Taking a more human approach to diagnosing churn and retention.

Aggelos Mouzakitis


Product-led growth lead


Aggelos Mouzakitis
Aggelos Mouzakitis

Episode Summary

Today on the show we have Aggelos Mouzakitis, Product-led growth Lead at GloFox

In this episode, Aggelos shared his take on product-led growth followed by the biggest misconceptions and mistakes companies make when they think about adopting a product-led strategy.

We then discussed why companies need to forget the data and analytics and take a more human approach to diagnose churn and retention. We dove into methods, tactics, and stages of the user journey that you can focus on to extract meaningful insights and help you truly understand the reasons for churn and not just the surface-level response we typically get in exit surveys.

Mentioned Resources



What is product-led growth? 00:01:59
The biggest misconceptions with product-led growth. 00:03:57
Humanizing our approach to analysing churn and retention. 00:07:09
Methods you can adopt to gather churn and retention insights. 00:10:00


[00:01:33] Andrew Michael: Hey, Aggelos, welcome to the show. 

[00:01:35] Angelos Mouzakitis: Hi, how are you, man? 

[00:01:37] Andrew Michael: I'm getting you for the listeners. Anglos is the product led growth lead at Glofox and all in one management software built to sport fitness operators of.

He is also the growth product manager at growth sandwich. The research agency, he founded as well as a product led growth mentor at growth mentor and gross was previously a product manager at octopus investments and the head of marketing and started bootcamp. [00:02:00] So my first question for you is what's your definition of product led growth.

And what is the biggest misunderstanding you think people have with the term? 

[00:02:09] Angelos Mouzakitis: Okay, very good question. There are lots of misunderstandings about playing them, the product that grows the product like that. My definition about a product led growth. Very simple. I don't, I don't have a handy by the book definition for you, but what is product led growth is when we lead with the product, when we put the UX and the experience of our users.

At the center of our attention and we lead with that. So the, what is product led growth? So it gets highlighted even more when we compare it with what is sales led growth and marketing led growth. So there is a principle that differentiates product led growth and everything else. And that is the principle of show instead of.

So we show what we're working on. We show our product, we try to demonstrate our value by [00:03:00] showing instead of telling, instead of arguing, instead of explaining that is basically the, I think Mr. Humane definition of product led growth, that doesn't include a lot of buzzwords. 

[00:03:12] Andrew Michael: Yeah, I definitely, I think that the ending there, the buzzwords, I think this is something for me.

That's really. Being a buzzword term, that's been picked up a lot. It's nothing really new. It's more just about like the framing of a method that people have been adopting for like four years. But definitely it's very, very powerful. And I think for me, one of the misconceptions as well, is that like it's one or the other?

Yeah. This is something that's like a lot of it comes out. We were product led growth company, and then just like, we don't care about anything else, but in retrospect, like that's not reality and that's not how you build and grow a business. Like you can just have one that's more dominant than the other, but really at the end of the day, you still need to be working on marketing.

You starting to be working on sales. It's not just good enough to have an amazing UX and express. Uh, There are 

[00:03:57] Angelos Mouzakitis: a lot of a lot of misconceptions in the market. Let's [00:04:00] start with with a basic on product led growth is good. That's the first misconception product led growth is what it is. It doesn't necessarily mean that it's good.

Doesn't mean that sales led growth is bad. So a lot of companies start committing to their investors that all we are going to become proactive. Without being ready, being product led without examining the fact that, Hey, your company is growing very nicely at the moment without being product led.

Do you really want to change that at the moment? Have you tested your readiness? They consider product led growth as a, as a silver bullet. It's it's the new growth hacking let's say as another misconception, very common in the market is that, oh, as soon as I have a product lead on board. It means that I'm product led.

So just the onboarding doesn't make you product led. And the third thing is what you mentioned, it's either one or the other, which is, which is very, very problematic because SAS founders meet realize that [00:05:00] a company can be 30% product led 50%, 80% plus. It really depends on literally on a dozen different factors on what is your product lead level that you can be in product like this isn't one thing specifically at the end of the day, it's the show don't tell principle.

So you can achieve that with multiple different ways. It's not necessarily sticking a user onboarding and going for PQS, 

[00:05:28] Andrew Michael: not just that. Absolutely. And just like, sort of it's many different aspects and inputs that go into it. Like somebody were chatting just before the show and the notion of like general retention.

