Setting up your data stack for impactful retention analysis.
CEO and Co-Founder
Today on the show we have Claudiu Murariu, CEO, and co-founder of InnerTrends.
In this episode, Claudiu shared best practices on how to get started setting up data analytics tracking, metrics and properties to consider, and why a tracking plan is essential.
We then discussed how to identify and select your onboarding and activation metrics for companies at different stages in their growth cycle and we finished off by discussing how to effectively analyze churn and retention with your data.
[00:01:23] Andrew Michael: Hey, Claudio. You're welcome to the show.
[00:01:26] Claudio: I Andrew, very happy to be here.
[00:01:28] Andrew Michael: It's great to have you for the listeners. Claudia is the founder and CEO of inner trends, a product, and it looks company with plug and play pre-built analytics reports designed by soft growth experts and built by data scientists.
Podio is also a partner of micro analytics where you spent almost 20 years helping companies like orange ING and Samsung with their data and analytics. So my first question for you. What is the biggest data challenge? You see SaaS businesses facing.
[00:01:57] Claudio: Well, it's not the [00:02:00] analysis itself, which might come as a surprise for many, but rather structuring the data and preparing it for analysis.
They often say that data scientists spend like 80% or 90%. It depends on who you ask on data cleaning and preparing. And a lot of the people that get into the industry and want to analyze data, don't realize that. So without having the data structured in the right way and without having it clean, Or validated, it's almost impossible to find insights or patterns or anything valuable in it.
So that was one of the main challenges we knew we had to tackle. If we were to launch pre-built analysis, basically allowing people to click on a link and get the report without doing anything without building the report. But that only works if your data is well-structured. Clean and validated. So yeah, that's something when
[00:02:58] Andrew Michael: I had, [00:03:00] yeah, it definitely is a huge challenge.
I think we previously Idela Dorfman who was at segments at the time where we talked a little bit about their user onboarding experience and how they actually added friction to it and slowed the onboarding experience down. By forcing like bigger customers to go through an experience where they created a data tracking plan, because what they'd realized was those that just got started very quick, ended up creating the mess that segment was designed to fix in the beginning.
And how important sort of the data integrity and data governance is really starting off with a good, solid tracking plan. Uh, is that the best way there? So you've obviously then experienced and seen this quite a lot yourself. Um, What are some of the recommendations. Now, if, uh, people are thinking about a cow, want to go about starting to understand much inner intention better.
I want to start my understanding that in looks like, where do you get started with tracking? Like, what are some of the events and things people should be thinking about, um, to set them up, to be able to do some of the more complex things later. [00:04:00]
[00:04:00] Claudio: Yeah, no good question. And I'll start from what you said about segment, because if I love something, but it's good friction and I can't recommend enough building good friction inside, inside the product.
Um, and for us, it was the same. Like when we started, you know, trends, we started with a managed setup. So we said, Uh, before we got to product market fit and to understand exactly how customers would get value and so-and-so will do the setup for you. Uh, and we actually witnessed that friction ourselves, and we saw how difficult it was to get there.
And yes, we got to the same conclusion like segmented with which was you need a tracking plan, but we actually. Went even further. What do you put in the tracking plan? How do you decide what events to track? How do you decide what account properties or event properties to add to the tracking plan? And that's when we that's, [00:05:00] when we, when we realized that you need to start from the business needs.
That's something that customers understand their needs. That's something that's very clear to them. So we put another step in front of the more friction, good friction, which is what we call the customer journey metrics map. So before you track everything, anything. You need to build your customer journey.
And we go for the five pillars of growth, acquisition, activation, retention, revenue, and referral. And for each pillar, we define different stages. And for each stage you need to define steps or metrics. And I'm going to focus on one of the most important products, which is. Activation onboarding under activation.
We have two stages onboarding and goals and onboarding. We have a very strict definition for this, especially to help customers get to the bottom of it, because [00:06:00] there are so many ways in which you could define on boarding and the definition for board onboarding processes critical for. Analyzing retention and verify churn.
So our definition for onboarding is the process that takes people from creating an account to experiencing the promise of your product for the first time. So you need to go and brainstorm with your team, not brainstorm, cause you don't need to brace where you need to align with your team. What's our problem.
