Paul: Welcome to The App Guy Podcast. I am your host, it's Paul Kemp. I go around the world interviewing guest startup founders. Today, I'm in Belarus (it goes to shows you just how global my show is now). My task here is to really help you out as an app entrepreneur. I find co-founders, founders, entrepreneurs, app-preneurs so you get to learn from the best inspirational interviews.
Today, I've got the co-founders of SplitMetrics. Max Kamenkov who works alongside Eugene Nevgen. Max and Eugene - welcome to The App Guy Podcast.
Max: Hi Paul, it's a pleasure to be your guest.
Paul: Thank you for coming on the show. Tell us about SplitMetrics.
Max: Well our service, SplitMetrics, helps app entrepreneurs to A/B test their AppStore pages. It helps them to:
figure out what variation
what idea of their graphical assets is the most converting one.
I bet your listeners know that advertising costs are getting higher and higher these days. It's getting very important to optimize the costs of user acquisition to get the most out of these ad campaigns. So we are trying to help people to get the idea of:
how their pages are converting
what to improve
how to improve
...and things like that.
Paul: Okay, so you do a lot of work with websites then, where people are going to the app page on the website and looking at split testing those to see which can convert more downloads.
Max: Yes, you got the idea right. So we create pages that look similar to the App Store or Google Play and analyze user behavior on the pages, tracking all their activity, analyzing them and showing the app entrepreneurs the breakdown of what the users do and how they convert. Using this information, they can make the right decisions.
Paul: Well this is a great episode already. I'm really excited about talking to you Max, because there're a lot of questions immediately coming to mind. So for example, in your experience - and if you need to confirm with Eugene who's near you - how many downloads tend to come from a website, compared to directly from the App Store. Do you know?
Max: It really depends on the quality of traffic that you drive to your website's landing page. We've seen that these numbers are not very different. Actually, no one knows the conversion for the App Store, because Apple is a kind of a black box now.
Paul: Yes, that's right. A black box that no one can open.
Max: Yes, so we are trying to get a brief sight on that kind of data. Different kinds of apps or games have different conversions. We've already tracked over 10one million users on our experiments. I noticed the following examples:
- mobile games have a conversion at about five or six percent
- dating apps have a conversion of about 70%
- travel apps may convert people at a range of 45%
.....and so on and so forth. It truly depends on the category of the product that you have in the App Store or on Google Play.
Paul: That's fascinating, so let me just try to understand the data that you've just given us there. So what you're saying is conversion - I'm guessing that's conversion from landing on a website to then clicking and downloading the app. Is that right?
Paul: And what you're saying... What a variety of information, 70% would convert when landing on a dating website and they want to download the app. So 70% of those people will actually go and download the app, whereas if it's travel it's less, it's 45%, and even lower for the other one you said. That's a wide variety, isn't it, depending on the category?
Max: Yes. Again, it really depends on the quality of traffic, but if you drive your target audience to the landing page... Well, naturally we call it here not the website, but the landing that looks similar to your App Store. So if you drive your target audience, it will act similar to those people that get to your App Store page, so yes, you will see the conversion like that.
Paul: Okay, I've just had an aha moment, which means that I've totally clicked on something. I'm guessing a lot of people listening to this are sending traffic to the Apple page, where you can download and it opens up iTunes, or to the Google Play page, where you get the install button. But I'm getting it's a lot better to send it to your own server, your own website, and then start split testing those to see how you can get more people to click that install button, or go to Google Play and click the download button.
Max: Yes, correct. Well, let me tell you how this system works, to avoid any misunderstanding. So basically with SplitMetrics you enter the URL of your AppStore or Google Play page, you create or upload different types of variations that you may want to A/B test... Say you have a bunch of icons and you are not sure about what to choose, you can create the landing pages that will look similar to your real AppStore page, but will have different icons. You get one link, you'll use this link for your advertising campaign. People that click that link will get to the web landing page that will look similar to the app page that you have in your AppStore or Google Play. They will click there, engage with the screenshots or whatever elements they want, we will track all this information and show you the breakdown of the data. When people click the Install or Get button on that web landing page, they go to the real app store, download that from there and then start using the app.
Paul: This is great. First of all, the testing of the icon... I'm already thinking actually for myself, testing the podcasting icon and getting people to the page and see which is better. But I can imagine that's really useful because it could be quite a disparity between different icons; one icon could lead to that conversion of 70%, another icon could lead to that conversion of 45%. I mean, this is a big difference.
Max: Yes, absolutely. That's basically the idea that stands behind this service. When we were working on a special project a few months ago, we got a promotional artwork request from Apple, and we were not sure what kind of artwork to choose and to upload, because every designer has his own vision, his own ideas, and the more people on the team you have, the more difficult it is to choose only one thing. So we were spending hours talking, discussing and decided to upload one variation that fits some extent to everyone's vision. But the reality was a bit different, and we got fewer installs than we expected. So then we decided with Eugene whether we have to find if there's any kind of tool that will help us to define the quality of our ideas, before uploading them to the store. Unfortunately then we didn't find anything and decided to build it ourselves.
