How To Get ROI On Your App Marketing Budget

APS2013 logo - large
aps panel
The panel discussion titled “How to get ROI on your app marketing budget” took place at App Promotion Summit London and was chaired by Berenice Kalan, director of the Mobile Partnership Agency. Other participants who took part in this discussion were: Nick Marsh, VP EMEA Sales and Publisher Development at Mojiva, Noemi Mckee, Head of UK business development at Surikate, Cristina Constandache, SVP Sales and Marketing at MobPartner, Rube Huljev, Global Partnerships, Infobip.  The discussion covered a number of areas relating to mobile media buying and mobile advertising, focusing on where to spend your budget to get maximium impact.  With the breadth of experience on the panel there were quite a few useful insights coming out of the discussion.
We have the panel recorded in high quality audio and available below:

There’s also a transcription available here:
Berenice: So we’ve got panel speakers here now and we’re just going to choose each of them. They’re going to tell us what they do and they’re going to give you a tip on measuring ROI, basically. So the first person is Noemi.
Noemi:   Hi. I’m Noemi. I work for Surikate, a company specialized in the promotion of mobile applications. We work with agencies and app developers to provide them innovative solutions to help them market their clients’ applications. My tip for an ROI is to measure all third-party tracking, to make sure you track to install and also look at what your organic users are doing. Some people just look at the dates that they pull from their campaigns and forget to look at what their organic clients drove.
Berenice:  Great. Thank you. Cristina?
Cristina: Hi, I’m Cristina and I work for Mobpartner. We are probably one of the first pure performance mobile-only platforms, because we have started back in 2007. To this day, we work on a performance basis and we do it quite successfully in over 20 countries now.
So I think the best tip that I can give you is to look for long-term relationships. You’ve heard all morning how birth campaigns with no next steps will not get you very far, but I say take a step back and think about what you want to do with the partners that you are working with for the next two, three years, because it’s only like that that you can really get a true return on your investment in the long-term.
Berenice: Okay, thank you. Rube?
Rube: Hello. My name is Rube. I’m a global executive at Infobip, working with partnerships, and I’m here to evangelize a bit usage of definitely performance-based, which is the model which everything seems to move to. But then again, direct marketing tactics, using as well lower tech and not always insisting on the newest technology because they convert so very well. Definitely the use of messaging whatever it might be and direct talk to your customer and focusing on the copy.
Berenice: Okay. Thanks.
Nick: Hello, everyone. My name is Nick Marsh. I’m the VP of Sales and Publisher Development for Mojiva. We are a global mobile ad network and we also have a sister company called Mocean, who’s an ad server.
My tip of the day is all about optimization. So it’s not so difficult nowadays to actually get in stores, but how valuable are these in stores? Do they go on to be valuable users. My tip of the day is to track where you get your good users from and exclude your bad users.
Berenice:   Great. Thanks very much. Can we go to the questions?
Charlie:  I’m Charlie. I’m from SwiftKey. Our question was about the difference between advertising games and advertising applications that are trying to do something useful, rather than just entertaining. How different do you think your approach should be?
Cristina: This is actually a discussion we had yesterday in a more restrained circle. If you want to promote your app for something non-game, you should really look at why users want to use an app. For games, it’s pretty obvious, they are interested in games around play. They are hardcore gamers, that’s what they do. But for apps is all about offering the user something more than they can get on the web. Especially in the UK, everybody is talking about special offers available only for apps or discounts or any type of, let’s say, extended user engagement that they can only get with your brand through the app. That’s the main difference.
Nick: At Mojiva, we work on some gaming apps but we have a lot of other types of apps that we promote as well. I recently had the experience of promoting an email client app. We initially tried to exam, doing this across some niche channels where we thought it would work, such as news and business. But after a lot of experimentation, we found that we got the best installs, and actually, the best users, from Words With Friends, which, for me, didn’t quite make any sense. But the stats that came back from the client were incontrovertible, so the two aren’t that much so different.I think the key thing is as long as you have the tracking in place and you’re seeing what the people do after it downloads and also you have some set goals what you want to achieve, then the difference isn’t huge.
