Published in: Expert Column
There is systemic install fraud in the app economy, according to business intelligence platform Adjust.
In a recent post, Adjust CTO Paul Muller revealed evidence suggesting that certain ad networks have been claiming “a good part” of genuine organic installs as their own.
Working with publishers who helped the networks to offer their inventory across the app economy, Muller identified certain providers as using click spam to run and support a form of fraud currently running within the mobile advertising space. Click spam benefits publishers and networks but can leave advertisers significantly out of pocket.
I agree that install fraud is a major issue in the mobile advertising space, but there is more depth to the matter. Although fraudsters do use attribution mechanics to “click spam,” many publishers have used those same techniques for legitimate purposes and for legitimate reasons.
It is only through understanding why publishers use these techniques that we can begin to see click spamming as a symptom of the wider problems with advertising attribution. And it is crucial to make the correct diagnosis early to ensure that we don’t mistreat problems.
Pre-cached clicks: Some publishers use click-spamming for fair reasons
Before I dive into explaining the wider issues, it’s worth taking a step back and examining what click-spamming means in practical terms. It has two fundamental incarnations:
The first is the problematic practice of using bots to generate fake device IDs and corresponding clicks. This is fraud, pure and simple, and should be stamped out.The second is used for legitimate reasons — pre-cached clicks.
If you’re not familiar with pre-cached clicks, the idea is fairly simple to grasp.
Instead of tracking a user after they’ve clicked an app install ad, a pre-cached click happens before a user hits the ad — similar to a view-through attribution approach. If that person doesn’t interact with the ad, then nothing happens. If they do interact, they get taken straight to the relevant app store and their pre-cached click is cashed in.
Pre-cached clicks have proven popular among many legitimate publishers at grassroots level. And the reason for this isn’t criminality; it is due to clear problems within the attribution space.
The root of this problem is click-thru attribution. This form of attribution measures and attributes the install after the event. So when someone hits a mobile ad, it very clearly tracks who the user is and where they came from, which is great, until you look at how it works practically.
Due to limitations in the attribution space, ad that use click-through attribution don’t take users directly to the store. They have to bounce a user into the mobile web to register the click and then redirect them to the store.
In talking to publishers my company works with that run launcher apps, we found that this inconvenience typically means that publishers lose 30 percent of the installs they generated. In some instances, these go down as organic installs and the publishers simply lose their money. But in the worst circumstances, publishers can actually lose that money to other networks. In particular, the largest networks with the most amount of inventory and the prime relationships with advertisers will often claim a user as their own further down the line due to their priority in attributing the installs.
Click-through attribution (especially fingerprinting) therefore costs publishers money, both through inconvenience and by accidentally attributing installs as organic or allowing influential publishers to capture the credit for the installs.
But a pre-cached click can help smaller publishers defend themselves and save their revenue. Returning to my sources, pre-caching clicks increases revenues for publishers by 20 percent compared to similar publishers (mostly launcher apps) that are not pre-caching clicks.
By using an attribution method that ensures an ad takes a user directly to the store, and by claiming clicks ahead of time, pre-caching helps publishers defend their position almost as effectively as click-through attribution unintentionally damages it.
Transparency: The true solution
For what it’s worth, I don’t actually believe pre-caching clicks is a good practice for publishers to get into. Nor do I dispute concerns that the use of pre-cached clicks, let alone techniques such as bots, can rapidly turn fraudulent.
But I do believe that by taking a step back to look at why publishers pre-cache clicks, we can reveal a more nuanced picture. Publishers are using this technique because it plugs a problem for them in the app economy that is caused, arguably inadvertently, by the ascendancy of publishers with the resources and clout to claim attribution.
They aren’t, in short, fraudsters. They are businesses that are negatively impacted by problems in attribution that are inconvenient to big businesses but which, at 30 percent of potential advertising revenues, could be harmful to a smaller company.
Therefore, we need to improve transparency in attribution across the board by:
Allowing grassroots publishers access to the attribution technology used by their larger peers (namely, view-through attribution and server-side click attribution),Making it easier to implement within their apps.
I, like all individuals working in the app economy, want to see the end of spammers, fraudsters, and criminals profiting illegally from our industry, but it is essential that, in our fight to stop them, we seek to fix the bigger problems.
This isn’t simply a battle against fraudulent installs in the end; it’s a battle against fraud across the whole ecosystem. So we must keep our eyes on the bigger picture to solve the industry’s problems.