make this month's target look good. And that's the kind of thing where we're leaning against. You want to think about this in terms of trends and have every month trending in an upward direction, irrespective of the spikes from month to month around the trendline.

Samuel (00:05:30):
And when you say trends, you don't mean trends in the industry or trends in your market. You mean the way that the effectiveness of your offering is trending, the reliability with which it produces value for your business and for the users.

Yohann (00:05:50):
Yes. In graphical terms, you want the trend line to be moving up and towards the right. That's the kind of trend I'm talking about.

Samuel (00:05:58):
Yeah, exactly. So, perhaps that's a convenient segue to outlining what, in our opinion, are three pillars for approaching growth more in the fashion that we do recommend, and less in the fashion that we just walk through. And those three pillars are what we currently call "Path Design," another pillar is called "Performance Valuation", and another is called "Super-Outcomes". And we will go over each of those in detail, but they all work together in a high level degree as well. So we wanted to cover each of those at a very high level first, then we'll do a deep dive on each of them. So Path Design is a concept that we view as being underrepresented in the product design literature. It's a difference of paradigm where you might be thinking of creating a valuable product that you have designed well, that you can then provide access to to users and make money off of it. Or you can be thinking about why your users are finding that product to be relevant and thinking about the process that they're bringing that product into and engaging it in and designing for the whole path, the whole process that makes that product relevant to begin with.

Samuel (00:07:33):
So from that perspective, that's what we call Path Design. And from a Path Design standpoint, we ask questions like where are users trying to get to and what are all the steps that they need to go through in order to get there? And how much support is your product or offering providing to them step over step to be able to actually make it through kind of the obstacle course of what we referred to in the last episode, making pancakes or whatever kind of process that they happen to be engaged in at that time. So a lot of those kinds of considerations would take place under Path Design.

Samuel (00:08:16):
Performance Valuation on the other hand is a different beast. Yohann, do you want to start this one off?

Yohann (00:08:23):
Sure. Once you have designed the path, or once you have a path that you're analyzing, you want to measure how many people are making it from one stage to the next. And you can get really granular with how you do this. And we will dig into this granularity further into this episode, but at the moment, one big concept that I want to mention is that apart from just measuring how many people make it from each step to the next, you also want to assign a value to that whole process. And in the last episode when we talked about how to turn revenue into a leading metric, this is essentially what we're talking about. If you can assign a certain dollar value to people making it through the path, and you know the number of people that are actually making it through that part in average, and where they're dropping off along that path, then you can make very accurate guesses about how the path is performing. And how you expect the path to perform forward in time.

Yohann (00:09:37):
So at a high level, we're not just talking about measuring performance, but we're talking about valuing that performance at a certain dollar value, and that helps you make changes to the path and measure the revenue impact of those changes as well.

Samuel (00:09:56):
I love the description. I have nothing to add to it. I'm locked in.

Yohann (00:10:00):
Okay. And that leaves Super-Outcomes. Samuel, all yours.

Samuel (00:10:04):
Super-outcomes is the term that we use for when you can get users to a particular valuable outcome or milestone that they're seeking. So back before when we were talking about CAC payback and having users stick around long enough to actually pay off the amount of money it cost to acquire them (and then hopefully stick around significantly longer so that they can produce revenue for you in their subscription or whatever their usage is). In that same way, you have to be focusing not on the outcomes or user behavior that takes place within your product... It's not going to give you those kind of bigger contexts in which your product becomes relevant. And when we talk about designing an entire path around a particular outcome, you really want to be specific about which outcomes you select because it does take a lot of effort and rigor to design an entire path and to evaluate its performance and to tune it and iterate and improve over time and have it trending upward as we were discussing before.

Samuel (00:11:19):
So you want to be really selective in which outcomes you're designing for. And to start with, they're not outcomes that take place within your product, they are outcomes that take place exterior to your product that bring your product into someone's life. So there are a number of other considerations that make it either more or less reasonable of something to invest your time toward. And we will dive into those in richer detail when we cover it in the deep dive. But between the three, Path Design, which is all about designing across an entire process instead of designing a single state product, Performance Valuation, which is all about being able to measure the impact that your changes to the product or to your offering or to the path, are having on key outcomes for both the users and for your revenue. And then we also have the concept of Super-Outcomes, which is getting really specific about what those outcomes for the users are that really drive the most revenue, have the healthiest LTV, pay CAC back faster, etc. So all three of them work together and we are going to dedicate the rest of this episode to talking about each of them in more detail. Unless you have anything else to add to the overview, that is, Yohann.

Yohann (00:12:49):
I would like to say one more thing about how they work together and that thing is the deeper you get in each of these, the better you can do the others. One thing we see is that people think about these things as separate. Path Design is usually constrained to in-app flows and it's usually a design problem. So marketing, who's thinking about super-outcomes that they should put on the homepage like value proposition, how do we get people to care about our product kind of stuff, they're not thinking about the path, they're not thinking about how that marketing material tees people up for product use later on.