What are we promising our users? And what's the first time, when is the first time when somebody experiences, but promise? Well, that's the first core event you need to add to the tracking trend? Because this probably one of the most important ones, the first time it happened. You have a user on important when it happens repeatedly, you might get to your north star metric.
And when it happens on a regular basis, you have high [00:07:00] engagement. So that action alone, if you could only have, if you would need to have only two events tracked in your product, it would be created account and experiencing that promise. And you could get so much value only from these three.
[00:07:16] Andrew Michael: Yeah, I really liked the simplification and the focus on a couple of events.
Cause I think this is one of the problems that I see a lot of, um, people getting started, uh, getting set up with analytics and metrics is they start thinking, okay, we need to track everything and we need to have an event for X and Y and Z. And what if we don't have this? Or what if we don't have that?
And they end up sort of. Digging their own graves before they even get started by creating this mess. And it really is sort of that simple, as you said, like there's maybe five to 10 metrics over time that you're going to focus on the most. And the, the other ones are just really peripheral, a little things that add very little value in the grand scheme of things.
And I liked the focus that you mentioned, like really just picking it to that. What is that one onboarding metric that we know? Like that's the value that customers came to us? I think [00:08:00] like, Exist today, we develop products and services to deliver value to our customers. They have certain expectation of what that looks like, and if we can measure that with a single metric, it's, it's amazing.
Um, and most products you can do, uh, because, uh, we software as a service businesses typically. So the. Going about determining the net, then it's just an alignment issue. You said like internally, cause like there'll be some sort of debates on our, this is the value that our customers believe this is the value.
Um, how do you sort of deal with those issues internally of like really deciding what is that one key onboarding metric? Cause I'm sure you've seen a few of these discussions. Yes,
[00:08:38] Claudio: yes, yes we do. And that to two different directions here, whereas one direction of the. Uh, companies that already got to product market fit.
So they have acquisition works well for, and the market has a need for their products and they come in whenever and even those [00:09:00] companies are not aligned often, like marketing would see save it. Uh, our value is something and sales to see something else. Uh, so. I always go to them. I say, open your homepage.
What's the big heading on your homepage saying, what's the action in your product that delivers that. Cause that's your promise. That's what people come and see you and they want to buy it. Uh, and often that a heading is really well written because they got product market fit. Those companies. Uh, no, what we are doing, but then you have the companies that didn't get there.
And when you go to the heading of the homepage, even them it's abstract, they don't really know what they is, the value of what we are delivering. And for them, it's kind of going to the jobs to be done framework. I can need to find out what's the job. What's the problem that you are fixing to your users and ventilated to the premise.
So you kind of need to go for that framework. I [00:10:00] actually had a great conversation today with a general Audi from forget a funnel about this. Uh she's uh, she has a great framework on getting to, um, to vet, uh, primary. And yes, once you identify it, you put it on your homepage because yes, you want people to see it, and then it becomes the definition of your onboarding process.
Uh, but then these a scenario when the promise. It's not easily tracked with a single event and let's say a reporting product product, uh, I know stock market reporting for. I go, I signed up for a SAS where I get reports on the stock market. The promise is I'm going to give you reports that will make you make better investment decisions, but what they event that I delivered my premise, because simply loading a report doesn't mean that you actually understood it.
So you need to [00:11:00] find a proxy in those situations. And yes, that's where we kind of come in and help. For example, for such a product, a good proxy is that. I created a report that I came and viewed on a secondary through my own. That means that I probably understood something for that. And the chances for me to understand something from that report has much higher or much higher.
If I come back on a second day to sit there, to see that report compared to just open the report and that's it. So yes, you need to sometimes look for proxies for. But most of the times it's straight.
[00:11:40] Andrew Michael: Yeah. Um, I definitely see all three of those cases very clearly. Like you said, like when you have product market fit, it's clear, like everybody on the team really knows what the value is that your product is delivering.
I think though, sometimes in product market fit, you have a similar situation, uh, to when you don't have where you might have multiple use cases of the product and yeah. Do [00:12:00] you have different audiences coming to you for different reasons? So like, uh, at Hotjar we had this way, a hot shot at the time when I was there, we still had like eight products in one.