This is one of those things that really bothered me before starting the show was like how people try to dumb down things to single digit single numbers, to single inputs when really sort of The challenges faced by companies when it comes to sort of growing the business and then retaining customers really, there's just so many different inputs that go into it in so many different aspects that [00:06:00] influence it.

So I think as humans, we try to simplify things to their own form and just give one definition extremity. It very rarely. Do the really successful companies that Excel with product great growth or retention understand that it's not just a single number or a single focus here. It's really about the full package and looking at and understanding all the inputs.

[00:06:20] Angelos Mouzakitis: Exactly. And especially in the product that there were functions come way, way below. Where we break the silos and we bring growth very close to customer success and customer success, very close to products. Everything starts being very, very interconnected. So treating different problems as these functions problems or that functions problems is isn't a very effective way of thinking.

[00:06:44] Andrew Michael: Exactly. It's the customer's problems we should be focusing on. Switching gears a little bit to, to the topic of today, obviously churn and retention itself. This is something you've had experience with quite a bit in, in your prior roles and where we are today as well. We chatted and we thought it'd be [00:07:00] good as well to talk a little bit about the human side of general attention.

And I'll hand over to you to sort of explain what you, what you meant by that. When we started talking. 

[00:07:09] Angelos Mouzakitis: Okay. Yeah, sure. So I come on a common theme that I, that I encounter access businesses that sustain stands to consider churn or retention as as a number they tend to detach from what that actually means.

And they see it as a number and they see it as a figure and they see it as a graph into their product analytics. But what we, what we fail to understand is that when you are using a software or a solution or anything in life, and at some point you stop using it, it's because you had some expectations about that, which either has changed, have changed now, and you have different expectations or your expectations were.

So lots of people tend to [00:08:00] treat churn reactively, to tend to give it as a task to customer success, to save customer success. When in fact they don't really treat the shores of the problem. This is the symptom journey is ultimately the expression of the problem, how the problem expresses itself and not the problem.

So if we, if we think of it like humans, the perspective starts changing. 

[00:08:28] Andrew Michael: Absolutely. I think definitely. The ability to empathize with your analytics is something that it's a very, very hard skill. When I was previously as well working a tacho and a head of business intelligence there, one of the things I was really pushing for and encouraging analysts to do was actually go out and speak to customers and and truly understand what their pain points are because.

You can look at a graph and you can like understand what Shannon retention doing. But you only really see like what's happening. You don't understand the why. And having that additional context behind really helped [00:09:00] you and make sure that analysts to being able to really empathize with the user, be able to empathize with the pain points and the problems.

And I think it's something very, very healthy to do. And something that I think a lot of companies go through sort of this maturity stage when it comes to analytics and it's sort of like a. And Alex is everything. And then it's realizing it gets slow. So like, we don't understand the why behind these things.

We really need to learn on top now research and my view on this as well is that there's a big problem that companies make. Today's where we, we separate these functions, where we have UX research on one side of the company, we have a product analytics on the other side, and very rarely do they work together.

They're really, they bring together the Watson and the Y The numbers and the people behind a downturn. So let's dive into that a little bit. Like, what does your process look like? When going into a company to try, trying to understand sort of what are some of the, the why's and what are some of the reasons for churn that where expectations are not being met?[00:10:00] 

[00:10:00] Angelos Mouzakitis: Okay. When a company comes to me with a churn problem of course the first thing that I have to do is Examine the problem and check all the relevant data and see if they actually have a term problem. A lot of businesses think they have a problem, and in fact, they might not have the problem or have another instead.

So what the data are saying, let's assume that the data is telling us that there is indeed a churn problem. Then there are a couple of questions that that arise Is it a temporary trend problem? Is it the term problem that started now that was persistent in the past? How is how's marketing performing?

Are we managing to bring customers in? Are we struggling to bring customers in? Do we have a situation where we it's easy for us to bring new customers, but then we're easily losing them as well. So just to, just to shape a picture of what's going on across the. As soon as this picture is is a shaped, then I typically follow the process of analyzing the value gap.

