This is a new season of App Growth Talks, series of interviews with ASO, user acquisition, mobile analytics and app growth experts. This episode’s special guest is Peter Fodor, a prominent expert on mobile marketing, user acquisition, strategy and growth. Peter is a founder & CEO at AppAgent – Strategic & Creative Mobile Marketing Agency based in Prague.
It’s the start of the year and I think everyone is looking for some tips on how to optimize app store creatives – maybe you’ve noticed that something works particularly well for specific app categories?
We work for lots of gaming companies, ecommerce, health and fitness, education verticals. But honestly, I don’t see any patterns on the category level. I see one pattern for years and that’s that companies are not very clear in how they communicate the value of their product, a unique selling proposition. And simply put, this is the reason why people should actually download your app or game. And that’s the way how to be also different on the market, how to be really unique. And therefore that gives people a reason to give it a try and download the app.
So the very first step I’m giving to anyone, my team and also the clients, is to define the selling proposition well, based on the context on the market, your competitors, your product, your future product roadmap, the motivators of your users, and build it from there.
And if I should highlight some of the other tips, some more practical ones (this is more high level), try to save some time by focusing on the first impression frame mostly, because that’s where most of the users make the decision. They are not scrolling that much as marketers think to the screenshots, not expanding description. So the first impression frame plays a huge role. And the third, fifth screenshot is something that you can keep for later, once you have a great first impression frame.
Also, what especially newcomers to the industry do, they try to say too much in this small, limited space in the store. So when you are speaking about the assets, for example, screenshots, there should be a single message per screenshot, just one thing and just highlighting the certain feature, a part of your app or what you want to communicate. [It should be] enlarged, highlighted, and really make it super simple.
And we have one kind of a test here at AppAgent with Richard, our Creative Director. And that’s when we create these assets and watch them on the device,
We want to make sure that if you are commuting in a tram or tube, you are standing and holding some handle, you are still able to understand what we are telling you in the screenshot.
You know, the device would be probably shaking. You wouldn’t enlarge anything, you wouldn’t just put it in front of your eyes. You will see it from a certain distance. And we have to make sure that this is still good enough to communicate why people should download the product.
A few more things, small changes, as we see it in A/B testing, typically don’t bring big improvements. If you hit a certain local maximum, you have to be bold. Your success ratio will decrease, but it’s the only way. How would you find something truly different that will really resonate with the audience if you just change, you know, some small detail? Typically, you will see a very small improvement if any.
Yes, a few additional thoughts on this, let’s speak about ASO first and then we can get to some broader topics.
The first one is: testing a video cover can also bring a very nice uplift. And I can see quite a lot of examples in the store where marketers don’t pay much attention to the frame that is selected as the cover on YouTube and then shown in the Google Play store. It’s in a prominent position. It’s actually the most visible and the biggest asset when you open the app store page. So that’s what I recommend to do. And technically on the App Store, if you want something more eye-catching, communicating the value of the product and what you can select in the video where you pick the poster frame, you can just ask your motion designers to put the still frame on the beginning of the video, and then you are good to go because that will be the poster, really the poster that communicates the product the most, and then the video starts to play.
Another idea that we have tested in the recent past is to get rid of one-liners – the headlines in screenshots – and just to show the game as it is. And you would be surprised how often this works nicely. And also, it’s a great way to force designers to actually design the efforts so they speak for themselves, without the help of the copy, the marketing slogans and so on.
So that’s another thing to test, because ultimately, especially if your game is globally popular, you will spend a ton of time localizing the assets. And by measuring if it has any impact, you can actually save dozens of hours of your production team if you skip this part of the production. And in case you see it actually works and your one-liners are play a role, then we also combine, let’s say, keyword optimization with conversion rate optimization by picking the keywords that people search, and we see in Apple Search Ads they convert well, and have good like lifetime value of the users, and we include them in the screenshots.
And it goes back to what I said before about this user journey. If someone searches for your game and looks, for example, for TBP fighter, and you see, this is a phrase that’s highly searched and converts right users, then it makes sense to include it in your very first screenshot or individual cover, so that people immediately can make sure that this page where they landed is actually what they looked for. So that’s how you can combine the data from Apple Search Ads with your marketing efforts.
Yeah, thank you, Peter. I think that’s really, really interesting for our readers and listeners. I’m personally surprised by this tip with the game screenshots, without any copy, without any captions and CTAs. But just like Peter said, you should first test it because, yes, it might work, but you want to make data-based decisions. But the tip is really great.
