Don’t just read the reams of 2019 trends reports that call out in-app header bidding as the next “Big Thing” in mobile app monetization. “Get involved now, or risk leaving money on the table” is the top advice we get from Michael Brooks, SVP Revenue at WeatherBug, where he manages direct sales, programmatic partnerships, business development, growth, and more. Mike debunks industry myths and shares best practices, providing straight answers to tough questions based on his first-hand experiences with in-app header bidding at WeatherBug, the media property that connects 14 million consumers each month to accurate and innovative weather data.
Hello and welcome to Mobile Presence. I’m your host, Peggy Anne Salz, with Mobile Groove, where I plan, produce and promote content that allows my clients to reach performance goals and scale growth. And most of that growth is around apps because that’s what we talk about a lot and for good reason. The app economy, if you hadn’t noticed it, last year, 2018, the first decade of the app store. Here we are, 2019, and we’re looking ahead to a lot of trends in the mobile app economy, mobile app monetization, a big one, how are you going to make money in 2019 with your app. And to get the inside track on both, I’m bringing in a guest who has seen it from the start and is going to share a lot of these insights and I’m sure history repeats itself, so if you’re listening in, you’re going to learn a lot from Mike Brooks, he is Senior VP Revenue at WeatherBug. Mike, great to have you today.
Fantastic to be here, Peggy, thanks for having me.
I’m excited because you have been watching the app economy since 2013, not just watching it, actively in it, watching the evolution of it. I mean, you’re at WeatherBug right now but that isn’t what you’ve always been doing so give me an idea of what you have been doing in mobile.
Correct, correct. So, I started in 2012 actually as an analyst at AOL, so AOL post the advertising.com acquisition had a very strong maths-based team doing yield analysis and delivery at AOL and I had just gotten my Masters in Statistics, so it was a great transition into the analyst-type world. And 2012 really mobile was just beginning, there was one guy at AOL, his name was Chad Gallagher, and he took the role of managing all mobile non-reserve revenue into AOLad.com, at the time a very small business. Again, I was a young just right fresh out of grad school analyst, been there about six months when they had their fast ad download campaign.
So, I was at AOL but didn’t really work with this part of the business and it was a campaign for Priceline and this was prior to us knowing what a postback was, prior to us knowing what app install attribution was. As an organization, we kind of had no idea but who turned out to be my boss, Chad, had to run a campaign with really no understanding of it whatsoever and that’s a fair place to be, right?
2013, the world – and we’ll talk about that in a little bit – was much different than it was today. You know, what ended up happening was we ran the campaign, I wasn’t involved, it was just Chad, and it was horrific, we didn’t know how to track, we didn’t know how to target…
I was wondering what was going to happen, I thought…
It was a train wreck.
Was it pray and spray or something like that…
It was – I mean, I think we probably optimized to CTR – back in the day, everything was click-based, so you know, I think that we didn’t know what the windows would be, we didn’t have a way of ingesting postbacks, all these things that we know as kind of commonsense today from a media-buying perspective and an analysis perspective, didn’t exist. Certainly not in the mainstream.
So, AOL ran a campaign, it was a train wreck, I think the CPI was well into the thousands of dollars – every company has a few of those…
Kind of good for me because when that thing blew up, it became a very big priority to say “Okay, how do we play in this game?” and it was still early on, this was pre-Facebook coming heavy onto the scene, it was pre-Google, certainly well before UAC and all of the players in the media buying space were networks, it was all ad networks off of the 320 by 50 banner ad, video wasn’t around, retargeting wasn’t really around and it was early.
So, because that campaign bombed so poorly, performed so poorly, I got the job. It was like, “Hey, somebody figure out how this works” and that kind of became my job. So, we starting in February 0f 2013, launched the app download demand business at AOL which was the function of basically advertising.com mobile and it’s optimization algorithms as applied to the app download world. It was a wild time.
Indeed, I mean, you stayed in mobile to your credit, you could have said, “Uh oh, this is a wild west, I’ll go back to online where it’s safe”. Since then, you’ve moved on to – you went straight from AOL to where you are now or…?
