This is the third episode of App Growth Talks, a series of interviews with ASO, user acquisition, mobile analytics and app growth experts. Today’s special guest is Johannes von Cramon, a prominent expert on mobile marketing, Apple Search Ads, app store optimization, mobile analytics and retention. Johannes is also a co-founder of Growfirst, an app marketing agency based in München .
To be honest, founding an agency wasn’t my idea at all. Reckoning 2015, I worked in the marketing department of an app development agency called NEXT Munich, while I was still studying for my bachelor’s degree. And then my former boss asked me if I could imagine hire of everything I do for their clients into an independent agency. Although this sounded scary at first, I wanted to take the opportunity, because entrepreneurship was always more exciting for me than being just an employee.
Well, and then I needed a co-founder (because I am more a business- and data guy) that would be strong on the creative side, which is, as you might know, one of – if not the most important – discipline of app marketing these days. I knew Benjamin from university, he was already one of my best friends at the time, and a good counterpart, especially with his Photoshop and After Effect [Adobe After Effects] skills. We took it from there, and now we are here, five years later.
The first thing that comes to mind is performance marketing as a key success factor for app marketing nowadays. ASO is still important for sure, especially because it creates many synergies with paid advertising but you can’t grow without organic or with organics alone, which was different when we started. And due to Facebook and Google moving more and more away from major campaign management to the machine learning algorithms, conversion optimization is now the heart of almost everything that we do also inside the app stores.
And the second thing is that with ASO and performance marketing coming closer together,
Incrementality has finally become a trend within the mobile industry.
So you cannot just buy downloads anymore without questioning where the user came from and if he might have ended up in your app anyway, or what the organic uplift could be. This is really difficult to measure, and the duopoly is even working against it by obscuring placements and mixing organic & paid data. But every modern marketing manager should really think about incrementality.
Yes, mostly on Android with Google Play Experiments and also with SplitMetrics for some clients in the past. To be honest, we used to do more tests in the past, but nowadays we do a lot of creative testing in our campaigns and then adopt the winning concepts to the app stores in order to offer consistent messaging to the user.
But sure, A/B testing of screenshots is still a key role of those experiments. And maybe it’s even more important for videos. We create app store videos for almost every client because we can use many of the sequences and animations for our ad campaigns afterwards, but we noticed a long time ago that the app store videos are a little bit complex. They can be a huge advantage because of the autoplay and the awareness they create, but are highly dependent on the category of the app.
For example, in a very competitive environment with strong brands a video can make a big difference, while more innovative and yet not so well-known app concepts might be better off with static screenshots which convey the core value and messaging of the products a little quicker than videos can do.
Personally I think that proper localization is very important in terms of creatives and texts in order to drive better conversion and also produce more trust into a brand. In our case, with Growfirst, most of our clients are targeting the German market and a couple of English-speaking countries, so localization is not that of an issue beyond those two languages for us. But in the case of, for example, mobile games aimed at the international audience, localization is sure the key for success.
All kinds of ad managers: Facebook, Google, Apple, Pinterest; MMPs, like Adjust and AppsFlyer; Amplitude for behavioral analytics, and currently AppFollow as our ASO tool. And If a client is eager to do that, also A/B testing tools like SplitMetrics, of course. Because we get a lot more data from such tools than from app store consoles.
Generally, we try to work with a small group of providers, common-visioned, and where we’re able to dig into the product. This is because, since the tech stack is so important these days, we have to know exactly what they do under the hood and how they interpret the data. Also you might get better support as an agency… [Listen to the audio version of this Talk to learn how]. We are just choosing the best products and working on a daily basis with them.
The great days of ASO hacks are over, in my opinion. However, there might be an interesting one I’ve learned from a new client of ours recently. They uploaded two app previews to the iOS App Store: one landscape and one portrait, which led to the videos being shown on different places on the product page. This part was not entirely new for us. But the new mechanics is that the organic listing doesn’t show the first three default screenshots if they are already can be seen on the Apple Search Ads ad banner. This seems to result in having two videos looping one after another on the same page because on the organic listing you see the portrait video with two first screenshots and on the ad above search results – the landscape one.