Rube: I would say for non-entertainment apps, for example, that connect to the offline world, like retail, hospitality or whatever, you should use all of your available channels, not just the online ones or the mobile ones. Try to connect them to online and mobile. For example, if you have a staff that is in contact with the customers, have them present them with the tools that can lift conversion. Then transfer it into installs and engagement. But definitely word of mouth, I would agree, and specifically if it’s digitized and direct and supposedly personal, which can be made, this converts . . . as my friend from my company has shown, this converts probably better than anything else.
Noemi: I just have one point. To reiterate what Nick was saying, it’s exactly that. You can be quite surprised about that your strongest conversions come through. At Surikate, we quite often advocate the 80/20 rule, which is invest 80 percent of your budget in sort of areas that you’ve tested and you know that you’re going to get good performances on, but always keep 20 percent of your budget to invest in new areas, test new formats, test new channels and see if they convert, because you could be surprised by the results.
Berenice: Great. Thank you.
Junior: My name is Junior. I’m the cofounder of a company called Bahcoin. What we do is we are a multi-sided mobile rewards network and we provide guaranteed retention for developers and indie developers. At the same time, we have a patent pending financial algorithms software for big brands and retail advertisers.
Berenice:  What’s your question?
Junior: Yeah, sorry. My question is what’s an ideal number of users that an indie developer should have before he or she starts putting advertising networks on top of the games and what’s a good fail rate?
Nick: Could I just clarify that? Are you saying that users so that you can actually put banner ads and show them advertising?
Junior: Yeah.
Nick: So, I don’t think there’s an exact answer to that. I think you do obviously need a reasonable amount because otherwise, it’s not going to be of interest for somebody to connect to you to [inaudible 07:17] hands. I would say you need in the realm of tens of millions of impressions a month to begin to make it relevant. Otherwise, it could be an awful lot of work and you just won’t make any money, make enough money to justify the work done.
Cristina: Yeah, as Nick said, there is no magic recipe or magic number. I think it’s just a matter of once you have acquired your first users — well, not your first two users — try to understand what works within the conversion rates and then go to the monetization partners that are going to be able to, first of all, provide you with a service and the type of campaigns that are going to help you grow together with them because I think all platforms, at the end of the day, appreciate publisher loyalty and if they can help you because you’re helping them, they’re going to be most willing to work with you and take it to the next step.
Noemi: From my experience, we actually from the word “go”, the minute our application was ready, within a couple of weeks, we started putting banner advertising in.
Berenice: So there’s no particular time, okay? Next question.
Cecile: Thank you. Hi, I’m Cecile. I’m an entrepreneur, so I’m actually here before I launch my product, or before it’s actually off the shelves. So it’s more or less a market study here today. I have a question that’s actually similar to this one. When it’s your first product, you’re actually reluctant to spend a lot of money, obviously, and the notion of return on investment actually makes sense when it’s your second or third or fourth product and you actually know what you’re talking about, even if the customer base is different. But at least you know what metrics you should go for.
So would you recommend any startup packages or would you recommend having marketing off the bat, of course, but if you have the luxury to start six months ahead of time, how do you want to ramp up your marketing spend in parallel to your development and testing so that it’s your first product but you’re still out there and you’re making the right choices?
Rube:  If it was my first product, I would try to get a lot of traffic to my website and my mobile website. Then from there, go for conversions, because it’s cheaper and it’s more effective and it’s completely measurable and you would get people then to really go through the copy, through the interest and you would get conversions of very relevant people, which are crucial in the beginning. So that’s what I would do and maybe put that widget on which converts easily the ones that are really interested to install.
Noemi: I’m actually going to answer your question but I’m going to make it very specific to what my company does. But one of the things that when we’ve worked with developers or people who have been launching their first applications, it’s really important for them to get feedback on their app. One of the things that we do provide is we’ve got a panel of testers, so [inaudible 10:10] the application, they’ll try it out and then they’ll give you feedback, and that’s going to help you go on for your next version, whether it’s different versions or a new products, but understand what your users like and what improvements they’d like to see in an application. Because, ultimately, yes, installs are great, but what you really want is retained users.