Uh, and obviously like, I'd say 80% of our users came for, um, heat maps and recordings, and we could say, okay, yes. In one way that. The an activation onboarding metric is like, once they've seen their first heat map was in their first recording. Cause they've got the value now, but this other group of users are just signing up for feedback.
And I think like these are some things that in the beginning, maybe because, you know, 80 or 90%, that's the use case focused on that. Uh, but then over time you can layer on that sophistication of like understanding of why did they come to us to begin with, and that's where you said, like the jobs to be done.
Um, framework is very, very powerful. Even at the stage when you do have product market fit, it just helps solidify and, uh, further, um, give clarity to.
[00:12:53] Claudio: I live in dive deeper into VAT because, uh, I do see that scenario a lot. And I should [00:13:00] say for bigger companies, the established companies and what we do there is actually still focused on a single definition, but on an abstract one, people go to Hotjar to learn how.
To improve products, maybe vet that's the promise like, and they can do that through multiple tools. I can do a heat map. I can do a video recording. I can do multiple things, so you can go and abstract thing. View a report might be on a second time, or if you free different reports from three different angles.
So you can go for ups accusation to find a single definition that covers all the eight scenarios, because at the end of the day, Hotjar. The purpose of Hotjar is the same. It's not doing also counting and banking and, uh, analytics supporting or phone. It only does one thing and it is difficult. You need it better when you [00:14:00] need to go for a bank brainstorming session, what the abstract visitation of our onboarding process, but tells us that people got value and it actually made them.
The tracking plan so much easier to configure after you don't get hundreds or tens of, uh, events that need to be tracked. You actually get to a very small list, but makes a lot of sense and a couple of event properties to distinguish everything. And yeah, it becomes very beautiful.
[00:14:27] Andrew Michael: Yeah, definitely. And I, again, like you keeping things as simple as possible, I think that's like the message.
If you're thinking about putting together a tracking plan is like, start to Claudia saying there's even like one or two events and then layer on from there. And you're going to get so much more valuable than that than trying to think. You need to think about everything from the start. Uh, you mentioned properties then as well.
I think this is another area where. When we start to think about like sending events or something. Okay. What all the properties can reconsent and what, all the ways that we're going to want to use these. What is some of your advice that you have for people thinking about like what event [00:15:00] properties to be sending with your events?
[00:15:03] Claudio: So, um, let's we first focused on account properties and then we go to, and properties, uh, account properties are much more important because they allow segment. And segmentation is very, very important in, uh, in data analysis. Uh, now it entertains, we are cohort-based by default. And when you do cohort based with segmentation, some real magic can happen.
So first thing you need to focus on is the account properties and the first account property that you should focus on from our perspective is. To, uh, on the, on the account properties that come from the onboarding process or which, uh, which are related to the qualification of your accounts, a lot of businesses have questions that you need to answer when you sign up.
So those are probably one of the most important account properties that you [00:16:00] should track that segmentation alone will give you a lot of insights of who should you focus on? Who. Where is easier to onboard accountants on when we move towards event properties, we look at event properties as way to, uh, to give more context to the event themselves.
So the idea is to, uh, to have everything simple at the event level, we typically have all events. Covered in three categories, core events. We are yet to find a business that needs more, that has more than 40 core events. And those are really big businesses. Most businesses have five to 10 events, core events, um, venue.
We have UX activity, UX events, which is. That what everybody refers to. Like you need to track everything. Yes. And [00:17:00] they S UX activity, track every click, every page view everything, but you can identify with accuracy, so no false positives fair. Um, but those UX events are not so important as the core events, the purpose of the UX event is to identify.
What influences users to do more of the core events. You don't even need a hundred percent accuracy on the UX events, because mostly it's mostly statistics fair. And the first type of data is integration data because a lot of your customer data sits in Stripe or Intercom or in other like marketing automation, HubSpot CRM, and so on.
All of the data needs to come together because it's, it says. Now even properties we mostly care about for the quarterback and for the core events, we want to add the context that makes it very clear for us. What happened, an amazing, a great. [00:18:00] Tracking plan is a plan that when I, uh, it, it generates a tracking that when I go and analyze a user, by just looking at the event and event properties, without seeing images, I get a very clear understanding of where he was and what he did.
