Do’s and don’ts for optimizing UA and performance ad campaigns
Welcome to Edition 31 of MAMA Boards, an AppsFlyer video project featuring leading mobile marketing experts on camera.
For today’s mini whiteboard master class, we have Andy Carvell, the co-founder at Phiture, a mobile growth consultancy that helps apps to grow, and focus on acquisition and user retention.
Andy shares how to improve retention using attribution data, and how to create a tailor-made on-boarding experience, achieve cross-team alignment and ultimately improve user loyalty.
Real experts, real growth. That’s our motto.
Enjoy!
Transcription
So what exactly is retention? It’s the percentage of users who are actually coming back and using your app again and again in a given time frame, or in a succession of time frames. And healthy retention is really the cornerstone of growth. If you don’t have that retention in place, that at least some percentage of the users are sticking around over the long term, you’re gonna find it really hard to grow your app sustainably over time.
One of the problems that I often see is what I call a silo mentality, so this is what I’m showing here, where we have, for example, a brand team, user acquisition team, and a product team who are not really aligned and not really talking to each other.
What can happen here is that you get a situation where you get what I call a disconnect, where the user expectation is not clearly aligned and you’re not delivering clearly on that user expectation through these different functions of your company.
So the brand is maybe setting a certain expectation, certain message, user acquisition is then advertising, trying to get installs, maybe at a cheap price, which is maybe optimizing for that price, and maybe adjusting the message in order to get that price. And then in the product, it’s also delivering a certain experience, which maybe is not aligned with the expectation the brand was setting and the expectation of the adverts and the app store itself. So if you don’t have these things in alignment, this is where you can really start to have problems with your retention.
Retention starts with acquisition
So retention starts with acquisition.
The user expectations are gonna be set from the very first touchpoint they have, if they see an advert for the app and it’s advertising certain features, the user’s gonna expect those features or benefits when they come into the app.
And if they’re not there, this is where you have a disconnect, and it’s gonna be very hard to retain those users. So it’s not really helpful for the user acquisition team to say, well, retention is a product’s problem. It’s actually everybody’s responsibility to have good retention. Here’s an example from a PDF app.
Let’s say you have an app which opens and allows you to view and maybe edit PDF files.
Hopefully, that’s made clear in the advertising, that you’re advertising these benefits. Hopefully, that’s clear from the app store presence, from the screenshots, from the description. And hopefully, that’s clear when, in the first use of the app, the user opens it up, and they can actually find that they can edit and open PDF files. And if the user opens the app, having installed something which they expect to open PDF files, and they find it can only open word documents. Then you’ve got this kind of disconnect where it’s gonna be very hard to retain those users. That’s a pretty extreme example, but my point is that you need to really carefully align the product experience with the expectations that you’re setting in the advertising and in the app store.
Measuring retention
So how do we measure retention? Well, the first thing to understand is that you need to be measuring every user session in order to actually do this kind of analysis. If you were using a commercial package like an attribution provider or an analytics solution, this is gonna be measured automatically by the SDK.
If you’re doing your own in-house analytics, you need to make sure that you’re recording every session that every user has, so that you have those timestamps to refer back to when doing this analysis.
Let’s look at some of the common visualizations for retention.
What we have here is a cohort table. A cohort, in this case, is a group of users who’ve had their first session on a given day. That’s what we see on the left there. And in the columns going through on the right, we see what percentage of users were retained on each of the subsequent periods.
In this case, it’s showing daily retention, so we see day zero, actually, 100% of users are always there on day zero, because that’s the day which they were first seen in the app.
Day one, it’s looking at what percentage of that cohort actually came back and were seen in the app, and had a session on day one.
Day two, et cetera. In this example, we’ve just used shading, often a cohort table has shading, but also a percentage number which shows the exact percentage of users who came back from that cohort. More intense shading means higher retention, less intense, lower retention.
