Socking Simians

30 Sep 2016

For years, app developers and users alike have struggled with the App Store’s search. Quality apps have often been outranked by profitable garbage. Now, thanks to Search Ads, this highly visible shovelware can finally be outranked by the developers that truly matter: the highest bidders.

Online ads have long been dominated by bidding systems, which allow advertisers to dictate exactly how much they’re willing to pay for a user to click, tap, or otherwise palpate their ad. The key advantage to the auction approach is that it tends to show the most profitable ads. The key disadvantage, however, is that it tends to show the most profitable ads.

In many contexts, the most profitable ads are the flashiest, the most distracting, and the most deceptive. Whether it’s some guy bragging about his Lamborghini account or a banner ad imploring you sock a simian, there has always been a category of ad that tries to suck in eyeballs using a finely-tuned blend of weirdness and deceit.

Thankfully, App Store Search Ads are static and will likely be policed. As such, the ads we see on the App Store will typically be profitable not because they’re annoying, but because the apps they promote make a lot of money. More precisely, since App Store Search Ads charge the advertiser when a user taps on the ad, the ads that will win will be for apps that make a lot of money per App Store product page view.

Ad fuel to the fire

Thanks to Apple’s App Analytics, I know that about 10% of views for our WeddingDJ App Store page result in a purchase. I know that a lot of pages online convert at 1% or less, so this seems decent for a $8 app. We make $5.60 after Apple’s cut, so using an advanced field of math known as multiplication, we can determine that a WeddingDJ App Store view is worth no more than 56 cents. Worse, that metric is inflated by visitors that are explicitly looking to buy our app, so the value of a view coming from a wandering soul just browsing the store is perhaps 40 cents or less.

As such, when Search Ads launch on Wednesday, we’ll have an experimental ad running in the App Store for a few wedding-related keywords, with a maximum bid of only 30 cents per tap. Based on my previous experience with ads on Google, Facebook, and elsewhere, I expect two things to happen:

  1. Our bid will turn out to be too high to be profitable for us, and
  2. Our bid will turn out to be too low to be profitable for Apple

I’m always open to trying new things, but mathematics is not, and the math is seriously stacked against anybody trying to profit from ads at $8 per sale.

This episode is brought to you by profit

Think about some software ads you’ve seen or heard recently. Likely, those ads are for software that generates a lot of revenue per user. Likely, that software is priced with a subscription. Advertising studs like Squarespace, Backblaze, and Hover all drive subscriptions. Subscriptions are jet fuel for what spreadsheet addicts call Lifetime Value, or LTV.

Now, it’s intuitive that subscriptions increase the average revenue per user – that’s just common sense. But if your renewal rate is good, the effect exceeds our intuition. Let’s take Backblaze for example. This simple backup service is famously “only $5” a month. Assuming they have a 90% yearly renewal rate, one signup would bring in $600. If you can drive such a signup for only $100, you’re doing great – you’ve found a way to turn money into a bunch more money.

Meanwhile, the poor schmoes charging just $10 for an app can only justify ads if they can motivate a purchase for under $3 per sale. At today’s ad rates, you can barely drive an install of a free app for $3, let alone convince some cash-strapped millennial to cough up $10.

Of course, a naïve auction system would reduce the price of a keyword until it becomes profitable – at least once the dumb VC money takes its turn. This could mean that some App Store ads actually do run for 40 cents, 10 cents, or as cheap as they need to be for developers to make a profit. Unfortunately for those of us that sell products with low LTV, there is a threshold where it’s not worth it for the ad network to show your ad at all, and typically 40 cents won’t cut it.

Showing ads, you see, comes at the cost of slowly training viewers to tune them out. Given this, we see veterans like Google presenting no ads at all for most queries. If a search term doesn’t drive demand for highly priced legitimate ads, no ads will appear. Meanwhile, profitable keywords, which Google refers to as “highly commercial queries”, often have an entire screen full of ads before the first organic result.

Most search terms that indie app developers would be interested in will probably fall into the “not profitable” category. If Apple is willing to show our ads for a few cents each while the system learns, then there’s no harm in putting them up, but that train won’t last long.

In the long term, Search Ads are just yet another argument for charging serious, sustainable money for your software. Whether you’re looking at the new options for in-app subscriptions, providing a fully fledged cloud service, or taking Omni’s lead in making use of App Store changes that allow a form of trials and upgrade pricing, the first step is shipping something that makes serious revenue per user.

Then, and only then, will it be worth asking them to punch a monkey.

© Allen Pike. See also Twitter and Steamclock.