We’re just digging in into what this means for TTR and other KPIs but it looks quite promising to have two videos in this place.
I guess the most overlooked ASO tactic is still review management, and it always has been, because it cannot be done without design & developments, and therefore requires much more time and energy to do.
But in nearly every ASO project I’ve been on in the last years, we achieved the highest impact in terms of keyword rankings and conversion rate by doing proper review management.
And what I mean by that is prompting happy users to leave a review when they achieved something inside the app, whatever that might be, and build a mobile focused support system alongside for negative or maybe just constructive feedback with in-app messages, chat bots, FAQs and all that kind of stuff. So that people know you care about support and there is not just an email where someone can send feedback, and they may never know if it even gets read. On the contrary, it’s very easy for users to stay in contact with a publisher. Also, as for the happy users who tend to not review an app at all because they are satisfied and don’t have any intent to do so – you will be able to push them to write reviews in the app stores. I think that’s really important.
Yes, definitely. Apple Search Ads is a little bit different than other paid marketing channels and since it is the only source of valid conversion rate, install rate & in-app event data on a keyword level we have right now, you can learn so much about your target group by analyzing MMP data, and use that for your ASO as well as for other marketing channels.
It is very valuable to know which keywords drive active users, retention and revenue and other indicators besides solely installs.
I can recommend everyone who has an iOS app to spend at least a little amount of money on Apple Search Ads to gain insights for generic terms, defend their brand and learn about their target audience.
And although Apple Search Ads can be a complex channel to manage if you are doing it at scale, it is a good channel to start with performance marketing. It feels a little bit like Google in the early days. I know you guys at SplitMetrics have written a lot of very good articles about the algorithm, campaign structures, bidding strategies and all that kind of stuff – and that might be a good point to start.
It depends on the category and on what your competitors are doing there because if you have higher bids, then you need to spend more money to get a decent volume of installs. But generally speaking, I would say $1.5-2K per month should do the trick to begin with.
The great advantage of Apple Search Ads is that you don’t need so much of a distinct volume that the algorithm really starts to work. You have this on Facebook with about 50 conversions per week and on Google with even fifty [conversions] per day, and that’s not that kind of problem with Apple Search Ads. So you can spend a little bit less [than with other UA channels] to get started and see what your bids are and what the outcome is through the MMP data.
Yes, we tried it when Apple released the feature at first, but had a little bit of a hard time interpreting the results of those tests. Phiture released a calculator some months ago, which helped us to do more statistical research.
But at the end of the day, Creative Sets suite wasn’t built to do A/B tests. And because you’re forced just to test one or maybe few keywords where you expect similar TTRs or CRs, as the algorithm otherwise will continue to show creative sets where it considers the best performance will be achieved, they will not be evenly distributed like you would want in a reliant test environment. And that’s a little bit of a problem.
However, I find it very useful for testing videos, since the video test you have on Google Play is available just on the product page and not on search because there don’t exist any graphical assets within Google Play. And you have this mechanic on the iOS App Store. So it’s very good to test your app preview as well against a screenshot set in order to see what kind of leverage you have inside search with app video ad. This can be huge due to the autoplay feature.
That’s a good question. We are a little bit early on, and I have very mixed feelings to be honest.
As an Apple user myself, I like that they are the only ones who take privacy seriously and act constantly to protect their users. But as a data-driven marketer, I think this is probably one of the worst things that can happen.
Without MMP data, we would be sent back to the dark ages of analytics. But at this point I’m quite confident that the MMPs will develop some kind of workaround. Probably with a new SKAdNetwork API, I don’t know. I’m not a fan of this death-of-something culture, if you know what I mean, because I heard this too often in the last years, and it was never true: for apps, and for some marketing channels and so on. But it’s very likely that our industry will change in the next few years due to this [IDFA removal]. And I think if Apple enforces this (and I don’t see any reasons why they shouldn’t), then Google will come under pressure and maybe have to do the same thing because users will demand it.