Even if you see the results that will be the same and there won’t be any uplift, it’s still worth pursuing this direction and just removing the captions because you will save the production time. So it doesn’t have to be better. It could be just the same. And already that’s a good reason to simplify the production process.
Is ASO able to mitigate the consequences of the IDFA deprecation?
It’s a hot topic and also a hot potato. You know how I feel about that, because there are so many question marks still left and the time is running by. So probably as many others are, we are still not 100 percent sure of what this will bring. And we are probably not the only ones and companies such as Facebook are even more worried, attribution providers as well.
Speaking of ASO, I don’t think that the iDFA depreciation will significantly change how it’s done today. The fact is that we will see as marketers less qualified traffic because the targeting will be limited. That’s for sure. So naturally, the conversion in the store when it comes to conversion optimization will be over. That’s also something that we can anticipate.
But until then, Google really improves the tools and the reporting with marketers at our disposal, then I don’t think that this change will actually influence the way how we work. It will be more difficult. There will be less clarity where users come from. And our approach at AppAgent is always to understand the major traffic source, the level of prequalification, how people actually learned about the app, what they know about it when they land in the store. So we can actually build a story on the existing foundation that the user already might have at that point. This will be, of course, more difficult.
But, you know, if we can’t even today design a specific store page for a Facebook campaign compared to TikTok or, I don’t know, an email campaign. And we still have very limited options on how to customize the experience for every user. So in these terms, I think that the workflow will more or less remain the same.
Do you run Apple Search Ads for your clients? What role does it play among other UA channels?
Of course, we utilize Apple Search Ads quite often and we have two typical use cases. The first one is that we use it for data collection that I’ve already mentioned. For example, when we are designing screenshots but more for defining the search optimization strategy, in order to understand what’s the behavior of users, which keywords are worth targeting, how we can combine our paid efforts with the organic optimization.
Secondly, of course, for user acquisition itself. And it pretty much differs client by client because, you know, there are certain products that are strongly benefiting from the search traffic. Now in January, health and fitness, every app that’s about becoming more lean, fit is actually now searched a lot. So there Apple Search Ads plays a huge role compared to, let’s say, a casual match game, which is typically discovered by exploring the store.
What we see is that the bigger the brand, the larger the share of brand-related keywords, which is not surprising so often the strategy about defending the brand works for this type of clients. So you are not outbid by your competitors who can steal the traffic. For small apps or non-existing brands, it’s quite difficult to scale Apple Search Ads. And simply the search volume is very dependent on the number of search queries and also the number of advertisers fighting in this limited pool. You can’t scale it as you can, for example, Facebook. So it’s almost always part of our media mix, but almost never it’s the key channel or among the top two where typically Google and Facebook are.
Thank you, Peter. And I just wanted to add for smaller apps, they could just apply the strategy of nipping off some traffic, from the bigger brands, from more prominent apps on the App Store, because Apple Search Ads is actually the legal way to run campaigns for your competitors brand names. This might actually work.
I would recommend to definitely first use attribution tool, to connect it together with Apple Search Ads and really check if the traffic you are getting from the competitor brand terms is of high quality. We have seen a few times in the past that we are getting quite a big volume on competitor keywords and not always this brings users that are spending in the app. So it pretty much depends, you know, what the attitude of the user to the brand they are searching is, how much they are willing to change their preferences – and that also influences if the strategy going after competitor names will work for you.
What UA channels will gain popularity due to the IDFA deprecation or maybe for some other reasons this year?
I believe that Facebook and Google will be still dominant, at their advantage when it comes to data they are collecting and the profile they build of us, users, is enormous. But their share might be smaller. Eric Seufert has just predicted that Facebook will see a 7% drop in revenues in the upcoming months, which will probably be the case.
I think that the whole situation on the market will open space for more experimentation with new channels and channels that are actually more reflecting certain interests of users.
And I think of Taboola or Outbrain as content ads. Some audience prefers, let’s say, Wall Street Journal, where Taboola ads are shown. Some – the Economist, some, I don’t know, sports news. Thus you can somehow tap into people of certain interests. The same could be Spotify ads: you know, if you look at their library of podcasts, just the fact that I’m listening to, for example, Reid Hoffman’s podcast defines it that I’m probably into business topics and entrepreneurship. So that’s the way how I can be prequalified by some B2B company out there.