Yes, well, it was interesting because over the course of my career at AOL, we were heavy in the MMA space at the time – AOL’s always heavy there but we had acquisitions of the advertising arm of Microsoft, we had acquisition of Millennial Media, we had acquisition of Adapt TV on the video side. So, all of these pieces kept pouring in and there was no me at the other – Millennial aside, there was no app download expertise at any of these companies but they were good assets, right? We’re talking about video, we’re talking about a much stronger mobile data presence and then you fast forward to Yahoo and Flurry is what Oath ended up turning into, or became…
So, I got to see the evolution while at AOL of what started as a core of advertising.com but then was supplemented by video via Adapt TV and then supplemented by the expertise of millennial and then the native piece coming over from Yahoo Gemini.
So, it was a very cool place to kind of watch the industry mature into what it is today, almost via acquisition at AOL. So, I was able to accumulate that knowledge from these companies as well as my basis at AOL and basically put together the best product that we could. And I think one of the most foundational points of that was the acquisition by Verizon to when we were acquired by Verizon, my team built – I mean, the pre-load business for Verizon was already built with Digital Turbine but we really brought that business to market in a big way.
So being able to use all those resources and kind of pick and choose how to build the best business was a really funky and cool opportunity and it was an awesome time.
You know, you mentioned pre-load, I have to smile here almost because I was just at an app event convention, conference, rather, and there was someone there saying “Hey, we have to take another look at pre-load as a way of getting to the user, keeping ad dollars under control because obviously if your app is pre-loaded on the device or the operator is pushing it for you, that’s part almost home-free for an app company”. So here we are, that was last year, 2018, and people were talking about that as a model and still has legs. What would you say are, you know, the biggest shift in monetization models? What, for you, works, or what do you think works best if you want to make money with your app?
Oh yes, I mean, on the monetisation side, it’s really about understanding the economic efficiencies you can make, so it’s really hard to say that you should be making money via a native or via search partnerships or via display or video because it depends on how people are interacting with your app, if we’re talking app-specifically here. So, I would say never you should be a native, you should monetize via native but something that I do want us to talk about later is the evolution of the maths that comes along with monetization especially in the mobile app publisher system. It mimics this evolution we’ve seen in the media buying ecosystem where worlds go from ad networks over to DSPs and they go from disparate partnerships to big, strong platforms.
So, I am a big fan of whatever it is you’re using to make money, just make sure, one, you’re not getting conned and, two, make sure that you’re always improving and trying to create optimal efficiencies. So, the biggest thing I could say for the back half of 2018 and all of 2019, the unified auction within mobile app is a huge, huge deal and something we can talk about a little bit in the future but to me that is the biggest way, the biggest change in the way that apps make money and one of the those things where – I’m only on the publisher side really about nine months now – was one of the biggest shocks to me – we did a report with InMobi and they found that 15% of partners work with one single ad network from a monetization perspective.
So, understanding the ecosystem and maximizing is really – as long as you’re pushing to make it better and like to use maths to do that, that’s what I think is important. I think if you go into it saying “Video is going to be big” or “AR is going to be big”, if it doesn’t really apply to your user base, it’ll never catch on, it’ll never make you money, it’ll never help any advertisers. So, it’s not about what you do, it’s how you do it.
I mean, I wasn’t aware at all of the InMobi report, I’ve seen others similar to say sort of along the lines that we’re decreasing the number of ad partners we work with simply because it’s just a lot of heavy lifting – for some companies, it’s heavy lifting to manage those partnerships – and you came out and found 15% use one partner?
It was insane, it was wild and that would be a single ad network or DSP that’s monetizing and what that is is a miseducation.
You’re short-changing yourself, basically.
For sure, exactly. The way that I – even if we talk about the waterfall which is something like 30 to 40% now, the analogy I like to say is like you go to a Sotheby’s auction and when the auctioneer starts, he doesn’t walk up to every chair and say “Hey, do you want to buy this painting?” and then if you say “No”, go to the next person. He says, “Who wants to buy this painting and for how much?” And that little vignette has been around in the desktop side with what people have been throwing around as header bidding in mobile for a long time.
So the desktop side has had this for three or four years and if you’re a desktop web publisher and you’re not using pre-bid or a header bidder, I think by this point you know that you’re very, very, very far behind. I think in the app world, there was a, you know, there was almost a misunderstanding – the technologies, different, we call it the same thing, unified auctions in mobile behave differently than they do on the desktop side but it’s one of those things that when I came onto the publisher space, that was the biggest shock to me. It was that “Why is everybody not adopting this yet?” and then once we adopted it, even in small chunks and we saw the value it provided, it’s even more befuddling.