So what is the value gap? The [00:11:00] value gap is basically the incongruence between expectations that people come in and reality. So what are the expectations that you were having and what is the reality that. Practically speaking for me to, to do that. I start by interviewing. The power users of the business, the best users, best users are different across different businesses.

Typically on a B2B SAS, they mix in order to define the power user. You need to mix both engagement, data and financial data and retention data. So we need to understand what is it, what is the baseline here? What is that the best customers are telling? Because by establishing that we can do, we can then highlight the difference between that and anything else.

So as soon as we ha we, we established that, we start speaking with churn users, fresh churn users that remember their disappointment. Then they remember what. [00:12:00] They stopped using our solution. It's very important to be fresh because you will realize by doing that job, that people tend to either forget the reason that they churned or they might attribute.

Two different reasons. For example, 

[00:12:15] Andrew Michael: the way you started, the way you're describing it as though it sounds like you're picking off like a tomato of the tree and this fresh churned users, I'm really getting the juice from them. So I go on 

[00:12:28] Angelos Mouzakitis: the people are attributing the reason of their churn too, and I'm pretty sure you will empathize with that too.

It's too expensive for me or I'm not using it anymore. These are the two most common reasons that people stop using a solution. The question is why did it start being too expensive for you now and too expensive compared to what? That's the first question. And then I stopped using it. Why did you stop using it?

When did you start? When did, when did you stop using it? When did you [00:13:00] consider stopping using it? What happened. What did we fail to do so businesses don't really recognize that instead what they do is they take the feedback at the face value and they're like, oh, we are too expensive. We should get cheaper.

Or people are not using us anymore. Perhaps we should bring more people I dunno, all sorts of different solutions. Yeah, 

[00:13:24] Andrew Michael: for sure. And definitely, I think that face level. Problem the load companies and a lot of companies are so very early in their understanding of what user research is and how you can really go deep and understanding sort of the jobs to be done.

And the pain points of your users, because it is very easy to sort of put out a survey at the end when people in and say, what is the reason, and then to realize, okay, everybody says it's too expensive, but like you say, like there's a reason at some point they signed up for your product and it wasn't too expensive at that point in time.

So what changed to in between that time and. There can be a number of different reasons. You alluded to some of them like [00:14:00] marketing and the promise fit, not being there, potentially sales, overselling, and that solution at delivering. Like these are some of more of the the deeper reasons. And I think like what technique do you typically then go in and pro to really try and understand and go deeper below this.

[00:14:16] Angelos Mouzakitis: I use both qualitative and quantitative data. And I also use both the ad hoc research, but also try to establish processes that bring these data automatically back to the business and create feedback loops. So, as I said I will start with user interviews in the same time I will tweak the CA the cancellation experience, but also the onboarding experience.

To collect, of course not the depth of the data that I can collect with interviews, but to collect expectations that users come in with and expectations that are not met. And in the same time we might run surveys exactly for the same purpose as well. So in that case, we have both qualitative and quantitative data go hand in hand.

The one are telling us the. [00:15:00] And they're giving us the directions here is where you have to look further. And the qualitative gives us the why the deeper why's and following that process, I was surprised in the past to realize that a very, a very good number of people, of companies that suffer from churn, they actually suffer from positioning.

So they might incorporate a lot of resources or they might put pressure to their customer success team. While instead they should have they should have done a reality check to their positioning, to the expectations that they create, because at the end of the day, what they did, where was trashing Sasha.

[00:15:41] Andrew Michael: Yeah. I think that's like for me positioning, it's like really surprising. The deeper you go into sort of understand the problem, general attention, how much of an influence really does have a, on the final outputs and it's. Most people I don't think would even think to think about when it comes to churn and retention of how much influence it [00:16:00] really does have on the product.

And you can position a product in one way and just have terrible churn, terrible retention, and then just slightly tweak the messaging, potentially slightly tweak the audience and you can have a great business. So words, words, like really, really matter, I think in this case, so. I liked as well. The sense that you're, you're looking at it from both the lenses, you're asking questions, like what are some of the questions you're asking in these user interviews on both ends and what are these surveys?

[00:16:29] Angelos Mouzakitis: Okay. Okay. About the questions, the line of questioning is pretty easy. It's pretty easy to understand for everybody that that is listening to that to this conversation the difficult part starts from the attitude of asking and bias questions and the specific wording that you have to use in order not to bias the answer that you're getting.

