The Story of Signal Zero: Appstore Domination
In the begining, we pwned the rankings
In late 2013 I decided to found my 3rd company. I think in many cases entrepreneurs feel the pressure to build world changing companies because of the start-up culture that exists, but these massive enterprises are often as much about time and place as they are about the starry-eyed dreamers that set themselves to the task of “disrupting”. Instead I wanted to focus on some basic business fundamentals and build revenue.
Now, if you have never bootstrapped a business before, I wouldn’t advise it. Its a bit like taking a few friends out in the middle of a forest with nothing but some undies and a knife and building a town out of dirt and rocks. It may appeal to your sense of toughness, grit or even your masculinity, but its just not wise. Naked and Afraid, but with stars in your eyes.
Signal Zero was a mobile focused company that by its end had developed and released a handful of games along with an advertising platform, 2 mediation platforms (offers and video ads), esports platform and data platform. It was initially conceived of as an answer to the reality of the mobile advertising market at the time (2013). Venture capitalists were pumping money into new mobile startups and those startups needed to show big growth and fast.
Enter TokenWall (Q4 2013 through Q2 2019)
Apple and Google had each arranged their (app) store fronts to be focused almost entirely on popularity. In grocery stores and the CPG industry, manufacturers can purchase better shelf space from retailers so that potential customers are more likely to see their product (and therefor make a purchase).
Ever wonder how songs make it on the weekly top 40? How about books on the NYT best seller’s list? More parallels. Companies spend in ad channels and / or they buy preferred space with retailers. There was nothing new here except for the market. Its never just about product efficacy, money is paid to shove things where you, the almighty consumer can see them.
Game and app developers in mobile were essentially given no way to purchase “shelf space” from Apple or Google, so they turned to the advertising ecosystem-at-large to solve the problem of visibility. Like any free market, the advertising eco-system answered! Behold, the “app-discovery” platforms offered mobile game and app makers a way to get their app discovered and to buy their way to the top of the charts.
These App Discovery platforms would incentivize users to install the app with some kind of loyalty style point. Signal Zero’s app discovery platform was named TokenWall, but there were a handful of others (Freemyapps, Feature Points and App Bounty to name a few). When the user had enough points, they could trade them in for something of value (usually gift cards or cash via Paypal etc). The approach was a simple but lucrative solution. Large game and app publishers in the space would calculate the number of installs they’d need in a 24, 48 or 72 hour window to hit the top of the charts and would pay these platforms to wrack up incentivized installs. In 2013, advertisers would often pay hundreds of thousands of dollars to propel their game or app to the top of the charts.
For publishers, the value was the visibility (or shelf space), which would in turn draw in loyal and dedicated users as a result of the visibility (or not). TokenWall installs cost much less on a unit basis than Facebook etc. For many publishers these TokenWall campaigns were the “litmus test” for the efficacy of their app or game. Once these burst campaigns ran their course, publishers would often allocate some type of “maintenance” budget to help hold their rank up in a respectable place (often the top 100 overall or the top 25 in a category). With 20 or 30 of these burst and maintenance style campaigns all happening at the same time, you can imagine how lucrative it might be.
When I started Signal Zero, I threw $15k into a bank account and only raised around $35k from friends and family. About 4 months in, we quickly raised another $115k because of the explosive growth. That cash along with a small team of founders working only for stock for 4 or 5 months propelled the company to nearly $7 figures per month in top line by its second quarter in business. When the company offered Bitcoin as a reward in addition to gift cards, the growth continued and along came Fox Business coverage, ABC News and more.
In February of 2015 the fun hit a significant speed bump. Many of TokenWall’s campaigns did not come directly from advertisers, but instead ran through large ad networks and exchanges in the space. One of these networks had around 350 employees at the time. I received a call from the company’s CEO informing me that they would not be paying for 6-figures (US Dollars) worth of installs from the past few months because it had all been determined to be fake or otherwise invalid. The CEO was sympathetic and apologetic, confessing that his network had recurring issues like these on a monthly basis with several of their partners. He reassured us he wanted to continue, but we had to come up with a solution to stop this type of behavior at this magnitude.
We’d detected this sort of behavior on TokenWall before and blocked it, but I was still learning about IVT detection. An investigation with cooperation from executive staff at this network revealed that TokenWall users had created invalid traffic. Since we provided our users cash equivalent incentives, it gave them a strong motive to cheat to try and get more, without providing the value to the advertiser. There are currently billions of dollars per year lost to IVT in the advertising eco-system.