Well, I’ll tell you what, Mike, for me, the biggest shock is that we have run out of time and we have to go to a break, that’s for me, but keep that thought because, listeners, you can see we are going to be talking about in-app header bidding which is going to be top of mind, top of your agenda in 2019, so don’t go away, we’ll be right back.
And we are back. Welcome back to Mobile Presence. I’m your host, Peggy Anne Salz with Mobile Groove and we have today my guest, Mike Brooks, Senior VP Revenue at WeatherBug. We haven’t talked about WeatherBug yet, Mike, but we will in a moment but you’ve really started into the topic that is just the buzzword – it’s in-app header bidding – it sounds really complex, it really isn’t. We’re going to talk about how you can grasp this opportunity. We know it’s an opportunity, you made that very clear, but give me some idea of like how or what do I need to be as an app company even to begin to want to get involved? I mean, maybe if I’m a very small company, I just leave it alone?
So, this is something that actually came up because I did – when we were looking at the report originally, there was this myth – we had people submit questions, advertisers and publishers, and there was this question around what was the scale required to be able to benefit from something like this which I didn’t understand was a mental challenge were taking on. It was one of those things that mathematically it is kind of always an improvement but people had this question of like “How big does your app have to be, how many impressions do you have to show, how do you have to monetize?”
And I guess the way I would answer that is if you monetize in a type of publisher unit, like we’re talking display, we’re talking video, we’re talking native, that is currently traded on the open market to put it almost crudely, you know, if we’re talking about something where there is an exchange, if we’re talking about a type of execution that is not custom and you’re making even a dollar a day, this is for you because now you can make $1.25.
I’ll kind of walk through how WeatherBug took on this challenge originally and to be honest, we did not go whole hog in the beginning, and I’ll explain a little bit why because we understood the maths but we dipped our toe. Philosophically though, this is for anybody who monetizes via mobile display, video or native because those are the three things we’re using it for today and what it is generally used for as well.
And to be very clear, this is about everything beyond Facebook and Google…
Correct, well, it actually includes Google too…
Well, yes, UAC, OK.
And Facebook audience network too.
Beyond Facebook because I’m hearing a lot of people say, you know, that’s just it, I’m not dissing anyone, I’m just saying what people tell me and they’re looking beyond Facebook because if everyone’s there, it’s like the blue ocean, the red ocean – if everyone’s there, you need to find a new place to go and this is where it’s at. So, tell me about that experiment at WeatherBug, Mike. You said you dipped your toe in it, you just said “This is it, we need to look elsewhere, lots of opportunity, let’s dive in” but maybe you learned something by diving in, toe as opposed to foot. Maybe you’re supposed to deep dive from the start…
Yes, I mean, so, look. If I was giving someone advice today, it would be to jump in head-first but only because we had the learnings that we did. I would say that this is really about the transition that a publisher makes between a waterfall monetization system and a unified option. So, that’s really the comparison that’s being made here because for the most part, I think publishers understand the waterfall, they understand their ad serving partnerships, their mediation.
That’s a part of the world people understand but I go back to the same analogy I brought up earlier about the Sotheby’s auction and the mechanics of how it works make it so much better. So when we were thinking about this originally, and I’m saying “We” – Ed Arundale on my team manages programmatic and he had pushed and brought a lot of this to market right when I had started – so I’m going to give credit where credit is due – and he put together this system where we wanted to test unified auctions because we had heard what I’m saying now, “They’re great, how could you not be doing them”.
That said, time is money in the publishing world, for every second that you’re down, for every second that you’re not optimizing your demand partners, you’re losing opportunity costs. So, that’s why people in general I think are concerned about changes. If we go down for a week, you know, that’s 2% of our yearly money likely. And that would be a very, very long time to go down but this runs through people’s minds. And we were in the same boat, a little bit lighter, but we said, “Yes, we want to try it”.
So what we did to start is we didn’t start whole hog and put everybody into a unified auction, we just added unified auctions into the waterfall. So a perfect set-up, a perfect set-up today is you have a unified auction, all of your demand partners bidding through that auction and they all compete on the same impression for who can generate the highest price.
