It's Bots All the Way Down, What's an Ad Worth Anyway?

Overall, bots—good and bad—are responsible for 52 percent of web traffic, according to a new report by the security firm Imperva, which issues an annual assessment of bot activity online.


So many bots, and the ratio of bot vs human traffic is continuing to increase. The same report tells that 29% of all web traffic is just malicious bots, while 23% are benign. The latter of course includes search engine crawlers and feed fetchers, for example.

One kind of bot that neither the article nor report mention is the kind that fakes ad clicks. Looking at the numbers reported that sort of bot must be much less than 1% of web traffic. The size of a rounding error.

Or maybe it's just indistinguishable from human traffic. That is, after all, the very idea. Farms of smartphones pretending to be humans surfing the web and clicking ads so that site owners get ad income.

The cost of ads is as far as I know still measured in clicks. An organisation pays for each click its ads get, and each year the price per click drops significantly. Google and Facebook and other ad networks have had to expand their client base exponentially each year to stay profitable.

With GDPR and other privacy measures in place, how will the price of ad clicks be affected? What's a badly targeted ad mainly visited by bots worth? Even without GDPR I suspect that an increase in bot traffic must drive the price down. Do organisations want to pay for clicks that don't lead to conversions? (For those unused to the lingo a "conversion" is a visit that leads to a sale of an item or a sign up to a service.)

At a previous workplace they used both Facebook Analytics and Google Analytics. Both claimed to keep track of conversions, in this case new subscriptions to the service. The interesting part was that when one added the claimed conversions from Facebook to the claimed conversions of Google the number vastly exceeded the actual number of new customers for the period. And that doesn't even include conversions gained from other ad networks.

The marketing department didn't seem to care. It was possibly worth paying double for some customers. After all measuring it must be hard.

But that leads to another question: are even clicks that lead to conversions decreasing in value? I doubt that marketing network want more accurate measurements if they lead to lower billing. Precision could be a selling point, but how does one prove it?

All in all I hope that the ad network market collapses from low profitability. It's possible that it happens, with the multi-pronged challenges of privacy demands and uselessness of clicks. I hope it's soon.

-- CC0 Björn Wärmedal