Samuel (00:13:33):
Well, and generally speaking, the product people are not thinking of how do we take the baton of what marketing inspired people to do and actually fulfil the rest of our end of the value proposition and confirm that people are actually getting what we're advertising.

Yohann (00:13:51):
Right. And on a metric level, each team is so focused on their own metrics that the more you zoom out, the more aggregate the metrics get. So if you zoom out and look at a CTO, for example, the kind of metrics he's tracking are all aggregates of the impact of different teams. So we see each of these pillars treated โ€” 1) superficially and 2) in isolation. And what we're trying to do is not just bring them together, but point out how achieving a certain depth within each of them can help the other two pillars perform better. The better and more specific you can get with a superior outcome, the better you will be able to design a path later on. The better you analyze a drop-off between steps in the path, the more specific you will be able to make your Super-Outcome. Things like that. To use a terrible word here, there's a lot of synergy between these three pillars.

Samuel (00:15:04):
That was the word. I was like, "Samuel, don't use synergy. Samuel, don't use synergy." Alright. I agree. So with that in mind, let us boldly proceed into diving deep into each of these three pillars. First one up Path Design. Yohann, would you care to tee this one up?

Yohann (00:15:26):
With pleasure. So when you're thinking about a path...think about your geometry class and your introduction to geometry. Drawing lines and making triangles and so on was thinking about points. In the same way, when thinking about paths here, we're going to think about points โ€” starting points and resulting points โ€” because a path is just a connector between two points. And even the term "starting" and "resulting" is contextual because what's a starting point in one particular context is a resulting point in another. The signup page, for example, is a resulting point in a marketing meeting, but it's a starting point in a product meeting. That's the kind of thing I'm talking about. Anything to add here, Samuel, before I keep going?

Samuel (00:16:25):
I think that you put that quite well. I would agree that the idea, it's called Path Design, but really the point is to make it more likely that people generate the desired outcome, which is the end state of the path. So the point, let's say, I mean, to use an ultra basic example, let's use the user interface of a light switch as an example. If I become a user of a light switch, it's either because I probably want to make the room darker or because I want to make the room brighter. And depending on which direction I flip the switch, I will be changing the status of my situation. So if I am sitting here and I'm ready to go to bed and I go to flip the switch because I'm going to bed, then contextually speaking, the resulting state that I'm seeking is to have a darker room so I can go to sleep. Is that a fair example?

Yohann (00:17:31):
Yes. I think that's a great example. Yeah. So you're working with resulting points and these resulting points are outcomes. And when you're designing the path, you're kind of working backwards from these outcomes to figure out how you can make those outcomes happen more reliably. But that being said, you're also working forwards because your starting point is defined by conditions that users have at the moment. And you have to meet them where they are. So you're working-

Samuel (00:18:12):
Strongly agreed.

Yohann (00:18:13):
You're working forwards from where users are at the moment and backwards from where you want them to be.

Samuel (00:18:19):
Yep. The two main things that really stand out to me in what you just said is that you always want to have a general idea of what the user's state is so that your system can guess what their desired state is, and then organize it's experience around forming a path that guides the user to that desired state, rather than just saying, "Hi, I'm a product. And if you learn how to use me, then you can hopefully get there."

Yohann (00:18:53):
That's a great segue actually, because it leads me to the next thing I wanted to say, which is that you already have paths. If you have a product, you have paths that your users are working within at the moment to achieve the resulting states that are meaningful to them. But, if you're not thinking about the resulting states themselves, and you're thinking about ... Okay, let me take a step back. If your resulting point is something that isn't in product action or behavior, and if you're optimizing the path to make that product behavior happen more, it's good that you're doing that, but by not focusing on the resulting situation that the user is pursuing, you're kind of forcing the user to do all the work by themselves after they make this product behavior happen.

Yohann (00:19:55):
So say you want this button ... sorry, this template, that's a better example. You want this template to be used more. This template is useful because it is helpful in a particular situation like a team meeting for example. You're making a template that is helpful in a team meeting. You want users to interact with the template more so you design a path to make that happen. Users then, once they do make that happen, once they do interact with the template, have to do the work of making that template actually helpful in the situation of a team meeting. And you would be much better off if you thought about the team meeting as a resulting point to design to, rather than the interaction of the template. Does that make sense?