Cristina:  In addition to that, I think you want to really look at the step-by-step approach. So test in a controlled environment because you are in the beginnings of when you don’t have millions to spend on advertising. Analyze, this is the most important thing, analyze your results and then scale it up. You need an analyst probably before you need anything else to understand what works. In fact, user feedback and everything else should get you to the next stage.
Nick:  I think, to add to that, it comes down to having a very clear business model, some very clear goals as to how am I going to make money from these apps in the long-term, and then put the track in place so that you can test the user journey through the app or your mobile website or whatever you’re developing. So that when you are in a position to start testing, you can see what’s working and what’s not, where do people fall down in the process and should you maybe shorten something. So you can redesign, so you can start off slowly and experiment and then take some stuff that doesn’t work and then work on the stuff that does. So that at the same time, it’ll give you a better idea of what sort of partners will be the best ones for you.
Rube: One, I think, encouraging thing that we all heard today is that there is no special moment in the life of an app. It’s a continuous process. If you measure everything, if you analyze, and if you tweak based on those results, you will always find another good or great moment for you to ramp up your marketing in that.
Cristina: And keep an open mind. The things are changing so quickly, I don’t think you can afford to have preconceived ideas about what works and what doesn’t. I think that with everything else.
Nick: If it’s magical, it doesn’t work.
Noemi: I think, just from my perspective as well, do some market research yourself. Get friends, get family, see if you go out to a local school or college or university and get them to give you feedback. That’s your starting point for anyone who does marketing. That’s free to do.
And I think the next thing is there’s quite a lot of app stores and lots of ad network and stuff out there. You can use it yourself. You can start and try yourself, and it’s trial and error, with a low budget. Then the other one’s, obviously, will do the campaign for you with more investment. But you could that in app number two, app number three, etc. It is a learning process, though.
Woman: Hi. We’ve heard a lot of people talking about measuring conversion, measuring action, measuring engagement, and now we’re talking about return on advertising investment and I haven’t heard a lot about cost per action advertising. Do you guys see that as something coming that’s going to get bigger? Because some companies are trying to do it. Is it working on the publisher side, and is there enough volume for it to be interesting for advertisers?
Nick: I’ll answer that. By “cost per action”, do you mean a cost per action within the app? So somebody doing a purchase after installing it? I think the market, even if people are judging you on a cost per install, they’ll also be judging . . . let’s say I run a campaign [inaudible 13:40], lots of different apps, some people judge you not just on the install but on what happens afterwards. And quite a few people who, some feedback very quickly and say, “Yeah, I’m getting loads of installs but these are not valuable users because they are not going on to do things.”
So I think the market is if not moving specifically to that metric, it is very much demanding that type of, shall we say, that scrutiny. So they’ll want you to perform not only to the install, but to those following actions as well. So it’s definitely going that direction.
Cristina: And what was your second half of a question? Was publishers ready for it? I think they are, as long as you are able to provide them with a hybrid payout that makes sense. Because at the end of the day, we know very well that they are looking to eCPM. So if you’re going to come up with a formula that says I pay this much on the install because after that I can measure the second action, I know how much I should be getting out of it, as long as their eCPM is good, I think they are going to be willing to take risk in the long-term. It’s a matter of testing at the beginning but we’ve seen this shift on the web as well and it’s only logical that it’s going to happen on a mobile app.
Noemi: And also, there’s no kind of magic wand to assuring a cost per acquisition. There’s also a responsibility within the application itself, so even though you’ve got the user on board, to keep them loyal and keep using the application. So it’s also the responsibility of the app developer to animate its audience, whether that’s by pushing educations which are really targeted or proving regular updates to the content. You can’t just get someone on board and hope they’ll stay. There’s work to be done behind as well, which has to be remembered.
Cristina: So that’s when we’re going to see some very interesting transitions from the web publishers that they have mobile traffic, and we know that 30 percent of traffic’s coming from mobile, but that are not capable to animate the communities the same way they do on web. It’s when they have a second action that they are going to know what users want and then you’re going to see lots of special offers and a lot more engagement with users on mobile.
Berenice: Great. Thank you.