So open any user profile in any analytics tool you have what we'll look at what he did event after event and ask yourself, is it very clear for me? What. What he did. And if it's not, it means you are not tracking the right event properties for the right event names.
[00:18:38] Andrew Michael: Can you give an example of Blackwood, a profile would look like then,
[00:18:43] Claudio: so let's do the onboarding process.
Yes. Created account would be the first, uh, the first event, uh, then would be, uh, a UX event. Um, let's say onboarding questions screen. Uh, w we, we we'd say [00:19:00] screen onboarding questions. That's what was loaded, uh, onboarding questions submitted would be the next event, but under event properties would be question one, answer, question to answer question three answers.
So you would know exactly what liens are answered then would be another event business pro screen business profile, uploaded picture. Um, uh, set a business name and you could put the business name fair, and I hope this gives an idea of how the tracking should look like. You'd have both UX, so different screens, loaded clicks that happened, but you'd also have what was saved.
Uh, and, uh, those are typically the quarterbacks.
[00:19:45] Andrew Michael: Yep. And then you mentioned as well, like the credit counts and, uh, that, um, that's why as well, you'd want to be sending some of the like firmographic or demographic data you're collecting at signup. So company size, um, role that they're currently in, uh, [00:20:00] employees and so forth.
Yeah, exactly. Um, cool. So. We established and we, I think we both definitely agree that like, if you're thinking about trying to track and understand, you're trying to retention from a data perspective, like the best place is always to start with having a very solid tracking plan. Uh, you mentioned as well then that like one of the carers that you always say to focus on is the activation and onboarding, uh, experience.
And like, just get that in. Let's say now we've managed to add our first couple of events. We have the account credit and we have, uh, the value prop, uh, received for onboarding. What can I do with these metrics now? Like what can I start to understand about my users?
[00:20:42] Claudio: Okay. So, uh, let's focus a little bit on maybe, cause that's a lot of companies that come to inner trends.
Say we have a channel problem prior to signing up. I like it will be [00:21:00] a statistic that comes out of my head. I didn't really do it manually, but I would say that probably 80% of them are wrong. Very few companies that come to us actually have a retention problem. Most of them have an onboarding. So the first report that I always recommend people to go inside in our twenties.
It's a pre-built reports. How are people converting during the onboarding process? Because in our trends knows your customer journey metrics and has your tracking plan and everything tracked. It just generates the report automatically without needing to. Manipulating the data. And in that report, it will tell you how many people get to the end of the onboarding process and not what step you are losing most of them.
So from here, there are two directions you should go first is why should I care about onboarding process? Uh, and that takes you to answering two questions. What is the retention of the account, but finished the onboarding process and other prebuilt report by default, we only care about the retention of the [00:22:00] onboarding.
But to prove to you that it's so important to look at the retention of the onboarded accounts. We also deployed a, another pre-built report, which is what is the retention of the accounts who didn't finish the onboarding process. And I'm amazed at how every time when people load that report and it goes to zero.
In week one or two, uh, they are like, wow. So important to have accountant boarded because yes, if you don't have them on boarded, there is no reason for them to use your product. I think that a message that came across a lot in your, uh, in your, in the, in the Turner fan podcast. So, uh, Yeah. Yeah. Like when we see that, then they immediately go back to the onboarding process.
Like how do we improve it? Um, and the, the onboarding process, we identified the steps that were used, the most users and where we have another pre-built report, which is one of the most powerful Latina trends. What are the actions that people do between onboarding stuff? [00:23:00] So vet, uh, automatically categorizes all the activity that people do between the step where, uh, uh, to the step where they use most accounts and the next one, and all this as those actions are categorized in three buckets actions that are specific to accounts that are successful actions that are specific to accounts, where to dropped off and action.
Uh, not specific to either of the groups from that important only kind of know immediately what you need to do in your product, in your communication in order to address the onboarding problem and the last, uh, probably, um, Pre-built report that people want to go really quickly though. It kind of needs at least four months of data to have gathered is what are the differences in actions between onboarded accounts that return and those that return, uh, vet report came, uh, [00:24:00] our need to have a clear definition for the north star metric.