Samuel (00:20:52):
What do we want the effect of the template to be on the team meeting? And then just say, let's not even think of it as a template, let's just think of how to get team meetings to be good and the way that we're trying to get them to be good. So if we think about how we can actually produce an outcome for somebody if we have an understanding of what their current state is, the real question then becomes, how do you create the path in between? And what we like to do before thinking about it in terms of screens or interfaces or design patterns, we like to take a step back and look at it as a process instead that has a number of steps. So our background is in user onboarding. If you really look at your user onboarding flow under a microscope, there are a number of different steps that you are asking the user to do, often. Maybe it's to create an account, maybe it's to invite other colleagues, maybe it's to confirm their email address, maybe it's to let you know what the size of their company is and what the industry they're in, or all kinds of different things. And all of those screens that are included in that flow have to be acted out in reality by the user.

Samuel (00:22:24):
And if you think of those as a series of actions, especially within the context of generating the outcome that you're looking for, if you have an invoice sending software company and somebody's coming to you so that they can start sending invoices to their clients, and so that they can start getting paid, and you're starting out asking them what kind of industry they're in, or how many people work at their company, or having them go confirm their email address, these are not steps that you would say, "Oh, of course, if you're trying to have somebody get an invoice paid more reliably, you got to start off by having them tell you what their industry is." Because frankly speaking, it's probably not really that relevant to the user in that step. And a lot of times when we look at user onboarding flows, that's like a question that a sales intern put into the flow three years ago, and they're not even there anymore and nobody thought to take it out.

Samuel (00:23:28):
So being really, really specific about what are the actions that take place to produce the required smaller changes that result in the desired, bigger change. And our nerdy term for those smaller changes are compounding sub-outcomes where you have the idea of the end outcome, the resulting outcome in mind. And you have to think about given the user's current state, how do you create a sequence of smaller outcomes that compound together almost like crafting in Minecraft, if you're familiar with that, or if you've seen your kids play it or whatever. How do you pull together the right resources at the right time and in the right configuration so that you can produce something that's bigger than any of those individual changes on their own.

Yohann (00:24:31):
There's so much to unpack in what you just said, but I think focusing on the term compounding sub-outcomes kind of encapsulates what we're trying to say here. First, we're trying to say that changes are made of changes. So when you're thinking about one of the changes in the path to successful onboarding being users creating a password for themselves, that change is made of smaller changes. First having to click on the form field and then having to find the first letter of the password on the keyboard, and then the second and so on. It's every change is made of a smaller change.

Yohann (00:25:17):
So when you start thinking of outcomes as made of sub-outcomes and those sub-outcomes made of even smaller sub-outcomes, you're thinking along the right track of what needs to change. But on the other hand, the compounding side of things is that if you're organizing this path to contain changes, that don't progress the user towards the resulting state, then those changes are just dead weight in this path. They don't make the resulting point happen any easier or any faster. They're there for maybe your benefit, but not for the path's benefit. So they actually detract from the outcome happening more reliably.

Samuel (00:26:16):
Correct. Or there could even be things that are well-intentioned but just not relevant to a given user at a given moment, or all kinds of different things along those lines. So the reason that we focus on the compounding part of the compounding sub-outcomes is that you want to focus on what absolutely has to change between right now and the place that I would like to be. So if I want to see my son in junior league hockey, I've got to go buy him skates because he doesn't have skates right now. I've got to go buy him the equipment. I've got to go find a junior hockey league.

Samuel (00:26:56):
There are a million different things that I've got to do along the way that I know have to change. And if I was creating software to create the end result of this random example that I just made, I would identify what are the changes that have to absolutely have to take place and in which order, or when in the timeline and think about how I can position the resources that I offer as my software company to boost, provide progress boosts at every point of change. Not thinking about it in terms of like, how can we stuff as many cool ideas and different activities into this process as possible. But how can I just measure the absolute, what I personally call "critical pathway". The most limited, bare bones representation of what has to happen between two points in time and use that as a starting place. Because I think that that leads you to focusing on ideas and having design inspiration that's a lot more likely to actually be of utility to the users rather than, "Yeah, it was like a cool feature idea, but nobody actually cared about it."

Yohann (00:28:19):
It was a trap I fell into all the time when we started working together, where you'd be like, "We need to design this thing that gets people X." And I'd be like, "Oh, X needs to happen. They need to do this cool thing and they need to do this activity and they need to work with this worksheet." But I wasn't thinking about it in terms of what actually needs to happen in order for X to happen as well.

Samuel (00:28:44):
Yeah. And I think one other aspect of it too is the timing aspect of it, having a real theory of change, theory of how do we actually get people from a common state that they usually arrive in to the desired state that our marketing implies is possible with our products. How do we actually think that that happens and how do we go about breaking that process down and evaluating the resources that we put out through that lens? And there are a number of different design considerations at a UI level that become really apparent when you start looking at your product less as a product and more as sequences of steps in a process. And I could get into that too, but I feel like that's a whole other episode. But the general idea being evaluating the steps themselves and do they even make sense? And can you help create content that resonates with the users on where they currently are and help them advance to where they're trying to get to, and we can just leave it at that. But that's the general gist of Path Design to me.