Paul: That's what we want to hear, and that is the inspiring bit for any app entrepreneur who's listening to this right now: if there's something that is not being served out there, then go build it. Like you did.
Max: Yes, that's the only way.
Paul: Yes, and that's really where... Rather than trying to copy things that are already successful and that have big marketing budgets - that I fell into the trap of in the past - build something that's not out there. I think that might be the takeaway so far. Tell us about this Apple approach then. So Apple were promoting one of your apps, and they approached you and they wanted to feature you, is that right?
Max: Yes, absolutely. This app is basically a kind of an additional tool for the service that helps interpreters to sign their docs by a digital signature, so it's not like the mobile standalone product. But in any case, when we got that kind of request we were absolutely excited because you know, when big brother tells you, "Hey, get on board", you always get excited about that.
Paul: Normally it's all the other way. It's me constantly bombarding Apple, saying "Hey, can you promote this? Can you do this?" And they're coming to you, which is much better.
Max: Yes, so we were excited. But you know, after we spent many hours discussing what kind of idea to choose, we got absolutely exhausted and the version that we decided to upload was far away from the initial one. To be honest, we didn't decide to build this service from scratch, because before doing the first string of code we talked to our peer mobile gaming companies, friends or mobile marketers, and describe them this kind of idea; we got massive feedback, analyzed everything and it took about three weeks for us to get the idea and deliver it to all our target audiences, get feedback, analyze the results and start prototyping.
Paul: Right. So let me get this straight then, you guys are a team of app builders, app entrepreneurs yourselves, you're releasing apps into the Google Play and the Apple AppStore; but this is like a pivot for you, because at one point in time you had the idea, then Apple approached you, you got stressed out about what to present to Apple for the ultimate conversion, so you started to build your own tools. Then you started to push that out to your own community, and that's the birth of SplitMetrics.
Max: Yes, so we were doing the mobile app that I've already mentioned for service, although we weren't alone there; it's like a company that builds that project. Besides that, we had a mate in our team who has been doing user acquisition for mobile apps for about two years, and he probably got one of the first influencers for our service, delivering the most valuable feedback. It appeared that he had tried to build something like that, but using Excel spreadsheets and some other tools, on the knee, as we call it here.
Paul: Yes, that's just not going to do nowadays, is it? So hence the idea, and I'm guessing the most valuable is the split testing of various landing pages. Could you split test anything else other than the icons?
Max: Sure. Almost any point or element that you see on your AppStore or Google Play page, you can easily A/B test here. To name a few, they are video app previews, you can measure whether there is any impact on your conversion if you add or remove video app previews; you can A/B test different types of screenshots, the order of screenshots, you can change descriptions, you can even try to test the price of the app, placing different numbers on this buy button; you can A/B test and measure how people pressed Read [inaudible 00:15:20] and engaged with the description, and we currently finished developing the Apple Watch compatibility to give people the opportunity to A/B test whether they have to add Apple Watch support to their app.
Paul: Yes, you could almost add Download on Apple Watch, and not build the app, but just see how many people click that, then you know the demand.
Max: Then also you can decide whether you have to spend the time to develop that kind of integration. An additional thing is that you can also A/B test whether you have to translate the description of your product to a foreign language. I remember Matt was telling us that one day he translated the definition of his app to the Dutch language and saw no impact at all, just because people from the Netherlands speak English.
Paul: It's like their second language.
Max: Yes. Besides the Netherlands, there are a good number of countries that you can't be sure enough, so thus you can A/B test and see if there's an impact on conversion.
Paul: So Max, in all the episodes I've recorded, I feel I've got to ask you some generic questions I'm asked all the time, because these app entrepreneurs are constantly asking me these things, and maybe you can think about your own... We don't need to give specific numbers but give us your own examples, your own results maybe, from some of the campaigns that you've worked with. Do you get more clicks having a video preview, or just having the plain screenshots?
Max: Well if we speak about mobile games, then the answer is yes; a video preview is absolutely a good idea to add to your game because we noticed that gamers do like to engage with the gameplay of your product before installing these 600 megabytes files. So we saw that these video app previews increase the conversion greatly. If we speak about utility apps, which are basically non-gaming apps, it really depends on the category. We ran a few tests with travel apps and saw that there is practically no impact on conversion after they added these video previews.
Paul: Wow, so those depend on the category. Okay, so here's another one: should app entrepreneurs release on iOS or Android? And so I'm guessing if you must have a landing page with both icons, which one gets clicked more? The iOS or the Android?
Max: Well, to be honest, we don't have such information, because in order to measure that you will have to invite both iOS and Android users to that kind of landing page.
Paul: Well what about conversion rates then? How about if you do a campaign for an Apple AppStore, do you get more people clicking the Apple icon to download that app, as opposed to the Google Play store?
Max: I think that numbers are pretty similar, and the reason for that is again, the success of your experiment really depends on the quality of the users that you drive to your test. So basically if you have an Android app, you might be driving the Android users there, which are the target audience for you. That means that they will be more likely to click the download button.
