This article reflects information as of 2020. For the latest details, please contact us.
Written by:
Hello! Thank you for reading again this time! The PS5 has come out, hasn't it? Have you all gotten your hands on one? I lost the early lotteries and haven't gotten mine yet... I want to get one soon and play Spider-Man. Game consoles have evolved incredibly, haven't they? I wonder if a PS10 or so will come out before I die. —and there I went, talking about something completely unrelated to this article's theme. This time I'll be writing about "ad fraud"!
1. What is ad fraud?
Translated literally, "ad fraud" means just that—advertising fraud. It refers to the mechanism of intentionally making bots generate impressions and clicks on ads to fraudulently siphon off ad spend. It's meaningless no matter how many times a bot sees your ad, right?
2. A difference in awareness from other countries
In places like the United States, ad fraud has been debated for years. In recent years, as the web advertising market has grown, its quality has come to be treated as important. Along with that, the damage from web ad fraud and the like has grown. So in the US, countermeasures against ad fraud are already getting attention, and efforts are underway to build an environment that delivers more effective advertising. In Japan, however, this hasn't yet been debated to that extent. The reason is thought to be that Japan and the US use different metrics to measure the effect of web ads. In the US, impressions (the number of times an ad is displayed) are treated as a direct metric of effect. That's because their top priority is to convey accurate information without damaging the company's brand image. As a result, conversions and CPA are not treated as very important. Japan is the opposite: it emphasizes actually acquiring customers from ads, and when placing ads, it prioritizes ads with good conversion rates and CPA. As a result, bots may view and click ads but can't go all the way to a conversion, so in Japan bots have been judged to be harmless. Perhaps because it wasn't treated as a problem, according to survey results announced by the ad-tech company Pixalate, 81% of all impressions of desktop programmatic ads traded in Japan in Q1 2017 were ad fraud. Surely this ought to be taken seriously.
3. Countermeasures
Momentum Inc., which offers ad-fraud countermeasure services, recommends a three-tiered approach: "pre," "post," and "after." "pre" means using DSP providers that have adopted ad-verification countermeasure tools. Ad verification means the following:
Using measurement tools to check whether ads placed on the web are actually being seen, and whether they've been placed on inappropriate sites in a way that damages the brand, and controlling ad delivery accordingly.
By selecting DSP providers that are linked to blacklists of domains at high risk of ad fraud or brand damage, and trading ads through them, you can prevent wasteful buying in advance. Next is "post." By delivering ads through an analytics tag, you monitor ad trading in real time and build a blacklist while analyzing. Once a domain is on the blacklist, on its next visit you can display a dummy ad instead of the ad, suppressing the damage. Last is "after" analysis. The delivery-record data accumulated through ad delivery is thoroughly analyzed in detail through a combination of machine learning and staff visual checks, and domains at high risk of ad fraud or brand damage are blacklisted, keeping the list continually updated. By running these three cycles, they recommend, you can put countermeasures in place.
Reports referenced:
4. In closing
I was surprised too when I saw the data I introduced earlier. But we've also started seeing countermeasures nearby, like "select the photos" or "put the shape in the frame," haven't we? I feel countermeasures against bots will keep advancing. And since there are already many companies and tools specializing in countermeasures, advertisers should know a few of them. Thank you for reading this time too!