You have to diagnose what could be off on your end before making extreme blanket statements.
These ad systems are tools, and tools only work as well as how you use them.
My only expectations is being able to believe real people are actually clicking the ads.
Which I don't.
What is your website/landing page URL and what did your ad say?
And considering the pervasiveness of tracking, why isn't the fraud better tracked?
Because the people who are best-positioned to do something about click fraud also make heaps of money off of click fraud.
For example, browser extensions like AdNauseam that randomly "click" on ad links are maybe meant to be a middle finger aimed at the ad networks, but, realistically, if you, the person doing the advertising, are paying for clicks, then the main effect of such extensions is to transfer money from your pocket to the ad network's.
On the advertiser side, fraud has a small impact due to idiot brands bidding up prices and also another small impact due to the fact that you need to cull all the worst placements/domains when you first start using a network or launch a campaign.
On the publisher side, each fraud penny earned is a penny directly taken away from legitimate publishers.
Fraud is much more a problem for publishers than for advertisers.
>There's a lot of fraud
Makes no sense.
Everytime AdTech comes up on HN, there's inevitably people who defend advertising with "you're not doing it right". Sorry, but the industry is a cesspool.
Are you retargeting someone immediately after they purchased? Before their purchase even arrived? Is it a good targeting list?
We have 3 buckets of customers, based on historical buying cycle: 0-30, 31-60, 61-90. We run retargeting within their purchase cycle, which is friendly and relevant.
Customers get churned out of retargeting if not within their buying cycle, which saves us money, and saves them irritation.
I'm not referring to real conversions, where yes, skills in marketing and using the tools does make all the difference.
All I'm saying is that I paid for 200 clicks that appear to not even have real people clicking anywhere at all.
What you need to do is to exclude bad placements/websites from your campaign. But even before that, you have to see if your ads + landing page are compelling, or you'll just waste a ton of money.
And how do you determine that, ex ante?
At the end of the day we're still talking about sales, just in a new form, in a multi-dimensional bazaar called the Internet.
That is, people that have visited my site in the past(with separate ad groups to test "expanded audiences"), currently visiting another site with the target keywords, from a targeted location.
And ads are generic branding and go to the homepage.
Conversion goal: simply clicking anything at all.
Depends on the type of site. If I follow a link to read something, I'll probably read it and leave. If the site tries to make me click something, I'll probably just leave.