North star metric is an abstract concept. And when we looked at it, we said, North star metric should refer to a very specific metric that. We find it in successful people, but we don't find it in people that churn. And that's how we came up with the idea of this pre-built report. And yeah, w w you know, the, the north star metrics of Facebook or slack or Dropbox, they have been around for a long time.
They're like a legend now, but very right for every single business where we ran that report, it's always. Two or three actions done repeatedly. A number of times that make the difference between people that return and those that return.
[00:24:53] Andrew Michael: Yep for sure. This, uh, actually was one of the reasons why I started, uh, the show, [00:25:00] because I think it definitely applies like there's certain metrics that you can, uh, solve for and focus on that.
I think that the problem with that is that people fixate on this idea that I can need seven friends in five days and that's going to solve all my churn and retention problems and it really simplifies. The thing, which is nice, cause you give everyone a common target, but I think it oversimplifies it for most businesses.
Well, because it's not always that clear or that simple to get those five friends in seven days and understand what that is. And also there's also so many different inputs that go into the final output metric. So it's not just about getting those actions, but there are other aspects to it. It's like, here's your marketing on points?
Like how is sales trying to put a seal? So you're closing the right customers. So there's a whole bunch of other things as well that happen in between. Uh, but yeah, I a hundred percent agree with. Uh, on the, the sort of the onboarding actions. It's actually funny because I'm going through this exact exercise now for my startup.
So, and it's just as a result of like listening to the podcasts and, uh, like having all the guests on the show is really like how [00:26:00] important is onboarding? And it is really day and night. Like when we look at our data, we see, okay. Those that went through onboarding and completed a key action that we see as a psych, one of the main value props of our product versus those that don't like, the retention is like off the charts is like almost zero for people that don't do it.
It's normal. Like. And were looking for a service. They didn't do what they were looking for. So why should they come back? And those that do it is really, really strong, but on aggregate because we have a freemium product retention looks really bad, uh, as a result. But when you have be able to segment those out, you can only see.
There's product very sticky. Uh, and now it's just a matter of like, how do we optimize our onboarding experience to get people to establish that value. And actually just last week, when we run out a really, really successful experiments where we increased activation by 180%, uh, so, uh, but it was a good one always to be fair, like the onboarding experience was really, really crap.
So it was very, very easy to beat. Um, but uh, definitely see sort of like. [00:27:00] The huge amount of value in focusing on onboarding because of the compounding impact of the time. Even if you do have really decent retention today, like the way if you spend time and if it improving the onboarding experience, that means you're going to be keeping customers, uh, more customers for longer.
And that just keeps compounding over time. So, um, big, big area of focus, the thing then you mentioned now, so. We focused on onboarding. You also mentioned as well, like looking at the key actions that users are taking versus don't take, I think this one is acting quickly, like where you get a lot of gray area and it's quite difficult to understand.
So I'm interested, like what sort of levels of data does a company need to have for you to be able to get these clear signals? Because I think definitely early stage, we were talking about companies that don't have a lot of customer in the middle there. And even in our case, like we slowly. The volumes that we have, like, we can only really take the data at like a face [00:28:00] point and just use it for signals rather than this is reality.
Um, so in that case where you're trying to send these actions, like what is sort of like the size of scale that companies need to be at, where you can actually get some sort of significant signal statistical significance on.
[00:28:15] Claudio: But, uh, but, uh, that's a very good question. And I was very surprised. We, we introduced statistically significant, um, algorithms in all our reports by default.
So whenever we say that there is a, drop-off at an onboarding step, for example, we know it's statistically significant. It's not just, uh, by accident, uh, where for one day. So. And my belief was when I, when we built in a trance, was that, yeah, you need hundreds of signups to get fair. Um, I was very surprised to see, but not really, I would say over a hundred a month, at least, and depending a lot on what's happening, but big [00:29:00] drop-offs are, are statistically significant.