You know, in the world of waterfalls, you have to program them in individually, say what you expect the CPM to be, constantly moving up, it’s a lot more work but that said, the hybrid model that we approached at first was saying “Okay, we’re going to take somebody out of the waterfall, probably a few people” – I think we were rolling 15-18 deep at the time and we were going to take on a handful and put them through one of these unified auctions and see what happens. We understood the maths but there’s another thing to have it be seen on the balance sheet, so we did it. So we did that two partners and generally, you want to work with one unified auction because they all operate very similarly in-app but we took those two header bidders more or less, put them into the waterfall to see what happens.
What happens when you apply these questions that you have in the back of your mind to, you know, real money and what we saw was exactly what we expected to see which was CPMs went way up, you know, within the first month those same partners that were already in the waterfall were monetizing somewhere between 15 and 21% better depending on the month in that first quarter. So, you know, that is significant upside but you are missing out on if you are not doing the full unified auction and remember that was just a small chunk.
So, we put together the experiment, we executed the experiment and the hypothesis that this was going to be productive to our business, just from a CPM perspective, was validated wholeheartedly. Granted, I think the real value to unified auctions are a few other things involved in that but from a CPM perspective, it was stellar. So, to me that proved the point and over the past nine months, we have migrated to a system where we can be whole hog with a unified auction. So we tested, we dipped our toe, and I want to be frank with people listening, that’s how we did it and the result we saw, 15-21% growth in CPMs, was good enough to justify full commitment.
But that’s not where it stops, right, the CPMs are awesome and I manage the revenue team so that’s how we’re compensated, we love to see CPMs increase, but it’s deeper than that because what was really happening was our latency was higher with this waterfall so what we saw was, at some point last year, I wasn’t here at the time, our latency was 7-10 seconds to show an ad because the waterfall was taking so long. So, what that can end up looking like is 5 to 10% of your sessions not seeing impressions and that’s not crazy, that’s the norm in a waterfall.
So I looked at that and I said this seems like if there’s a system that can make this move faster when the impression call goes out and people can respond quicker and in this model much faster, there’s a lot of money from a latency perspective that’s left on the table and that hypothesis was validated.
And then finally the icing on the cake was the conversations we were having with everyone on the monetization side was SDK-related. Oh, you need the SDK because you can’t be successful without it because we’re able to bid better because we have it in, SDKs can slow your app down. So, what the other fantastic bi-product of this is that we were able to limit our SDKs from 17 two years ago to 8 last year to 1 or 2 this year. So, all those things together are really why I’ve gotten on this stump lately of guys, like this is one of those things that isn’t too good to be true, it’s just a maths thing and if everyone in the world does it, it doesn’t impact me and everybody’s business can be successful.
So, it’s been something obviously you can probably hear in my voice that we’re very passionate about because it’s opportunity cost sitting at the table for most publishers, you’ve just got to at least dip your toe.
I mean, the SDK number, now that’s really something because that really is an issue and I know people with 20 plus of those and thinking about what to do about that. There’s a couple of other points you raised, we do have to go to a break for the moment but I’m really excited about this, Mike, because you point out also the latency which is another one I’m hearing a lot about, I’m talking to people who are saying, not even for the money, just for the latency, they are going to in-app header bidding and setting up a completely different system which we will get to – that’s the point. Now that we are all convinced of this, you know, our listeners, we’re going to come back and we’re going to hear how really to cash in on this opportunity so don’t go away, we’ll be right back.
And we’re back. Welcome back to Mobile Presence. I’m your host, Peggy Anne Salz with Mobile Groove and we have today Mike Brooks, Senior Vice President Revenue at WeatherBug. Mike, it’s been a great show with you because you are so – I’m enjoying it because you’re so pragmatic, you’re so straightforward, just tell it the way it is and you’re not like a vendor with a mission here, you know – it’s WeatherBug – you’re more like a convert with a mission. You have been converted into… And you make a great case, no question, which absolutely dovetails with everything I’m hearing out there but we have to take a look at it, step back for a moment and the myth I was hearing – 2014/2015 – so maybe things have changed and you’re a practitioner, I’m an analyst, there’s a difference, right? It was basically this is all great but kids don’t do it at home because you need to have a system and some tech that is, you know, off the charts, not for everybody, expensive, hard to maintain, an army of data scientists – what’s it like now?
Yes, you know what? When we were going through the process, we had heard those, a couple of those things but they were never that large in our head but now that we’re talking about it, I think I understand where it’s coming from because you can talk about misinformation being spread in the mobile economy. I think this might be a function of the way people are doing things on the desktop side, being applied into mobile, right?