Yohann (00:30:09):
Right. That makes sense. I think that's a good place to leave it without getting more granular, but there is more granularity to get to.

Samuel (00:30:19):
So with our due respects paid to Path Design, we now segue over to pillar number two, which we call Performance Valuation. And the main theme here, as you may recall from the intro is that we should be able to track the impact that our design changes have on user behavior and especially user behavior that strongly correlates with the users getting what they want and/or the business getting what it wants.

Samuel (00:30:59):
So the idea here being that if you think, "Hey, maybe we should add a tool tip tour to our onboarding sequence, maybe that will result in better things happening." And you just add it, you're not going to know whether that had a positive effect or not, or maybe no effect. Maybe it's bringing the conversions down for all you know. So generally speaking, what we recommend is to start with revenue and not only measure the first time people produce revenue, which is oftentimes going from signup to producing revenue for the first time is often called your conversion rate, that we think it's crucially important, but we also think it's equally important to measure how many people go on to pay a second time, if we're talking like in a subscription model. How many people stick around for a third payment, a fourth month, fifth month, or six months, etc.

Samuel (00:32:01):
So that you can actually track how many people are actually making it to a point where your cost of acquiring a customer, your CAC that you're trying to pay back when that actually takes place from a cashflow standpoint. There's no reason not to be tracking these kinds of things. And if you're tracking that, it's the same principles of Path Design, where you're starting with a starting point of signing up and an ending point of let's say revenue one, just for sake of example. If you use cohorts, you can tell whether the percentage of the signups that you're getting in one given week is going up or down as you continue week over week over week.

Samuel (00:32:49):
And that alone is helpful information, but then you can also identify the required steps in between and measure where you're losing the most people along the way from sign up to the first time they produce revenue. And see which of those steps might be bringing the most attrition to your overall flow. And be able to design in a way that would have a much more sophisticated hypothesis behind it than what we often see growth teams going after, which is like, "Can we improve day 20 retention?" And just like, "How can we just get more people to keep coming back longer?" Or something along those lines. I'm not trying to talk trash. Am I in the safety zones Yohann?

Yohann (00:33:43):
Absolutely. We're not criticizing people here, we're criticizing an approach.

Samuel (00:33:52):
Yeah. That's fair. That's fair. That's a good one. That's very diplomatic. Is there anything that you would add as being crucial beyond what I just outlined?

Yohann (00:34:03):
Yes. Let's use a different example. Let's say you're not tracking ... okay. A previous example was you're tracking the drop-off from one step to the next between signup and rev one. Let's switch that out with tracking the number of people who make it through a cohort lifetime milestone like month 24, or maybe that's too far, say month 16. So you're tracking the number of people who make it from paying you money for the first time, rev one, month one, and you're tracking that all the way to month 16. And you're trying to figure out what the drop-off is from month to month. Like Samuel said, if you do this on a cohort level, you one, will be able to know how many people you're losing at each month.

Yohann (00:35:07):
So if you're measuring the number of people who signed up in the week of September the sixth, if that cohort starts out at a thousand people. A thousand people paid you in month one, that number drops down to 800 in month two, drops down to 500 in month three. And this is just a hypothetical drop off curve from month to month and onward onto month 16. What you can do if you're tracking these things is look at all your past cohorts and past weekly cohorts, like people who signed up in the month of August the 24th, the week of August the 24th and the week of August the 16th and the week of August the 9th and so on, and track what the drop off rate looks like from month to month, from month one to month 16, for all of these different cohorts as well. And you will be able to start to draw patterns.

Yohann (00:36:16):
You will be able to start to figure out what you expect to get next month from your most recent cohort, September the 9th. And this is a great way to turn revenue into a leading metric, is to figure out if the changes you're making today impact the amount the of money you expect to get from your most recent cohort September the 9th next month. If you expect to get, say that at this point there are a thousand people in the cohort and you have a $10 plan. So if you expect to get-

Samuel (00:36:55):
10$. What kind of BS businesses is this? #chargemore. #patio11.

Yohann (00:37:05):
$10 just so the math will be easier on my brain at this point.

Samuel (00:37:10):
All right. I will allow it.

Yohann (00:37:14):
We could take it up to $100 or $1,000, but we have to work in multiples of 10 for this to be easy on me.

Samuel (00:37:20):
Yeah. Let's keep the zeroes short in this case.

Yohann (00:37:24):
Okay. So let's say you expect to get $10,000 from this cohort next month. Can you make changes that result in the cohort beating your estimation of what it's supposed to pay you?

Samuel (00:37:49):
Beating the average drop-off rate that past cohorts have had. That's what your estimation would be based on?

Yohann (00:37:56):
Yes. Yes. Because they can't pay you if they drop off and if they don't make it to revenue event number two.

