We are down to the last 2 steps in the Saas Analytics Stacks. And today Ruben will be discussing the critical events and actions after a user successfully onboarded and how to use segments and amplitudes to track data for user retention and revenue reports.
SaaS Analytics Stacks Playlist https://www.youtube.com/watch?v=t8zAPTNlJMg&list=PLgr_8Hk8l4ZGDnOOa0biP1N7buN9BBts2
Amplitude Analytics: https://amplitude.com/
🎓 Learn more from Measureschool: https://measureschool.com/products
🚀Looking to kick-start your data journey? Hire us: https://measureschool.com/services/
📚 Recommended Measure Books: https://kit.com/Measureschool/recommended-measure-books
📷 Gear we used to produce this video: https://kit.com/Measureschool/measureschool-youtube-gear
All right, and welcome to our last video series, well our last major video. We will have a short conclusion video after this.
In the second video after the intro, I talked a little bit about the acquisition data in our here in our SaaS stack or SaaS analytics stack that we’re looking at here. Then I talk about onboarding in the last video. And then this video, I actually I’ll cover two things. I’ll cover the critical action section. This is typically where you find things like behavior analysis, retention, and then a little bit of the revenue, right?
SaaS is, of course, a special case when comes to revenue, just love very specific SaaS metrics, like monthly recurring revenue, annual recurring revenue, and so on. So we’ll talk a little about how to track those and what’s the best way of have a tool to the stack to look at those numbers.
Let’s start with critical actions.
And critical actions really is at this point, the user has onboarded, let’s say they onboarded successfully. And now we want to understand how they actually use the product, right? Are they using it in a regular basis, whether it’s daily, weekly, or monthly.
And are they retained. Are they coming back, or do they love the product, right.
And this can be a very broad area but we’ll focuses specifically on on critical actions. So some of the most popular things that users love about the product, from a product perspective, or from a product manager perspective.
This can be things like feature adoption and things like that it can influence your product roadmap.
And then retention which is our are they, you know, are you sustaining each of the product and of course, retention, very similar, but look at a few reports.
Now for events, it’s quite broad. So we have critical events and action. So this means that were some of the critical actions for your product, they need to track here.
In our example, reason of a job search site and we’re looking at employers specifically. So employers were posting jobs to hire people.
In their case, they’re critical actions are things like the job posted, the end of the job has successfully filled a job, and things like that. So there’s you know, there’s always a handful, maybe five or 10, critical actions that really explain or understand what the product is, and what users can do with the product.
Then for user attributes, we’re looking at some behavioral traits. And this will also be very specific to your product. For example, this could be things like, how many jobs has this company posted in total, right? How many jobs have been canceled, how many jobs have been successfully filled.
And then how many people have been hired, maybe how long the context of events, right?
These are traits that explain the behavior of a given user a company within the product, right?
From a quality perspective, you may also see something like NPS score, again, we’re looking at customer satisfaction, or how much people love the product.
An NPS score can be an easy way to get started in this kind of world.
From a technical perspective, once again, will keep building on this idea that we talked in last video about tracking data from the server side, likely be the best long term option for your company. But if you can’t, and if you’re in if you want to submit that’s a little bit quicker, a little bit faster, client-side can still work, of course.
In our stack will be the same as before, we’ll use segment and we’ll use amplitude to track our data.
Now, the first report we can look at here is an event segmentation report. And we’ll look at just the most popular events within this account. Right. And when we look here, we get to see events like you’re, you know, searching for candidates, they on board and so on. Right?
So we have a few different things, things that are most popular actions here. And we can filter things out, you know, we might not really care about the viewer loaded page.
We may care about a job application, or searching and and so on. Right? So this case give us a sense of where the most popular actions users are doing within the product in comparison to other actions, right? Are they searching for candidates as much as they’re posting jobs? Or are they successfully filling jobs as much as they’re posting them, right.
And here we are using little bit some of the filters available to us. And we’re looking at only employers, and specifically if person signed up. We know we can have employers who actually didn’t sign up at all right? We just started the funnel and it just never completed, right. That’s one event. And there’s a few different things you can do here right. We can do different ways of filtering and slicing the data. But we just really want to get a sense of where some of the popular actions within the product.
Then the second report, we have retention report. And this report, we can look at a corporate analysis, right. So we’ll start by saying show me all the new users.
And then show me the users that return and perform any active event. So an active event can vary for your product. But this is typically some kind of critical action, it may be posting the job, it might be filling a job, whatever that is. Viewing the page, firing the page view, those things typically don’t qualify here.
