Hacker News Reader

Comment by brookst | Hacker News Reader

> It would wipe out their revenue.

Disagree. YouTube's revenue comes from large advertisers who can measure real impact of ads. If you wiped out all of the bots, the actual user actions ("sign up" / "buy") would remain about the same. Advertisers will happily pay the same amount of money to get 20% of the traffic and 100% of the sales. In fact, they'd likely pay more because then they could reduce investment in detecting bots.

Bots don't generate revenue, and the marketplace is somewhat efficient.

View on HN

> YouTube's revenue comes from large advertisers who can measure real impact of ads.

Not necessarily. First, attribution is not a solved problem. Second, not all advertisement spend is on direct merchandising, but rather for branding/positioning where "sign up" / "buy" metrics are meaningless to them.

> In fact, they'd likely pay more because then they could reduce investment in detecting bots.

A lot more. Preventing bots from eating up your entire digital advertising budget takes a lot of time and money.

Making advertising more efficient would also open up opportunities for smaller players. Right now only the big guys have the chops to carpet-bomb the market regardless of bots. Noise benefits those who can afford to stand above it.

Pay per impression is a metric adexchanges use. So it is not in best interest of companies to remove bots. Bots create "economic activity" when such metrics are used.

Yes... but maybe also no. Well measured advertising budgets are definitely part of the game. But so are poorly measured campaigns. Type B often cargo cult A. It's far from a perfect market.

In any case, Adwords is at this point a very established product... very much an incumbent. Disruption generally, does not play to their favor by default.

Advertisers have spent decades pressing Google to do something about fraudulent ad clicks/views, and occasionally tried to sue Google over being billed for the fraud.

And yet Google is still the dominant player here and cranking in billions, so it seems that until some other actual competitor shows up, they have no reason to change their behavior.

Advertisers would have to pay far less, because of fewer fake impressions. Furthermore, advertising would seem to be more effective, since bots don't buy the product. The publishers, however, would hate it.

The problem is, the bots seem like a scam perpetrated by publishers to inflate their revenue.

What on Earth has given so many people in this thread the confidence to assert that marketing departments actually have any real way to gauge the effectiveness of a given ad campaign? It's effectively impossible to adjust for all the confounding variables in such a chaotic system, so ad spend is instead determined by internal politicking, pseudoscientific voodoo, and the deftness of the marketing department's ability to kiss executive ass. This ain't science, it's perversely-incentivized emotion.

The fact that sales go up in correlation to when marketing campaigns are run.

As opposed to, say, the million other reasons sales could be going up or down? This is the post-hoc rationalization of a random walk.

Every time the marketing guy does a campaign, sales go up.

Reply on news.ycombinator.com