Samuel (00:38:05):
Right. So what I'm hearing from you is that if you look on a cohort by cohort basis beyond just the first time that people pay you, that you will see that there's a drop-off curve that goes from one month to the next and hopefully flattens out or ideally if you're really smart, starts to curve back upward again. But the idea being that if look at the drop-off that takes place from one month to the next to the next, then that gives you a very reliable means for predicting how much value the entire cohort will be worth because the cohorts tend to when you're working on a scale of 16 months or whatever the specific amount that you mentioned was, they tend to be pretty inflexible, which means that they tend to be great for predictions. So the amount of people who you retain in month six from three weeks ago, or a cohort from three weeks ago, is very likely to be roughly similar to the amount of people that you retain in month six from a cohort from eight months ago. Right? Percentage wise.

Yohann (00:39:33):
Right.

Samuel (00:39:34):
Now, if you factor out the differences in the amount of users that you acquired, if you're just saying the percentage of people who pay us starting in month one tends to be 20% by month six or whatever.

Yohann (00:39:52):
Right. And if you build out a new feature that blends with the product without really affecting the path that users in this cohort are on, of course you're not going to see any changes to the drop-off curves.

Samuel (00:40:10):
It's not extending its relevance to the user's timeline.

Yohann (00:40:14):
Exactly. But if you can manage that, you can see these curves change because a smaller number of people will drop off. And that equals to a larger number of people paying you. So it's just a question of counting the receipts you're getting,

Samuel (00:40:34):
Just counting them receipts. I agree. The information is sitting right there. It's weird that we are in such an information centric industry and yet we are not paying attention to some of the most basic data that's literally just sitting in your user table. I strongly concur my good man.

Yohann (00:40:56):
At least counting the receipts from a weekly cohort. You might be counting the receipts of all the users paying you in this particular month, but that's an aggregate. And that doesn't tell you how your recent product changes are performing, how your recent cohorts are performing.

Samuel (00:41:17):
That's absolutely correct. If you look at, what you and I call a "decay curve", which is you start with month one and then see how low it gets in month two, and then you see how low it gets in month three, and then you see how low it gets in month four, etc. If you chart that out, it's sort of like inverse hockey stick growth, where you early in the process, your churn from month one to month two, it's going to be significantly higher, almost definitely, than your turn from month two to month three, which will be higher than your turn from month three to month four, percentage-wise of how many people you're losing at each step.

Samuel (00:41:58):
So again, from an onboarding perspective, it really puts it into perspective to look at how many people you're losing, not only from month one to month two to month three early in that process, but even just early in your acquisition or activation process, just going from signup to the first time they pay you or sign up to the first basic successful outcome that anybody would expect your product to deliver on. You're just losing tons of your signups at each of those steps. And only later on, usually several months after they become customers, does that cohort start to flatten out again, hopefully. If not, then you're in even deeper trouble business-wise.

Samuel (00:42:51):
So there's a difference to us between what we call "starting churn", which is just the huge amount of people you lose at the beginning and "sustaining churn", which is what you hopefully want to from a revenue standpoint, turn into net negative churn and turn it into a smile graph where the revenue from each cohort starts to go upward instead of downward. But from a starting turn perspective, you're just losing so many people early in the process, which to your point about net churn, really puts into perspective why net churn is not the greatest financial predictor in my opinion. And especially doesn't give your company a lot of control over being able to leverage it. Instead of looking at a pile of subscriptions and how much they're worth from one month to the next to the next and calling that net MRR, and then looking at how many people you're losing, either head count wise or revenue wise from one month to the next to the next on a net level, that includes people who just signed up and people who have been with you for nine years.

Samuel (00:44:10):
So when you blend those together, it gives you a false sense of how quickly you're losing people. Where if you look at it on a cohort by cohort basis, you can see that you just, generally speaking, have just a huge drop-off early on and then a flattening out afterward. So instead of focusing on bringing your net churn up or down which blends the starting churn and the sustaining churn, I would instead really focus on trying to bring your starting churn down or in other words, to retain more and more of your signups from the beginning up through every step that they have to go through to become customers and then to become recurring customers and remain as such.

Samuel (00:45:00):
If you, instead of looking at MRR from your timeline, the timeline of your company and how much MRR it has made, it can really be revealing to instead look at MRR from a cohort's perspective on the user's timeline, how many people who signed up in the week of whatever continue to do activity X, activity Y activity Z become customers, stay customers in month two, stay customers in month three, etc. That churn tells a very different tale. And one that has a much faster feedback loop and lets you make design decisions where you can actually see the impact there.

Yohann (00:45:54):
Yep. Makes sense. The relationship between Performance Valuation and Path Design is becoming more clear, because just pushing a feature out is not going to make the resulting point of revenue happen more reliably. So you need to push out design changes...you need to come up with hypotheses that make this-

Samuel (00:46:24):
It might. You could put out a feature and it might impact revenue, but it's unlikely to be done in a way where you have the direct attribution to, we made this design change and it resulted in this change in people's behavior as users and as customers.