In this case, down here, we of course have different user types, we’re actually only looking at employers. We can also do it through the filter type, as we said before. Or we could do it like this, this you get roughly the same result. And here we see our retention curve, you might be a little more familiar with a more cohort analysis curve, right?
Where we have a numbers here, you know, how many users filter criteria on on this week? And then was the retention over time?
And same thing. Once you define the retention, and the frequency that is, you know, what makes the user retain, right?
And then how what was the frequency of users for your product.
That is should they be used on a weekly basis, monthly basis and so on. A typical example here is imagine something like a food ordering app, right, like skip the dishes, fedora, or whatever it would be available, wherever you’re watching this.
You know, for user to be retained for this kind of product would they have to order food once a week, once a month, once every two weeks, once a day, Right? You kind of figure out what is the expected usage for the majority of the users and you of course, have power users who are going to do them what’s more often, and maybe you should do it less. But this starts to play a little bit into , you know, this retention report.
One more thing I was going to show you here on the behavioral traits. So we have our default amplitude properties. But then we have things like this. So this is one of the behavioral traits that we might have, you know, this product, for example, has credits, there’s people buy credits to be able to post jobs. So you can have credits, you know how many jobs slots are remaining. So we have that limit on how many jobs they could post any given time, total job post, total jobs slots, things like that, right? So these are all behavioral traits that are very specific to this product. But from here, we can start to build more constant segmentation, right? Let’s see for users to post them more than 10 jobs, how’s the retention change compare to users who post less than two jobs.
Those are the kind of questions you can start to explore once you have the data in order.
Then our last step here in our SaaS Analytics stack is our revenue data. Right. And in here, we you know, for events, we’re really looking at some kind of subscription event, right?
So this is our user who starts a subscription. And of course, the subscription by renew on a monthly basis or annual basis or any other kind of frequency you might be looking for.
Typically, for that event, we’re looking at things like what’s the plan there on, what’s the amount, the frequency,
Sometimes it’s even helpful to track the old plan if they’re upgrading, right, or if they’re upgrading or downgrading. And I can give you a way to that track how many people are upgrading or downgrading, what’s the most typical upgrade, downgrade path and so on.
And for user attributes, then we’ll track also tricks around subscription. So these are things like what’s the plan there on, what’s the frequency, how much revenue have they had with the company, all these kind of things. So these are all very revenue specific trades. And subscription data by far and large should be tracked server side, right? We talked about here where you may have an option depending on what kind of capacity you have with your team plus description data ready for the best acquisition should be done server side. And this typically can be done by firing from an API. Sometimes you have integrations from popular payment processors but somebody should be aware of. In terms of a stack, we got segment, we got amplitude, and we have recurly here as an example. But there’s quite a few options here recurly charge me. We’re actually be looking at one called Data box. In essence, these are tools that help you process payments. And they also give you a lot of metrics out of the box. So there’s a lot of payment processors like recurly charge me and so on. But they’re designed for SAS companies.
So they’re able to create a lot of analytics dashboards out of the box for you and calculate logic for you. Typically the logic around when the user returns, when when the user renews, doesn’t renew, all this kind of things can be quite complex is still up in the beginning. And it’s really not needed because you can get them with a few clicks, right.
So our typical report will look something like this, right, we can see how many new subscriptions happen. And this month, how much new monthly recurring revenue, total subscriptions,
total monthly recurring revenue, why we get some comparisons right here, right compared to the previous month, break down by plan, want to grow by plan, and so on, right, we get some other metrics. So these are all things that you could go out and find the formula online for this and work them out and so on, and likely even recreate a lot of these metrics insights into like amplitude.
But it’s just in the beginning, you’re better off just taken some kind of pre-built report and get them done. And as I mentioned, you can you can build upon them.
In amplitude, what the you know, you can track the revenue data in amplitude and that really is just to be able to tie it into everything else.
So you have all your users who onboarded, all the users who are retained, can you then confirm their the revenue, Right? Can you say, you know, are most active users are also our users with the most revenue? Or maybe there’s maybe it’s not right. So in here, you’re taking the revenue data, the revenue events to analyze the product and product usage.
But when it comes to revenue metrics, like monthly recurring revenue, and so on very specific SaaS metrics, you you likely be better off looking at something like a custom dashboard from your payment processor.
That’s really it. We’ll make a short conclusion video to finalize everything. But that’s the final two steps in our SaaS analytics Stacks, our critical actions and our revenue.
Hey, Julian here from measureschool. If you want to follow along with the next video in this series, then we have it linked up right over there. And if you want to learn more about the data-driven way of digital marketing, then definitely subscribe to our channel right over there. Now, my name is Julian, see you in the next one.