We can read the cohort table two ways:
- One is scanning across from one particular line, which shows how well that particular cohort was retained.
- We can also scan down the columns.
- So if we look at, for example, day one retention, it’s usually a fairly critical number which companies want to improve, that’s showing the trend in day one retention and how it’s evolving over time. Ideally, it’s improving over time. Which would indicate an increasing level of product-market fit, which could be achieved by different marketing activities, which are bringing better users in, or improvements in the product, or some combination?
So let’s look at retention curves.
A retention curve is essentially a plot of a blended average retention for your entire user base over a particular period. What we see here is we have two examples.
- The blue line, showing very healthy retention. It levels off at a particular point, showing that, actually, a decent percentage of users are sticking around over the long term.
- The red line is actually more typical. It’s showing that actually it’s trending to zero over time, which means that, for any particular cohort, there’ll be basically none of that cohort left if you look at a long enough time horizon.
This is unsustainable growth, but it’s often masked by throwing lots of new users in so that the active user number never actually dwindles to zero. But if your retention curve looks like this one, the red line, eventually, if you stop acquisition, you’re not gonna have any users left.
Finally the stacked retention graphs.
This is essentially another way of looking at, and best visualizing, how monthly active users are affected by retention. So here on the y-axis, we’re actually looking at monthly active users. And what we see here is essentially a bunch of stacked retention curves, each of these representing one cohort. You can think about this very much like the curves that we see here above. Each of these cohorts is showing how it’s contributing to monthly active users.
The example on the left, these cohorts are trending to zero, so it’s kind of like the red line example above. And so it gets harder and harder to sustain monthly active user growth, and it starts to level off because these cohorts are dwindling to zero, it doesn’t matter really how many more you throw on top, it’s very hard to sustain the growth.
On the right, we have a much healthier example. These curves are flattening out, which means that some percentage of users are sticking around over the long term. So as you acquire more and more cohorts, this is slowly building and building and building on your monthly active user number.
How to influence and improve retention?
So let’s bring all this together.
- We’ve understood why retention is important, and some of the common pitfalls.
- We understand how to measure it, so how do we actually influence it and improve it.
Well, it’s a tricky topic, and it’s a big topic. Everything can improve retention, or it can influence retention, including external factors such as competition, as well as the marketing and user acquisition activities that you do, as well as anything that’s going on in the product. So there’s a lot of inputs to retention, and retention is an output metric.
Let’s focus on the actual input levers which I think are most easy to actually tackle. That would be new user activation, existing user engagement, and reactivation. So new user activation, this is how well you can actually get people to experience the value of the product in their first session, and get them activated such that they’re likely to actually come back and have more sessions within the app. So this is all about new users.
Existing user engagement
How well are you able to keep users engaged over the long term. This might be through releasing new content or providing other mechanisms that keep users engaged and they keep finding value, not just in the first session but time after time.
And finally, reactivation.
Are you able to bring users who’ve lapsed or churned back into the app, maybe with a special offer, or some new information which reinvigorates their desire to use the app and to come back again? And if you’re able to improve any one of these, retention’s likely to increase, albeit slowly, it’s a lagging indicator.
New user activation is usually where you can see the results quickest, and it’s usually where there’s the most upside, which is why a lot of people focus on that. So I would really advise working cross-functionally, both to get that alignment of your marketing message with your first-time user experience to improve your activation rate, and also to maybe consider doing what I call adaptive onboarding, which is to provide a slightly tailored or personalized onboarding experience based on the acquisition channel, or the acquisition campaign, that the user came in on.
Using your attribution data, you can actually identify which campaign a user came in with, maybe even what creative they clicked on, and then provide a very tailored onboarding experience that really speaks to that message that they saw in the advert that they clicked on. This is the ultimate in achieving that alignment.
So my final message really is that improving retention is tough, it takes a lot of work, but it’s totally worth it. And to work cross-functionally in order to achieve that. That’s it for today.