Yohann (00:46:45):
Right. It would be accidental if it does impact revenue.

Samuel (00:46:49):
Well, intuitive, I guess, maybe we could put a generously.

Yohann (00:46:53):
But we want it to be more scientific.

Samuel (00:46:57):
I'm just trying to be nice to people who we don't necessarily agree with.

Yohann (00:47:02):
Okay. Sorry. I didn't catch up to that. Okay.

Samuel (00:47:10):
All right. So this is getting us dangerously close to super-outcomes territory, talking about the user outcomes that correspond to revenue most closely and most empirically. So let's do a quick recap of what we just covered with Performance Valuation before we move on to pillar number three. We just talked about how you want to be generally in tune with what your conversion rate is from the starting state to the outcome state, whatever that might be. Maybe it's sign up to the first time they become a customer. Maybe it's sign up to the first time that they share a document with their colleague, whatever that might be. You want to look at what the conversion rate, the success rate ultimately of that is overtime. And hopefully see it trending in upward directions where if you can get more people to share a document with their colleagues and you can have that have approvable impact on more people converting and becoming and staying customers, then that's a pretty sound basis to be making product decisions around. But if you're just giving people a random list of activities to do, because you think that that probably will introduce them to features that they might care about, that's a little bit less strategic.

Samuel (00:48:32):
So talking about along those lines and being able to measure it using cohorts and seeing how quickly each cohort loses people or how steep the decay curve of each cohort is, which also gives you powers of early prediction, which leads to ... I'm getting tired recapping all this. Which leads to an ability to turn revenue into a leading indicator and to also connect the user outcomes in your app more closely to business outcomes in the form of uplift toward higher LTV, faster CAC Payback, more better retention, etc. Fair?

Yohann (00:49:18):
Fair.

Samuel (00:49:19):
All right.

Yohann (00:49:20):
I was on a call the other day and I used the term "decay curves", and the person said, "What? Are your users radioactive?" And I was like, "Damn, we should change that."

Samuel (00:49:34):
We can use "half-life" instead.

Yohann (00:49:38):
But these terms being flexible is just one of the many ways in which this is a collaboration. This is just how we see the picture at the moment as we're uncovering it, as we're coming up with a way to approach growth in a healthier manner.

Samuel (00:49:57):
I concur. Nothing to add. I couldn't have said it better.

Yohann (00:50:06):
If you disagree with any of this stuff, we would love to hear it. And that would only enrich the discussion around how we can come up with healthier, more sustainable ways to grow.

Samuel (00:50:18):
Yeah. The concrete is wet as we tend to say. It has not fully set yet. We've put in a lot of work behind the scenes to a cross our t's and dot our i's. But that doesn't mean that all of our guesses are correct. And we would love to hear about what you ... I mean, not that we're opposed to praise or whatever, but if there are obvious flaws, if we've got a little bit of parsley in our teeth, we'd appreciate it if you let us know.

Yohann (00:50:51):
Absolutely.

Samuel (00:50:52):
All right. So with that in place, we can now securely move on to pillar number three, which we are calling Super-Outcomes. And those are generally speaking what we've been hinting at while in the introduction, and also while talking about pillar one and pillar two, which is the idea that every time, every, every, every time somebody uses a software product, it is because they are trying to change something about their current situation. Every time. And there is particular criteria in place, whether the user themselves even realizes it or not.

Samuel (00:51:35):
So we can form a sense of clarity around the desired state that people have in mind when they show up at our doorstep. And we can design a path that we think makes it more likely that the people who are in the state they are when they arrive, will turn into the state they desired to be when they arrived with our help. So instead of creating a product where someone could conceivably do that and then leaving it up to the user to figure it out. Instead, design something that is specifically intended for turning people in the state that they currently are into a situation that they are asking us to help them get to.

Yohann (00:52:30):
And the upshot is that you will never have to worry about user motivation again, because users are so inherently situated to better their own circumstances that you don't have to convince them to do it.

Samuel (00:52:48):
You are setting up shop on motivation main street.

Yohann (00:52:52):
Right. Okay. So let's talk about specifics here. And I mean that literally. We have to talk about how to articulate super-outcomes in a specific way, because that's one of the hardest things to do. What I mean by that is it's very easy to talk in thematic terms. It's very easy to if your Buffer say, "We will help you grow your brand on social media."

Samuel (00:53:25):
No shade to Buffer. Much respect toward Buffer. We love you.

Yohann (00:53:30):
Yes, We love you Buffer.

Samuel (00:53:33):
But for example.