Man 1: Well, if an advertiser approaches you promoting his iOS and Android app, usually how the split in the budget looks like, allocating one part to iOS, the other to Android, maybe you can share some insights and some specifics to that question. And how did the picture change when he has a Windows phone app as well?
Cristina: I’ll start because we only do this everyday. I think it really depends on what you want and where your focus is. We recommend, and I think everybody would more or less say the same thing, we recommend that you test all of them and then once you have the first results in, you start analyzing the profitability and look what is immediate return on investment and long-term.
So while there is no, again, no formula, I can tell you 40 percent on iOS, 40 percent on Android and 20 percent on Windows, it really depends on what you want, it depends on the markets because all the markets are going to be different, and they’re going to be different between the different OSs. Then it depends on what the users tell you. So you have to analyze your first results and then go for it in the long-term.
Man 1: Maybe a recent example you can share some insights with?
Cristina: Game or non-game?
Man 1: Non-game.
Cristina: Non-game. Travel. Asia is starting to look like a very interesting market for travel because the last minute bookings are very, very successful. So there are lots of people investing in Asia. It’s quite challenging for non-Asian players because the traffic is a lot more expansive, but when you find a partner that has a good entrance into Asia, I think it’s worth definitely testing Asia.
Then in terms of iOS or Android, probably iOS to begin with, but definitely allocate some budget on Android as well. Then you start talking about US, you want to go for massive iOS but you want to start testing properly at the lower level on Android as well because the behavior is different. US, it’s all about volumes. Then you have to measure the user behavior.
In Western Europe, I think equal percentage, iOS and Android. And Windows is just emerging so I can’t tell you put 15 percent on Windows because [inaudible 18:04].
Nick: To give you an example, I was at a conference last year, one of the mobile master classes, and the mobile director for Trade and Media, which does all sorts of trade and all lots of other nice products, they’ve had a really strong movement to mobile. So they’ve got a tablet edition, both iPhone and sort of iOS and Android. He came out with a statement and he said the most valuable users are unquestionably on Apple in terms of people going on to do actions and buy. Now, you won’t all agree with that. I know it sounds a bit controversial but we’re here to be controversial.
So my statement would be essentially, I wouldn’t go 50-50. If [inaudible 18:44] you’re trying to achieve, it does seem clear that Apple users, for whatever reason, seems to have more spend propensity to spend more than an Android user. And I think going on to Windows, the big challenge with Windows, and not that they make good phones or bad phones, is that they don’t make enough of the phones. If there are not enough people using them, then any development efforts could be money down the drain.
I don’t like to knock Windows. I used to work for Microsoft. I think they came back in the market. But in the cold light of day, you have to decide how much budget you have and what money you have to spend. So my advice would be unless Windows gives you an awful lot of help and push to further the app, I would focus, put a lot of effort into Android but Apple is probably going to be the one where you get the most money back.
Cristina: Okay. Just to challenge me in combat, because he likes that. While you might not do a 50-50 split, you still have got what you have for that is the pricing, because the price is on Android and I also know the same. So you can go 50-50 but have different pricing.
Rube: And one thing is to definitely know your market. It’s depending very much on the regional market. Know your market, know your user. If you know those two, you can plan for other things. Yes, in Western Europe, real Western Europe iOS users are much more valuable. Elsewhere, you go to Middle East, I wouldn’t bet on it. You go even a bit more south, completely different situation. You go, for example, to one of the biggest growth markets like Brazil, and there you have a jungle. So I would concentrate on the user and then afterwards, measure on what this user is.
Blunt example, if you send them a link, you identify them on the spot when he clicks on this link, what exactly he is, where does he come from, which network is he on. So all these immediately can let you make a much more informed decision on what do you present to them, to this particular user. So go down to that user level. Measure some KPIs, always do B2B testing and present various calls to actions to various types of people.
For example, a mobile network, which mobile network user is on. Is it the pre-paid or post-paid users tells probably 50 percent about the user. Then of course, the device itself, the cost of the device tells maybe another 30 percent. Then you build a picture on the fly and present what you think in that moment is the best for this user. And measure, always measure. When you have your results after, I don’t know, a week or a month of testing, you know where to optimize and then you put your bigger money in it.