Much more numbers than bigger numbers. Yes. A difference of 5% needs big numbers, but a difference of 20%, it needs much more than numbers. And, um, yeah, I would say a minimum of a hundred signups when you need to start looking at a kit, but for the difference of actions between onboarding steps, Even a hundred signups is not enough, probably 300 accounts that get into that scenario in that get to that step where you have, uh, the most, uh, the biggest drop off, I would say from what I've seen, but it's hundreds, hundreds of signups.
[00:29:48] Andrew Michael: Yeah. And I think it's also what you like. It's a key, key points as well. Like free socks. We want to think about how do we measure the impact that we're having and how do we get like statistical significance is that [00:30:00] it's not. The volumes is not as important as how big of a lift up or down you're trying to measure.
Uh, so like you say, if it's like, uh, you're running something and you want to see a K it has this improved, you can tell with Lexis Cisco significance, if it's a major improvements, like if it's a 50% or 25% that like, if you have a, like, in our case, we, the experiment that we were running, we were trying to understand, okay.
Uh, how long do we need to run this test for, to get an uplift of say like 20 or 30%? And it was four weeks, um, that we needed to run the data in order to be able to understand that lift, but because it was so significant, we managed to get significance in like five days, because it was 180% uplift. And I think this is one of the things is like when we early stage, we want to be able to understand like how our metrics and the analytics is doing is like, Big changes.
You can measure it with the reliability. If it's like, you're trying to make this small nuance improvements. That's where you really need to have like data on your side and you have volumes to be [00:31:00] able to understand those changes.
[00:31:02] Claudio: Yeah. And the small businesses, I think the most important for them is to have a clear status what my own boarding grade.
Uh, what's my retention rate, the big, the big numbers and. You, you go through account by account, it's your numbers. That's what we actually like. We have integrations with hot jars and all the video record the session there and you can segment, okay. Who are the people that dropped off at the first step of the onboarding process, drill down to them, get very bitter professional recordings and watch them.
Um, basically we kind of, in those situations, we can only help. To pinpoint a specific group. It's not statistics. You don't have a problem of statistics. Anyone you have under a hundred signups a month, a start, you start to growing statistics becomes a problem. And that's where the algorithm, uh, Become very
[00:31:57] Andrew Michael: powerful for sure.
And I think [00:32:00] like what you're saying is that being able to combine like quantitative and qualitative together is really powerful. Like we similarly do a very similar thing as well with Hotjar now we actually have it, like now we've started with called popcorn sessions, which is something like. Companies actually do themselves where they get the team together.
Then they sit down, they watch recordings and like the amount of value you extract in an hour, just watching like, uh, quantitatively, like what people are going through and where, what they think. And sometimes we do sort of like, we see something in the data and then we'll say, okay, this pop constituent, let's try and see if we can answer these questions.
Um, so. Uh, and the specific, an early stage where we're at now, like this is a, it's a huge help having both signals. Uh, and then on top of that, just speaking to customers is probably the most powerful thing you can do. Um,
[00:32:48] Claudio: Yeah. Yeah. Yeah. But the data can guide you. What are the customers you should talk to and what you should ask them?
So that's why I think you should always meet all of them. [00:33:00] You should, should not only rely on data. You should not only rely on customer interviews. They should be mixed because they are just different sides of the same coin. I already say your role as a product manager or as a strategy is just like, uh, an army general.
You want to get as much Intel about the battlefield so they know what you are going to move in order to have a victory. Um, that's the purpose of data is the purpose of customer interviews to give you the Intel and not guess you do. Uh, they're still going to be a risk. There is always going to be risking everything you do, but that's calculated risk, which is much more different than simply.
Okay. What happens if we do that?
[00:33:44] Andrew Michael: Yep. And you're reduced the level of risk as well for every additional database, Jeff. So the more different tools and services you bring into your arsenal that like allow you to understand what's happening and the less risky you're leaving on the table as well. [00:34:00] I see, we are almost running up on time.
Uh, I want to make sure I have time for a couple of questions. Ask every guest. Let's imagine a hypothetical scenario you joined. I think I already know your answer, but you joined a new company. General attention is not being great. And the CEO comes to you and say Cleo, like we're really to turn things around.
We have 90 days, we need to do it fast. Um, you're in charge. What do you do? The catch. You're not going to tell me I'm going to look at the data and see the problem, or I'm going to speak to customers. You're just going to pick one thing that you believe, or you've seen to be effective in another company, and you're going to run with that playbook blindly.