Yohann (00:53:35):
For example: there's a huge gap between using Buffer to schedule tweets and actually growing your social media audience, growing your brand on social media rather. Buffer helps you do it, scheduling tweets helps you do it, but there's so much more that you have to do in order to get to the promise land that Buffer is talking about on their homepage. The more thematic your outcome, the less you can actually be in control of making it happen.

Samuel (00:54:12):
Let me jump in. Just for those listening, when you say thematic, what do you mean by that exactly?

Yohann (00:54:21):
By more thematic, I mean you're talking in THEMES of what the product can unlock for you rather than discrete outcomes, discrete practical, tangible outcomes that you can have at the end of using the product.

Samuel (00:54:47):
To me, when we say thematic outcome, it's an outcome that is not super discreet. It's hard to tell whether it's actually happened or not, or when it's happened. So if your outcome was having a life well-lived, it would be hard to say, "And...we're there." So, not that those are not worthy pursuits or things along those lines, but there are a ton of different things that people are wanting to do that do have discrete outcomes in the sense of you can actually tell when it happened. Even if you're not currently set up to measure it, and even if you never measure it, there's still a difference in focusing on an outcome where you could theoretically tell whether it was taking place or not. If I may give an example to illustrate my point.

Yohann (00:55:48):
Yes, please.

Samuel (00:55:49):
Okay. So I would say that a game like Fortnite for example, the super-outcome that we would be considering from a Value Path perspective would be that the person wants an adrenaline boost, that they want a little action. They want to experience some stress and anxiety of maybe dying right away, maybe having somebody sneak up behind them, whatever. They're getting their adrenaline up and running. They're trying to get a general thrill out of the game. Maybe that's not the only reason, but I would imagine that would be the case for many people. A core motivation.

Yohann (00:56:43):
A super-outcome.

Samuel (00:56:44):
Yeah. So you could theoretically measure the levels of stress, hormones or adrenaline or whatever that are in people's bloodstreams. And you could theoretically see if different versions of your game produce that, but you probably would be pretty creepy for doing so. But if you at least keep that in mind as a discreet organizing principle, it is something that is inherently falsifiable whether you're actually going to go to the trouble of falsifying it or not, it at least gets you thinking in a way that is more conducive to what chain of events needs to take place in order for this to happen rather than what are a bunch of cool ideas that we have that are along these lines. And maybe after somebody does most or all of them, then they will have had a life well-lived or whatever. Not a theory of sequential changes that actually lead to the outcomes that people not only actually desire, but are actively desiring by showing up at your doorstep because they're signing up for your product because they hope it will help them get to some different situations than the one that they're currently in, which is driving them to your product.

Samuel (00:58:13):
So this should be just like shooting fish in a barrel, but it's not really something that gets a lot more attention, maybe there's a corner of a persona template that has goals listed at the bottom. Or you've got your user journey, or you've got your internal sense of, "If we can get people to send 15 documents, then that makes them more likely to be customers. So let's just get everybody who signs up to send 15 documents." What the fuck are we doing here? It shouldn't be that hard to get an understanding of what it is that people are trying to accomplish. "Hello. Why am I here? What can I help with? Oh, you're trying to do this, let me show you the way." This should not be brain-busting for people to try to attempt. And there's a lot more that we can go into as far as how to actually go about doing so, and these are things that you and I have practiced and field tested over the years. But even just on a basic level, it's kind of weird to me that this is a concept that even really needs to be highlighted.

Yohann (00:59:31):
I think one of the problems here is that if you think of your product as an object, you have to recognize that as an object, it can be used in different contexts. So you probably have five different personas, and they probably come with five sets of persona goals. And if you're working with five different sets of persona goals, it's not like you can pick an outcome, it's not like you could pick one of those goals and then design your product around that particular one, because the system is stacked against you. If you think about it in those terms, by focusing on one persona goal, you're ignoring the other goals within that persona and you're ignoring all the other four personas as well. So the trick here is to stop thinking about your product as an object to begin with. And it really frees you up to think about the outcome better, because then it's not about personas and their persona goals, it's about you making one outcome happen more reliably. And then catering to the set of situations across personas that care about that outcome.

Samuel (01:00:54):
Correct. So the dog that wags the tail or the tail that wags the dog, the thing that determines the other thing, you can either say, "We have built a product that has a particular market, and that market has diverse needs. And we try to kind of satisfy the relatively one-size-fits-all experience of our product to better serve the general needs of our general audience." Or you can instead say, "We are in the resolution delivery business, and that we can index on one particular need or a handful of particular needs and just own the shit out of them." Because nobody's really competing based ... When we talked about main street of motivation, I'm not joking, it's the same idea of positioning your business in a brick-and-mortar sense in a place that gets a lot of foot traffic.