Luckily, we had recently moved the entire operation into Pivotal Labs in San Francisco. Pivotal Labs is reputed as one of the best global technology mercenaries in the tech industry. Their achievements include ‘fixing’ the Twitter “fail whale” that used to plague that service along with working with other lucrative clients. In short, we managed to get the issue with invalid traffic under control by collaborating on a proprietary solution with Pivotal Labs and engaging with third party security groups like MaxMind, Distil Networks (pointed TokenWall’s DNS to Distil’s servers) and others. Here is what the proprietary solution looked like:
We’d quickly learned that issues with IVT aren’t so much a problem that can be solved, rather a battle to be constantly fought. There was always a monthly back-and-forth with our advertisers on suspicious traffic, but we managed to keep our chargebacks to less than 3% on average. These conversations were also vital to keeping our defenses up-to-date with new capabilities. Our issues with cheating users were kept under control with the help of our own hard work and our partners. At this point, TokenWall was known as one of 3 or 4 platforms that essentially controlled what was on the top charts in the Appstore. By its end, TokenWall would be responsible for delivering thousands of the industry’s largest publishers to the top of the charts. The next 18 months or so were smooth sailing but we eventually began to notice a lull in the campaigns we had coming in. Now it was Q2 2016.
Apple had never liked the practice of incentivizing installs because they wanted their popularity charts to be a result of organic activity, not the highest bidders in the advertising ecosystem. Because of this, TokenWall and many similar platforms were mobile websites. Apple would never approve such an app on their Appstore. Apple had for this and other reasons also been moving away from focusing on popularity. There was a gradual shift in the AppStores that took place from 2016–2018 that resulted in Apple focusing more on editorialization and featured games and apps selected by their staff. Today, the popularity section still exists, but its pushed “below-the-fold”. As a result of these changes along with the aforementioned venture funds pressing their investments for ROI and exits (rather than just growth and popularity), Advertiser demand for burst campaigns slowed and a refocus on ROI based campaigns via channels like Google and Facebook claimed more advertiser spend.
At this point, the business was a NET success. We never anticipated a large exit as we knew the EBITDA multiple would be low on this type of advertising revenue, especially with this high of IVT risk. As a result, we had arranged the business as an S Corp to focus instead on dividends. Because of this, the small investment been paid back + returns. The business itself had some carry from factoring receivables, but our chargebacks averaged less than 3% monthly for over ~18 months, so this carry didn’t pose a big risk. The company still had great top line revenue. It made sense to try and expand or pivot into a new business line. We put TokenWall on the back burner, let it run, but the company’s focus shifted in mid 2016.
Oops You Died (2016)
As we noticed the lull in advertiser demand start to set in mid-2016, we decided to revisit the idea of developing and publishing games for Android and iOS. I’d personally designed and developed several hit mobile and social games earlier in my career and had even become known for my design theory VDWM-T.
We’d been working on a game off-and-on already since 2015 to serve as an additional “unlock-app” for TokenWall. You see, TokenWall was a mobile website, so it was unable to read the user’s device information (IDFA, IDVA etc.). We required users to install a separate app on the AppStore so we could get their device id(s) which the advertiser’s required for attribution. We called these unlock apps because when the user installed it, it unlocked all the offers on the TokenWall mobile website.
Oops You Died was a simple casual game inspired primarily by the success of Crossy Road and a few other similar games. It featured a crazy little creature named Giblet whose futile existence was simple, avoid falling obstacles and survive as long as possible. In between matches, players could interact with a Goat to get rewards and power-ups. When the game launched, it found itself in the most popular games for over a month without the help of TokenWall or similar services. We’d spent around $20k on Facebook ads and it was featured by a handful of prominent influencers. The game was a hit, but after 45 days or so retention started to decay and the game lost its juice.
We were learning the hyper-casual phenomena was already taking shape in the mobile market with established incumbents. Ketchapp, Voodoo and others dominated the vertical and their success had much more to do with the efficacy of their buying teams on Facebook’s ad platform and much less to do with the investment in their game development. The rise of these companies over the previous year or so had made it much harder for success stories like Crossy Road to have so much staying power, because Ketchapp and Voodoo were throwing millions in ad dollars at the vertical to keep consumers engaged with their games. In this space, the games were made around the idea that their popularity would only last around 30–60 days. Instead, the focus was on spending millions per month with on Facebook and Google’s ad platforms while simultaneously churning out several low cost games per month.
We realized that to truly compete in this space, we’d need to raise (tens of) millions of dollars into a company that already had a long history as an advertising platform, not a game development studio. We’d need to produce 10 casual games a month to find 2–3 with the metrics to support the millions in ad spend each month on an ROI positive basis. We’d need to do all of that fighting existing incumbents in the space that had massive war chests. No thanks. In short, Oops You Died ended up being a 4–5 month company development diversion including the launch that resulted in a small hit and a lot of market insights.
We did spend a few months exploring a Tamagotchi-style pet game, but ultimately determined that the amount of ad dollars needed to make a single game succeed in mobile (in any vertical) created a barrier-to-entry that was just too high, especially given the incumbency. In AAA the production costs will kill you. The mobile industry deceives many entrepreneurs because the games can be so cheap to produce, but the millions needed to be spent (often monthly) on ads (not to mention the buying and analyst teams) to even keep your head above water. We didn’t want to churn through development on simple repetitive games on a weekly or monthly basis while doubting our ability to raise millions. We already had TokenWall as an annuity and wanted to focus on something a bit more interesting without so many major incumbents to battle.