So, if we think about what header bidding is where you’re really bidding into a header on a web page, that can really only exist in the web. And that’s how the world of, you know, header bidding came about. I think Google was one of the pioneers. The real pioneer or folks were targeting folks like Criteo who had been doing this type of thing for a while but I think where this is coming from is the process of setting up a pre-bid set-up on the desktop side for a header bidder.
So, when I’m thinking about some of the challenges, I think at the end of the day it’s the risk is what people are concerned about and we’ll talk about that, but when you set this up on a desktop side, I don’t want to say it’s easier but it’s easier and you don’t need an SDK to manage this auction, you can do it with what is really opensource with the pre-bid methodology.
So you can build your own header bidder pretty easily on the desktop side. You would have to devote a little, not a little bit but a decent amount of engineering resources on the web side and a little bit of data science and it’s a pain in the butt because people are used to kind of having things served to them from a monetization and partnership perspective and that’s fair – it’s a partnership eco-system.
That said, if you’re setting up pre-bid on the desktop side, you need all those things and maybe that’s where these ideas from people are coming from is that I just set this up on the desktop side, am I going to have to do this in mobile? I understand even less about mobile and then I have web to deal with and an app so how do I look at that?
So I think a lot of those ideas about the level of technology required, the level of investment might stem from the desktop set-up that exists today. You know, obviously you can work with a header bidder partner on a desktop side but if you have some engineers with some gumption, you can put it together and handle it yourself.
Mobile app, I wouldn’t suggest it, I think there’s a couple of really good partners out there, we did a very serious vetting process for this and there’s actually a lot of good partners out there – we didn’t run into anybody who we didn’t think would significantly help our business but I think at the end of the day, there’s that mindset from the desktop side about what it’s like to go through a pre-bid set-up. But then you think it really comes down to the money value of time. So, the cool thing about having an app business is you go to sleep, people continue to use it, you can continue to make money, you know, weekends, holidays.
What’s better than that, making money while you’re asleep? I don’t want to put you on the spot, Mike, but I will ask you because that will be the question, you know, vetting the partners, maybe we’ll have you back just to talk about picking the right partners but do you have a couple of names you could share without, you know, naming necessarily – not any one in particular but of everyone out there, what are some names we should be looking at?
People who I thought were exceptional, and there were a few – I think Max Ads was exceptional, acquired by Applovin. I think Oath has done an exceptional job. I think AirServe as acquired by InMobi does a fantastic job. I think the Amazon TAM product is a fantastic job. And remember, people who are listening could be working with these folks in a different capacity to begin with. We started by vetting 11 companies for this and we had some particular needs that led us in the direction and because it’s not public I won’t announce yet but we knew that this was going to be a slog of an RFP process, we put a lot of time into it. At the end of the day, I was surprised at how strong all of the players were that we met with.
So, we didn’t meet with anybody that I thought “never in a million years” which was nice because I haven’t had an RFP process like that literally every. So, that was nice where everyone has it together. But those are the few that I think are doing some exceptional things.
I think that’s really valuable, our listeners are going to thank you for that one and I have those always in show notes and everywhere. So, don’t worry, you didn’t have to jot that one down, we’ve got it for you, we’ve got you covered on this one. In the meantime, Mike, we have come to the end of the show, that is so sad but…
Indeed, because I want to have you back, we need to talk about WeatherBug, we need to talk about location and data and weather and that’s exactly what we’re going to do at a later date. So, listeners, Mike will be back but in the meantime Mike, how do they stay in touch with you?
Anybody please feel free to LinkedIn me or shoot me a note at firstname.lastname@example.org and let’s catch up in person.
Absolutely, and we have those as well for you later. And in the meantime, if you want to keep up with me throughout the week or find out more about how you can be a guest or sponsor on Mobile Presence, then you can email me, email@example.com, that’s also where you can find my portfolio of content marketing and app marketing services and loads of articles that I write about the mobile app economy.
You can also check out this and earlier episodes of our show by going to webmasterradio.fm or you can find our shows on iTunes, Stitcher, Spreaker, Spotify and iheartRadio simply by searching Mobile Presence. So friends, until next time remember, every minute is mobile, so make every minute count. We’ll see you soon.