But the end of the day, it's very simple. When we speak with power users, we are. What is it that you are achieving with us? What is the. That you are doing with us. What is the job that you are doing? What is the [00:17:00] value that you are getting, we want to establish? What is it that they perceive as value?

What are the jobs that they're doing with us, how their life is changing by our solution, then there are some other follow-up questions. This, these are typically easy conversations because these people like us so we can ask them okay. Before starting with us. Well, what other solutions have you considered?

To do this job or to have this value and they don't necessarily need to be software. They do. They can be people that can be processes that can be. So they start giving you alternatives, not competitors, alternatives, and then considering that these are good good users in the last week, we might start questioning okay.

What are the goods and the bads of each of these options? And then lastly, what I do with power users is I ask them to give me a use cases. So I sometimes might ask them to share the screen and show me how they use something, or they might tell me different scenarios that they are using the. [00:18:00] And then I try to pair these up with their demographics.

So the jobs to be done, mindset typically conflicts with demographics. But in that case, it doesn't because it's a reverse engineering of the situation. We don't lead with demographics. We lead with jobs to be done. And then we try to correlate that with demographics. For example, if you are, for example, zoom, let's an example, a company.

If they do the research and the realize that, you know what, there are certain jobs to be done, that we are delivering well, but we are doing very well for our customers and with jobs to be done, typically are found there that's the correlation of jobs to be done on demographics. So then I do exactly the same thing, but open.

To churn, churn people. So what were the expectations that you had that we didn't meet? What was the value that you expected that will then give you? What were [00:19:00] the jobs that you wanted to do that you didn't manage to do and regarding alternatives? What are the alternatives that you are considering now?

To do that. So just opposite, opposite line of questioning. What are the goods and the bads of of these alternatives? The difference with Jenn users is that I try to uncover the timeline of actions. So that's why we need it to be fresh because I will go back and I will try to uncover the events that push them into actually performing the churn.

Because a little note here, when you see the. It's not the time that customers decided to churn. It's probably just the time that they realize that they're still paying for that, but they, but they don't want to pay for that at the time of churn probably is weeks ago, even months ago, you just didn't know it because you didn't monitor your activity churn.

You didn't. You didn't proactively deal with air, [00:20:00] reduce in engagement, which was leading to a customer. For 

[00:20:05] Andrew Michael: sure. I think this is one of the biggest reasons why it is so difficult to reduce churn at the point in time where people have actively decided to cancel their accounts, because it's not that point in time when they've actually actively made the decision.

The decision probably happened like weeks and months ago. This is just a point in time when they've actually. Been motivated enough to go and take out their credit card or delete their account. And, and at that point it's almost always too late. I think. There's obviously things you can try to win people back, but it should really be focused earlier on in the last cycle, early on in their journey.

How can we activate, how can we make sure that there's adoption there's engagements and that sort of prevention is better than churn religious focusing on never getting to that. Exactly. 

[00:20:45] Angelos Mouzakitis: Exactly. I totally agree with you on your cancellation experience needs to address that. Anytime that I was involved in cancellation experience designs, we had to incorporate both deflection tactics, save tactics, but at the end of the day, when the user [00:21:00] decides to churn, the best thing you can get out of this user is feedback.

[00:21:07] Andrew Michael: And what can you then I think one of the areas where I think for me is also interesting on top of this as well. And you alluded to in the beginning as well, like the expectations really also looking at newly converted users and going into understanding, like, what is the expectations that they have for the product?

Why did they sign up today? What nearly stopped them from signing up today? These are things that I think also really, really interesting to like really complete the picture then. So you have churn users, you have really power users and you have new users just coming through the door to understand sort of the full picture.

I even like to take it a step further and even look at panels then like people that are considering the product or the service and what are some of the things that they're considering. So you have. The full journey and lifecycle of the user understanding the different pain points at each point in time, the motivations and bringing it all together really gives you like clear picture into like who these people are.

What they're trying to achieve and clear is probably the wrong [00:22:00] word, because it is completely different as well for different individuals. And that would just sort of be bringing it back to the data numbers again. But for the most part, you get a really, really good understanding across. 

[00:22:11] Angelos Mouzakitis: Exactly. And a very empathetic understanding, which which bits every graph of this world because you, because when you start speaking with fresh users, just to complete the value gap analysis, you start understanding.