Samuel (01:02:01):
If you know that there's a motivation that's bringing a lot of users to your product, and especially if you know that your product isn't really fulfilling those, that's just such a no brainer growth area for you to focus on. In a way that's not creating habits or getting people addicted or tricking them into returning on day 20 so that you can say that you brought your day 20 retention up or whatever. Or relying on something vague like NPS or a number of other customer sentiment things. But instead say like, "When somebody shows up, can we have a general idea of what they're looking for?" And that's a process that you and I call "desire triangulation", where by paying attention to ambient information in addition to maybe a couple of strategically placed questions, you can get a pretty good sense of what the person is trying to do.

Samuel (01:03:00):
And then it's just a question of providing them with the preferable path to the outcome that they have. If they know that your way of getting to that outcome is going to be their preference, then you have their loyalty and you have them retained as a user. And if there comes enough times where you are not the path that works out for their desired outcome, especially if those are times where they go to you and you turf out, that is what drives churn. It's really almost as simple as that. Which isn't to say that it's easy, but it is possible and relatively straightforward to do. And it's something that not many people are really competing on right now.

Samuel (01:03:50):
So something to think about there. But if you combine this idea of super-outcomes, of the idea of identifying what it is that brings people to your product and how you can choose to design paths to better fulfil those desires, you can also identify which of those desires are worth more from a value standpoint through the same process of pillar two, Performance Valuation that we just mentioned. Like WordPress for example, I don't recall the specifics off the top of my head. That would be like a Rain Man level kind of thing. That's really speaking to a young audience, Rain Man. But anyway. Do you know what Rain Man is? The movie.

Yohann (01:04:36):
Of course.

Samuel (01:04:37):
Okay. All right.

Yohann (01:04:38):
I'm older than audience you're talking about.

Samuel (01:04:41):
We're trying to really get big on TikTok. That's our long-term plan. But when you sign up for WordPress, they offer you a number of different options like, "Are you trying to gain a larger readership or grow your network?" I can't recall exactly what they are, but different super-outcomes in our terminology. What are the main driving forces that bring you to us so that we can help fulfil those for you? And with that method of being able to measure ... If you can identify from one user to the next which outcomes they desire, then it's very straightforward to see which of those different kinds of users, A, what the distribution of those are. Maybe 80% of the people come for one reason and 10% come for a second and another 10% come for a third. Or maybe it's 33% across the board. That alone is information about what the distribution of what people are desiring when they come to you are.

Samuel (01:05:54):
But then you can also easily take it beyond that and measure what the conversion rate of those different intent segments are. And you can see not only on a volume level of how many people said that they were here for reason X versus reason Y, but you can also see on a value per signup level how much a signup who is coming to us for blank is worth versus a signup who's coming to us for blank two is worth. Does that makes sense?

Yohann (01:06:27):
Yes. That makes sense.

Samuel (01:06:31):
And so last point, just one last thing to just jump in and say. So instead of blending everybody together, the idea being that you really want to identify who are the people who are becoming really successful with our product? What are the commonalities that they have? What are the desires and intents that bring them to us? And how can we really just own the heck out of the whole path forward that connects them from where they are when they arrive to where we have created some intensely educated guesses around where they're actually trying to get to? And if you invest your design dollars in Path Design and your data dollars in Performance Valuation, that really gives you a platform to be able to tell what are the big motivational levers that actually drive the revenue of this company, because it's not providing people with access to a quality product. Nobody values access to products in any real capacity. It's always the value that you can generate with the product that creates the value in the person's life.

Yohann (01:07:44):
So we talked a little bit about how the outcome needs to be articulated in a specific, tangible way. And that's just one of many considerations that we've put together that you need to keep in mind when you're thinking about super-outcomes. There's more. And we should probably save that for another episode because we're running out of time at the moment. But if you're interested in something like that, let us know and maybe we could even do it next.

Samuel (01:08:17):
Ooh, what a tease.

Yohann (01:08:21):
I think that's good, a place as any to wrap up our conversation on the three pillars. Consider this a preliminary conversation on our three pillars. Rather than manufactured growth, think about growth in a holistic and sustainable way. That's really what we mean when we talk about healthy growth. We're thinking about how to create systems that make growth a natural occurrence of pursuing a path, rather than having to manufacture the path and really stress over aspects of the path in order to make growth happen.

Samuel (01:08:58):
Let's say you're Kickstarter, it would be very reasonable to say, "Let's try to figure out how we can get more projects past their funding point." That's a really clear alignment between the business generating revenue, but also the user's generating value for themselves as well. It worked out rather than it didn't work out in that case. So not only is it important that your entire company think about its role in growth and value generation for the company, but I would say it's almost even more important and healthier to rally your company around the major milestones of value delivery from the user's perspective.

Yohann (01:09:44):
And with that, we will call it a day for this episode. Stay tuned and we will be back with more Value Paths soon.

Samuel (01:09:54):
Keep fighting the good fight!







Want some hands-on help with your own app's growth?

Let's team up with our Coaching service!

Nah