Paul: Have you ever done an experiment with a paid app, where you are inviting someone to a landing page and you announce that it is a paid app, and then you look at the conversions and wondering if there's any difference between the Android users clicking that paid app and installing it, or the Apple users clicking? Because that's an urban myth, isn't it? We all often are told that Apple fans will more likely buy an app, compared to Android.
Max: Yes, so I can say that there is a difference, and the iOS users are more likely to play the button which a price on it, compared to Android users.
Commercial break [00:20:21] to [00:22:55]
Paul: Thank you. There you go. All those people that keep asking me that, there's your answer. If you're going to do a paid app, do an iOS app to start with, because you are more likely to get paid.
Max: Yes, well actually we've run a good number of tests and we keep running them, and sometimes we get absolutely amazing results. I can also share a very interesting point here. We were amazed to learn that only one percent of people click the "Read More" button and read the full description of your app. So it basically tells at least us, or some friends that develop apps that we know, that you basically there is not much need to pay a lot of attention or effort to create a very long description for your app, just because people don't really read it.
Paul: That is fascinating. So just to repeat what you've said, only one percent of the people actually click "Read More." So that is not the most important thing of what you do; the most important thing is the title, the description, and the icon.
Max: Yes, absolutely. Right.
Paul: Wow. And what are you seeing from the cost of acquisition then? You said it was going up. Have you got any guidance on acquiring new downloads, any good place you found to acquire new users?
Max: Yes, sure. I think that before doing any kind of user acquisition, you have to spend some time trying to figure out whether the screenshots or description that you have on your store are good enough, because even one percent change of the conversion may bring you a good number of new users, or may soak your budget. For example, we've been playing around with these statistics and calculated that, say you have a gaming app which has a five percent conversion, and you are able to increase this conversion up to seven percent. One day you want to run a user acquisition campaign, you buy one million clicks to your product, and by having the conversion at 7%, you basically save $70,000 per this user acquisition campaign, which means a good number. This is something that we want to help app entrepreneurs to do.
Paul: I see. So what you're saying is you're helping people save money because it's easier to increase conversions, rather than try to spend more money on getting more people to your landing page.
Max: Well yes, so basically you have to think about both sides. But if you have an optimized AppStore page, then it will be easier for you to run effective user acquisition campaigns.
Paul: Max, I'm loving this chat. I almost feel like we need to go deeper into the launch, and I'm tempted to invite you on to the AppStore launch stories podcast that I'm doing, which goes into more detail. But for now, that was excellent going through all that stuff. There's one more thing I'd like to do before we say goodbye to you, Max, it's the fact that we love to know about apps, we love to know what entrepreneurs have got on their phones. I'm wondering if you could give us one or two app recommendations from your phone; one or two that we may not have come across before. So do you have an app or two on your phone that you could tell us about?
Max: Well I have one of my favorite apps for traveling, which is MAPS.ME. Basically, it's a kind of an offline map of any country in the world, that you can download prior to traveling to that kind of place, and then use it there. Since from time to time I travel to different countries, like Russia, Poland, you don't always have internet connection there, so it's difficult to use Google Maps or any kind of online maps. So what I like about this kind of app is that I can download the app, it has all the points there, like subway stations, restaurants, bars, whatever you might want to find in your city, and I can use it on the go and get around the city easily.
Paul: Great, Max. I will make sure I put links to that on the show notes, so for you listening, just go to episode 299 and it will be Max Kamencov and Eugene Nevgen, and you'll see the links to the things that we're mentioning. Also, I guess I can put a link to some examples of the websites and landing pages that you've been talking about.
Max: Well, in terms of a good app combined with proper AppStore optimization, I could also tell you about WiFi Pro app. They have done a good job in terms of screenshots optimization. Besides that, the app is also kind of useful for me, because it's generally a database of Wi-Fi passwords that you may use while going through the city. If at some point in time you run out of your internet balance or your cell phone can't connect with 3G, using this app you can get connected to a public Wi-Fi or to a Wi-Fi with a password and use it.
Paul: Yes, I will make sure that we put a link to that so people can check it out, not just for the app but also for the landing page because you said it's a good example to see somebody who has improved conversion and what they've done to do that. So I'll make sure I put a link to that. This is great, honestly. What a great episode. How best can we reach out to either you, Max or to Eugene? What's the best way of getting in touch with you guys?
Max: Well you can reach out through the LinkedIn app, or Facebook, or just drop us a message on SplitMetrics.com website and we'll reach out.
Paul: Great. I'll make sure we put links to those as well. Thank you so much to Max and also to Eugene who's sitting there as well. I thoroughly enjoyed going through the chat with you. All the best with SplitMetrics, thanks very much for bringing the world and awesome service, and I really hope the best for you. I'm sure it's not going to be long until Yahoo! or Google are knocking on your door.
Max: Well, you'll be the first one to know.
Paul: You'll get another approach from Apple.
Max: You'll be the first one to know about that then.
Paul: We would love that. Okay, all the best.
Max: Thank you, Paul, thank you.