What 3Q Learned About Mobile Fraud in 2018, Part 1: Click Fraud
Published: April 2, 2019
Author: Chris Bral
Thanks for joining us on our road to MAU Vegas 2019, where we’ll be presenting our proprietary mobile ad fraud-fighting process! In the weeks to come, we’ll detail war stories and tactics for sniffing out and snuffing out fraud.
Let me paint a picture for you about the mobile ad industry today: Digital will surpass traditional ad spend this year. Mobile ad spend is expected to reach $113B in 2020, nearly double that of TV. Finally, advertisers are expected to lose $19B to fraudsters in 2019, and as much as $44B by 2022.
Mobile advertising is only getting bigger. Fraudsters go where the money’s at. Never has the phrase “Mo’ money, mo’ problems” been truer. So, what does your fraud program look like?
(Image credit: eMarketer)
3Q Digital is in the trenches with some of Silicon Valley’s biggest companies. We manage a lot of traffic, and peeking into our MMP’s fraud consoles makes it clear our clients aren’t immune from contributing to that $19B statistic. With this knowledge, we’ve made it our mission to minimize ad fraud exposure, and this post outlines some of what we found (and how we dealt with it).
You Need the Right Tools
First thing’s first: You need network transparency and an MMP (mobile measurement partner) with a fraud suite. If you just go by a network’s dashboard metrics, you’re not seeing what’s happening under the hood. Some of the best campaigns (and bundle IDs) look amazing until you start looking at click-through rates, click-flooding statistics, and other click-based fraud metrics. None of this would have been possible without our MMP’s fraud suite.
Where We Found Click Fraud
One of the biggest red flags for click-based fraud in our campaigns was “duplicate clicks.” Boiled down, this is when a fraudulent app or bots fire off a bunch of clicks claiming they are unique and legitimate in an attempt to steal organic users. Not good.
(Image credit: Kochava)
Here are some signals we look at via our MMP:
- When an app has devices spamming a ton of clicks.
- When many clicks come from IP addresses associated with bots, VPNs, or hosting servers.
- When clicks are vastly out of proportion to installs.
One surprise last year was finding out our click trackers were hijacked. This is an especially sinister type of fraud because someone, somewhere took these trackers and was firing clicks nonstop. Networks (particularly ones that buy on exchanges) sometimes don’t even know this is happening because these clicks are bypassing their servers. When you’re dealing with tens of millions of clicks a day across multiple networks, these fake clicks blend in easily, and diagnosing the problem can be difficult since only you and the MMP are seeing it.
You Found Fraud – Now What?
Once we verify traffic is fraudulent, we do the following:
- Blacklist offending bundle IDs, device IDs, and IP addresses in our MMP.
- Notify the ad network that drove the activity to get their help investigating (being careful to not over-share data that could give away our fraud thresholds or other sensitive information). After investigations, the networks:
- Pause the fraudulent DSPs / exchanges.
- Block the fraudsters (be it a publisher, range of IPs, etc.).
- Look into publishers to ensure ads are firing correctly, proper frequency caps are in place, etc. (Some networks needed to build this tech out.)
- In some cases, we push for reimbursement.
One of our clients saw click fraud drop more than 99% (not a typo) and install fraud drop 75%. That’s staggering to think about. Today, fraud hovers around less than 1% of total traffic.
We’re not naive; we know fraudsters will continue to find new, creative, and cunning ways to steal ad spend. We also know we’re not capturing 100% of fraudulent activity because there are simply too many ways to game the system. But we actively work with our MMPs to develop new fraud monitoring tools and reports, networks to implement solutions like click validation, and fraud tool vendors to make sure our toolkit is up to date and ready for the next trend.
Amazing work is being done by companies to utilize blockchain tech, machine learning, AI, and more to fight fraud. But don’t wait around and expose yourself to risk. As advertisers, it’s our job to do what we can, however we can, as soon as we can. Join us in our fight against fraud.
Watch for other installments in our series to see how we’re pushing for transparency and addressing impression-level fraud. And don’t miss our talk at MAU Vegas 2019 (War Stories and Tactics for Fighting Mobile Ad Fraud – Wednesday, May 1 @ 3:40PM), where we’ll cover this and much more.