Podcast sponsoring could be another option. We will see how this will be effective but I think there will be more experimentation for sure and probably also more brand-building campaigns, because this will actually be more about the traditional world of advertisement where the specific targeting isn’t that good as in the digital space where we are born. And also the ratio of brand awareness type of campaigns is higher because it’s typically a larger mix of tools these companies are using and the journey is more sophisticated.
We’ve discussed games, so let’s speak about the case from a non-gaming category – Travel actually, which is now a bit funny in the current circumstances. But I will speak about Kiwi.com, a company that’s basically a digital travel agent.
We worked with Kivi for three years before the coronavirus hit. And I like to give this example because it was a really holistic type of cooperation: we were responsible for the mobile marketing, acquisition strategy, data analytics, user acquisition, creatives production and also app store optimization. So the full mix was on AppAgent and we had only this ROI target and growth target. The rest was up to us. And because the ROI is actually a compound metric of ad performance, store conversion and purchases, it’s something that truly reflects the need to have all the pieces of the puzzle in place. And we were able to double the ROI in the year three on Facebook and Google Ads and also grow the volume by 70 percent year over year.
So that’s a tremendous success. And just to give you an idea how we got there, because that’s is quite an interesting journey, we actually switched, fully switched to Facebook dynamic ads, which allow you to use custom design templates and connect the product catalogue, which in this case was a list of flights, routes as they call it. And Facebook actually combines the data you provide them, imports the data into a template and then shows the ad to the relevant audience. So that was one step that changed the performance very positively.
The second was to find the right app event optimization, something that helps us to tell Facebook and Google that this is the action that a user should perform that to be considered as a good proxy for the future purchase. That could be, for example, finishing Level 3 in a match 3 game entering a clan in Clash of Clans – anything that data scientists at AppAgent are able to identify as a good proxy. We tested about 20 of those and found one that perfectly correlated with the future purchase and happened already in the first session, and this signal for Facebook and Google was how to improve the performance.
The third major change was switching from regional targeting to global targeting where we basically told the ad networks, mostly Facebook and Google, please find us anyone around the globe who actually you consider as high-value user. And Facebook because of this enormous database that they are having performs much better when you give it flexibility to find people all around the globe compared to telling them “find people in France / Sweden / Poland” or even if you target cities. And people who are into more local business, they probably know it, but the others might not: the more narrow you go, the more expensive it is. So if you want to acquire users in Hamburg or Berlin, then go after Germany or Europe as a whole.
Which app and game categories are on the rise today? Which ones are recovering from the pandemic? What are the most promising app and game categories for 2021?
In gaming, I see that the puzzle category is probably stronger than ever, and we see it also on our client roster where we have quite a lot of new games that are falling into this category. And also, what we can see as a trend that will only grow this year is something that Michail “Mishka” Katkoff calls “hybrid casual” – I don’t know if it’s his term that he announced in his article on Deconstructor of Fun or it’s a really category defining term for everyone in the gaming space. But it’s actually where the boundaries between the hyper casual and traditional casual games are becoming less distinct. And it makes a lot of sense that publishers of simple hyper casual games are adding more content, and features, and richness to user experience, and it’s moving a bit away from the super-simple hyper casual approach to something more complicated where users can be retained longer, therefore you can monetize them better. So this is another category that I expect to grow even more.
And, obviously, mental health and fitness is the category that will still be very strong. I’m a bit doubtful about travel recovering well this year because as far as we see it’s still very difficult to travel. And the same goes for dating, at least here in Prague with quite some restrictions and the same goes for the cities where our clients are based as far as I know.
That’s a difficult question, let me think. For sure, CPMs will decrease due to the IDFA deprecation, and that will help mass-market products, such as ecommerce, to grow faster. There will be more space for them to acquire users for less.
Ad creatives will become even more important that they are today, also because their role will be larger, a tool which needs to prequalify traffic to certain extent, which is now done by targeting algorithms of ad networks. I’d also mention the brand building, I think this will also be a bigger topic, and I already see some companies going in this direction – they are actually building a lot of brands to go beyond performance marketing and create something that can last more than you’re running direct performance campaigns on Facebook and Google.
Apps monetizing through subscriptions and in-apps only I think will have hard times to sustain the profitable growth. And those that are actually monetizing outside of the stores will have more opportunities because their LTV will be higher.
And generally there will be consolidation on the market, that’s for sure. Small studios will be acquired by larger ones, this big players can then cross-promote users that will show some strong KPIs. These are a few of my thoughts on what comes in 2021 and let’s see how accurate my predictions are.