So you're just going to say, okay, this is what I'm going to do to solve a general attention at this company. What would be the one thing that you would want to do sell produce?
[00:34:45] Claudio: It's a difficult question without looking at the data, but I'm going to answer it. Um, I'm going to talk to the, um, I would need to get this information.
Give, give me on the phone call all the customers that, [00:35:00] uh, Baden did not finish the onboarding for. Every single business has them. And, but the lowest hanging fruit, you have to improve retention companies that pay the account, your customers, but paid, but did not finish the onboarding process. You only need to help them to finish the onboarding process to stay longer for you.
Otherwise we churn the next month and removing that. It actually increases considerably your lifetime value of your customers because you remove, uh, the ones that only stay one or two miles, which affect your average is,
[00:35:37] Andrew Michael: uh, as I've had a feeling, you'd focus it somewhere on onboarding. I liked that they say.
Uh, focusing customers that have actually paid, but not established development because you do end up getting quite a few of those as well. They're like, ah, when I sign up, uh, like just pay with a credit card and come back to this at some point and then end up like three months later, we didn't even use this.
Like it's canceled. It's sort of a [00:36:00] nice, the next question is what's one thing that, you know today about general retention that you wish you knew when you got started with your career.
[00:36:10] Claudio: But you can't force it. Um, I used to think as a lot of people that I only need to find out what makes people, uh, retain and I'm going to shove it up to them and I'm going to have amazing retention.
Uh, it doesn't work like that. You don't force retention people. You don't like if you find out that they had seven friends in five days or whatever it is, You won't add them automatically and they will be retained. It's not. And I have an amazing story about this search engine optimization product. Find out that if people are five keywords for research, they're going to stick to the product.
And we say, okay, we'll add them automatically. And three months later, we do the analysis. That number of people to 10 [00:37:00] people needed to add five, 500 cents. It's not the number five is the experience to get there. You cannot force it. You need to improve your product? We enable it.
[00:37:10] Andrew Michael: Yeah, absolutely. I think that's also one of the dangerous sides when it comes to like metrics and analytics is when we see the specific numbers of specific actions, users need to take.
Like they, we can automatically start to think, okay, what are the behaviors we can engineer to get us there? Uh, Like the goal is not to get to the number. The goal is to get to the value. And I think that's also like something to be very careful of when you just get started with analytics and tracking is like not to obsess over the numbers themselves, but actually, what is the value that we're delivering and the numbers need to match the value, not the other way around.
Uh, so. Nice Gloria. Is there any sort of final thoughts say you want to leave the listeners with? It's been a pleasure chatting today. Like obviously all the things we've discussed will be in the show notes as always any final thoughts you want to leave your listeners with?
[00:37:59] Claudio: [00:38:00] Yeah. Um, yes. Always be. W what is question, um, all your decisions I do.
I'm a, I'm a data person, but I'm a true believer in gut-feeling and which is weird. You won't hear a lot of people in the data space. Talk about that. I would say, listen to your guts for the questions you should ask from your data, because that's where a lot of value comes in.
[00:38:31] Andrew Michael: Nice. I like that. Well, thanks very much for joining to plumbing has been a pleasure having you on the show and, uh, wish you best of luck moving forward.
[00:38:39] Claudio: Thank you. Great being here,
[00:38:42] Andrew Michael: And that's a wrap for the show today with me, Andrew, Michael, I really hope you enjoyed it. And you're able to pull out something valuable for your business to keep up to date with churn.fm and be [00:39:00] notified about new episodes. Blog posts and more subscribe to our mailing list by visiting churn.fm. Also, don't forget to subscribe to our show on iTunes, Google play, or wherever you listen to your podcasts.
If you have any feedback, good or bad, I would love to hear from you and you can provide your blend direct feedback by sending it to firstname.lastname@example.org, lastly, but most importantly, if you enjoyed this episode, please share it and leave a review. As it really helps get the word out and grow the community.
Thanks again for listening. See you again next week.
A new episode every week
We’ll send you one episode every Wednesday from a subscription economy pro with insights to help you grow.
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.