[Smiling] Yes and yes. I always grant those permissions because I know how hard this is not only for marketers, but also for product management, and development, and in the whole – to build a sustainable company. I am always giving the opt-in for the limit ad tracking as well. But I also know that only very few people think like that. I don’t expect opt-ins to be higher than 10-20%. Therefore, I think that a hash solution they’ve suggested recently could be a point in the right direction: just to relieve the IDFA on the device and create a hash upon that they can use for attribution of their service. But we have to wait how Apple’s answer will look like.
Of course this depends on the business model and the goals of our clients, but in most projects we’re optimizing for Return on Ad Spend (ROAS) and / or retention. In-app events are also important, especially in the upper funnel – something like registration or the first transaction, which can happen really fast after the acquisition – to get an early understanding of the campaigns performance because, if speaking of ROAS or retention, it can take really long time to get a proper understanding of the campaigns.
As you can imagine, as an agency we have to deliver results really fast. So we’re looking mostly at the in-app events, and in some cases we even look at the share-of-voice data on certain keywords we’re getting from our sales representatives from time to time. But those are companies that are not really into performance marketing but rather looking for increasing the visibility inside the store mostly against other strong competitors.
One step deeper I’m looking at tap-through rates, conversion rates and cost per taps to control spendings and know what your key competitors are doing: if they’re spending a lot on certain keywords or they don’t. And the last but not the least: limit ad tracking is always important to monitor. We try not exclude those users entirely since it really can hurt the reach of the campaign, but we split them into different app groups, and then monitor the performance of those separately.
Check the latest Apple Search Ads CPT, CPA, TTR, CR, as well as CPT & CPA trends for various countries and categories and Q/Q changes in the metrics in Apple Search Ads Benchmarks Report, Q2-Q4 2020.
They do like any other paid downloads. So even if you are running campaigns on Facebook, Google and other channels, you get a little boost on organic keywords, that’s for sure. But not on a keyword level. So I don’t have the experience when spending a lot of money on a particular keyword led to the organic uplift of this particular keyword, and various Apple reps also told us that this is not possible. But like with everything in mobile, a self-performed test is always better than the opinions of others, so I am curious to see other results people might have had.
Apple Search Ads was always the game changer if we discovered unexpected differences in user quality from specific keywords.
We had some clients who were sure that they knew what their most important keywords were in terms of retention and revenue, until we looked into the data after we had set up Apple Search Ads campaigns.
And we found that these keywords only generated downloads but not active revenue and users. I think in many categories there are prominent keywords which drive users to the product page and maybe also to the app but are a little bit too generic, so that you can’t build up a valuable user stream upon them.
And secondly, Apple Search Ads turned out to be a really scalable and profitable channel for some of our clients. Including those of them who had never known about this channel in the first place. This is especially true for very strong brands that can drain competitor traffic on a higher scale, because if you have a strong brand and people know your product, then you can build a very scalable competitor campaign with a decent amount of valuable traffic.
I think it’s quite clear that Apple will expand their inventory and ad formats, so maybe they will roll out ads in their News and Stock apps, and maybe also…
Listen to the audio version of this Talk to learn about another potential update.
Personally I would prefer further placements inside the App Store, like with UAC, to get with the first three tabs where featurings are now, since it’s very difficult for some apps to get featured because there’re so many submissions every day. And to get a little bit of a visibility in those first three tabs would be great. But not with UAC analytics and targeting options, we don’t need any of them [smiling]. So a display component on Apple Search Ads could do a lot for businesses that have a hard time scaling their campaigns with search only, because obviously, there are apps that are not that search-driven. This is important for Apple Search Ads right now. So this way would be good but depends on the placements and the value of traffic and analytics marketers can get out of this.