Fresher, we follow the path of a power user, or if a churned user based on what they told you, you, you sort of see the future. If you perform this analysis, now the, the magic part starts when, instead of doing that ad hoc analysis, which is okay for you to do it's very useful, but you can't do it every week.

It's a, it's an expensive thing. And I know that businesses prefer to snooze this qualitative research. So it's always better if you create processes and. That bring these data automatically to you. So just a little example, your [00:23:00] constellation experience asking about expectations, not being met in the right way, your cancellation experience asking about alternatives.

So couple of questions that you can ask instead of asking what is the. What is the competitor that you are choosing from now on and listing their names? You can describe solutions. You can say the, the solution that he helps me do that. He helps me do this. He helps me do X do Y so that was a little change that I have done in a previous client instead of listing their competitors, we listed the description of their company.

So that we could understand what is the job that they wanted to do with them. So that's one, then on the other side, you can go to onboarding and ask for expectations and demographics, of course. So by mixing expectations that they have, and that you source from the onboarding with demographics and jobs to be done, that are not met and alternatives with demographics on a chair on a [00:24:00] cancellation level, then you can do this sort of the same analysis, automatic.

[00:24:05] Andrew Michael: Yep. You can automate it and then continue to feel good insights to the team on how to work on things. Another thing is, what I think is also very interesting to look at is in the context of the churn exit survey, where we give a lot of focus to, I think, in general retention. And again, often it's too late is also the inverse is really like, depending on your cycles, like if you're doing annual contracts predominantly.

Posts renewal asking the question as well of like what nearly stopped you from renewing this time round. Because that way as well, you're also likely to get really, really strong feedback from people that are still managed to stick around. So it might've been a couple of things that were really just pissing them off about.

There may be needed it for too much into way. So asking those users will help you prevent a lot of the other, the churn that you see on the other end, because typically if some of the users or the users that are renewing their product and they're getting through, they're happy with it to some degree, but there are things that they're unhappy [00:25:00] with.

That's probably the greatest time you'll get some really, really strong feedback. What you can do to improve, because I've just laid out my credit card for another year. Like, I want to see some of these things fixed now as well. And instead of focusing the attention just on the, the turn exits, it's really about, okay, the people that survived, like what do they need after a year?

Because you mentioned the beam in the shows. You might come to us today with a problem sets and you have the specific use cases and needs, but your use cases and problems evolve over time. As you mature in a product or service, your understanding becomes different of the markets or the space of the challenge, and you have new challenges.

So thinking about that as well as another one, you can there on top to even further clarify that picture, I think there's so many different techniques you can employ, but I think the general message is really like. Speaking to customers gathering feedback like me, I think it's hands down.

It can be beaten. Like any analytics that you're looking at Alexa are super valuable and useful to give you a pulse of what's happening in the business for you to understand what is happening, but to understand why you [00:26:00] got to be speaking to customers, you've got to be collecting feedback. You've got to be on the ground.

[00:26:04] Angelos Mouzakitis: Ultimately the, your product analytics and your Google analytics and all sorts of analytics will give you the direction of which are the right people for you to speak with. Especially if you have a lots of use cases that you're that you are delivering. If you have for example, if you are a category software that delivers across a horizontal.

Or across a vertical category of clients. And you might have lots of use cases. For example, glove box, the company I'm working for at the moment, fitness software. They have a lot of use cases because there are a lot of different fitness software. So in that case an initial quantitative analysis tells you, which are the companies, the businesses, the people that you have to prioritize.

[00:26:48] Andrew Michael: Yep. Absolutely. And that's when it's most powerful is when they used together to be able to bring qualitative and quantitative into the business. So I see we're running up on time. I want to make sure I ask a question. I [00:27:00] ask every guest let's imagine a hypothetical scenario, which has probably happened to you in the past is you joining your company churn and retention is not doing great.

At this company, the CEO comes through and says, angular like will need to turn things around. We have 90 days to do it. You're in charge. What do you do the catch? You're not going to tell me I'm going to speak to customers. Like we've been talking about this whole episode and understand the pain points and go from there.

You're just going to pick one technique that you've seen. That's been effective in reducing churn fast. And you're going to run with that. Blindly, hoping that it works with this new company, what would you pick to try and reduce? 