Berenice:   And just to be a little bit more controversial with it, so basically, from my point of view, I wouldn’t necessarily just put all my money on the app [inaudible 21:51], depends on the country, obviously. [inaudible 21:53] Apple users because they tend to be bigger spenders, so they have more money. But it also depends on category, for instance, SwiftKey, who asked a question earlier, it’s probably completely different to them than to game developers.
So from Nick’s perspective, I would say to make maybe, like Cristina said, 50-50. At least start with Apple, perhaps, and then move on. It depends on, again, which category you’re in, which categories you really want to invest in, because obviously an Apple app store, you do have less territories. Whereas on Google, you have a lot more territories, but they might not be bigger spenders. Anything to add, Nick?
Cristina:If we are really speaking about travel, you might want to test, actually, a mobile site as well, which might sound like a dirty word, but in Latin America, like Rube said, and even Southeast Asia, Japan, Korea and all that, mobile sites can give some very interesting results.
Rube: Especially in Brazil and South America.
Berenice: We’ve got time for one more question.
Nick: Which in the end, is the same.
Berenice: Time for one more question.
Chris:So it’s Chris from Bardow. Fun question, just before lunch. So we’ve got a non-game app and we started advertising a couple of months ago. I tend to listen to developers more than marketers, and what they tell me is, “Spend it all on Facebook. The Facebook ad for apps is nailing it. Don’t spend it anywhere else.” So what would the panel suggest to counter that?
Berenice: Firstly, could I get your business card after this so we can continue this conversation? Yeah, definitely test Facebook. I think, I can’t remember who said it, I think it was Ben who said it during his presentation, but Facebook is new, it’s working really well at the moment. We’ve tested it through a gate; the results are really good, but that doesn’t mean in six months time that that’s still going to be the case.
And one point that we do keep coming back to, I guess, it depends on your total audience as well and it will depend on your application. So, yes, definitely try it out. It is a good channel and it works well, but you need to see if it works for you and you also need to bear in mind, its landscape changes very quickly so in six months time, it might not be the case.
Nick: I think a follow-up on that is it’s definitely worth testing Facebook but some of the messages I’ve heard earlier on is, “Don’t put all your eggs in one basket.” It’s worthwhile testing lots of different channels. Some of the channels won’t work, but some of the channels will.
One key issue might be is if Facebook suddenly stops working for you, you’re right back to square one. So it’s always worthwhile experimenting on the other avenues to see what different type of user you pick up as well, because if you use just one channel, then you could be missing out on a load of other people who could be very valuable.
Cristina: No, you have to test several sources of traffic. You can’t rely on only one. They are good. We had lots and lots of positive feedback, but you can’t afford to start with one channel and only focus on that one because you’re never going to be able to compare how users behave in the long-term and how you can optimize a campaign afterwards. It means if you one day decide, or after the initial phase decide to go away from Facebook and test other things, just starting from scratch, other money, other budgets invested, other tests, you have to do things from the beginning.
Rube: Definitely, definitely to spread your marketing mix because if I was starting right now and had, I don’t know, €10,000, maybe €50,000 of marketing budget down the line in two months, I would probably make the spread across all of this panel and see what works best. No, really I would.
For example, our marketing niche is 15 percent of UK’s mobile marketing budget. Nobody ever heard of it. But people are putting their money in it, so I lean towards performance. Definitely give the best reach. You do give results that might not maybe give the concrete user, but they give a whole lot of users. So you need all of this, because a travel app, an entertainment app, their users will just be completely different when it comes to what works, what ticks.
Noemi: Right. I think from my perspective, what I would say is just be careful, because what you’re going to do is you’re going to have a whole lot of people go there, not just yourselves and developers, you’re going to have lots of brands, it’s going to drive up the price, you’re going to pay an absolute fortune, where you could spend your money better elsewhere. So great at the moment, try it out, we do say that, but do test other roads because you’ll find that sometimes you might get more for your money elsewhere, not just on Facebook. But like I said, the price will definitely rise in the next few weeks, so watch over that.
Berenice: Okay, great. Just want to say thank you to my panelists, Noemi, Cristina, Rube and Nick.
Thanks to everyone involved for a great panel and you can find more coverage of App Promotion Summit here