[00:27:36] Angelos Mouzakitis: Mm, one technique. I would I need a car. I need a few seconds to think of it, but if I don't have this few seconds to think of it, I would use deflection tactics.

So what I mean by deflection tactics for every user that is visiting the cancellation page and that initiate. The cancellation experience, I would experiment with [00:28:00] different offers. So not just offers, but those are reminding the user of the benefits that they're going to lose. So reminding the user of what they're going to lose and the couple of offers to save them that could be a lot of different things that could be a little discount, but it could be an extra feature that could be anything.

That's why I said experiment. So. Deflection tactics, because when you have a next number of churn, there is a wide number that is part of that. That is not saveable, but you can do nothing about it, that they will live anyway. But then there is another part of churn, which is actually very big, that either is slightly disappointed or there is a misunderstanding between you and the.

So by by trying a little bit during their disappointment to save them, you might actually save a big, a serious chunk of them. 

[00:28:54] Andrew Michael: Yeah. 

[00:28:55] Angelos Mouzakitis: And now that I'm actually thinking, I might have chosen another tactic, [00:29:00] another tactic, another tactic monitoring the activity churn. So I would also. Or I would replace that.

Instead of monitoring the customer churn, I would start monitoring and sending signals to customer success. People when actively. Gets significantly reduced. Actually, that would be the one that would be one. 

[00:29:21] Andrew Michael: So you had focused in to enable, to sort of prevent Chen's tracks, seeing when customers stopped becoming engaged with the product to service and then push it to tier.

So they can sort of start working with those customers. Exactly. 

[00:29:35] Angelos Mouzakitis: The focus would be on prevention either one way 

[00:29:38] Andrew Michael: or the other. Cool. What's one thing that you noted about tryna attention that you wish you knew when you got started with your career?

[00:29:44] Angelos Mouzakitis: One thing I know it's something that can be caused from everything it can be, cause it can, the source of churn might be your positioning, your landing pages, your sales page, your price. Your product, of course your customer success. It can be everything [00:30:00] and anything. And it's also everything and anything it's everywhere.

That's a good insight that I, that I should have had earlier. 

[00:30:08] Andrew Michael: Absolutely. I think this is like, we talked about this role, but like a lot of companies in the beginning of this decade, it's customer success is job. God, and do it, but really. Everything impacted. It's the whole premise of the show as well as really just to sort of like there's so many different inputs and to really dumb it down to a single metric is really as it unjust unjust, not even a word, but anyway it's really just not a great great way to view it.

And coming from that lens of understanding that, okay, there's just all these different inputs that are influencing. It is like that's when alignment comes in with the company where people understand that. Something that I'm doing in my role and supports and something, someone in marketing both has an influence on general attention.

And if we work together and we have a good alignment of what we're trying to achieve, like that's, when I think you really see step change in a business, when the company is aligned, they understand how each of their roles influences the final metric and they can go out then in [00:31:00] unison to sort of reduce it.

Okay, well, I'm glad it's been a pleasure chatting to you today. Is there any final thoughts you want to leave the listeners with and how can they keep up to speed with your work? 

[00:31:11] Angelos Mouzakitis: Any final thoughts? Yeah, try to be more humane I, we are all in the tech industry. I know we are all into our graphs and into our numbers and into our very stressful jobs.

But we are we are ultimately building stuff for other people. So we have to, we have to remember our human nature and start exploring problems as. You can find more about me in my website growthsandwich.com. And of course, if you want us to geek out about a journal or anything else, feel free to find me on LinkedIn and just send me a message and let's speak 

[00:31:44] Andrew Michael: very cool.

We'll make sure to leave those notes and anything discussed today in the show notes for you to catch up on. But yeah, thanks so much for joining today and I wish you best of luck going forward.  

[00:31:54] Angelos Mouzakitis: Thank you. 


Aggelos Mouzakitis
Aggelos Mouzakitis

The show

My name is Andrew Michael and I started CHURN.FM, as I was tired of hearing stories about some magical silver bullet that solved churn for company X.

In this podcast, you will hear from founders and subscription economy pros working in product, marketing, customer success, support, and operations roles across different stages of company growth, who are taking a systematic approach to increase retention and engagement within their organizations.


Listen To Next