It is possible for advertisers to configure it so that this does not happen (in theory), but a lot simply do not bother.
I've seen some compelling case studies that it does work, and even at low conversion % for big-ticket items (like hotels etc) it's insanely cost effective.
If I shop around for a lamp on 5 different websites and finally buy from a 6th site, how do those first 5 websites learn that I bought a lamp?
The pervasive tracking of which retargeting is the most visible form is mostly what drove me to use an ad blocker. I wonder if the cost effectiveness calculation factors in lost future purchases?
> how do those first 5 websites learn that I bought a lamp?
Even Amazon's ad servering systems don't seem to be told that I bought the lamp from them never mind some other store.
Maybe they are hoping I liked it so much I'd decide that I need another lamp, so they want to remind me where I got the first one from?
> is the most visible form is mostly what drove me to use an ad blocker
My reason for blocking ads was because of the increasing incidence of malware riddled ad hosts often being used by high(ish) profile sites.
imgur.com was the final bail of hay that made me install network-wide protection at home. I also rehost any image I decide to share from there, to protect people outside my LAN. I don't want the tech support workload of cleaning other people's machines after what I'd too often seen their (obviously not sufficiently verified) ad partners try to pull.
My privacy is fairly shot already, but at least I can try to protect myself and others from other malware.
Assuming that five different websites would all run a re-targeting campaign for the same lamp - it's tricky, and you would probably be relying on the retargeting cookie expiring, or you hitting the frequency cap for each.
It's possible that at the level of the ad network the products could be disambiguated, and then cancelled so that other companies did not waste ad spend. This would be difficult as the products would have different SKUs.
At the level of a single retailer, the retailer places a burn pixel on the checkout page which registers to no longer place ads for that product.
Certain hotels and services are already targeting a captive audience.
If you're looking for hotels in London, you're already intent on finding a hotel.
A single 3 night booking in New York City will easily produce $100 of revenue for an online travel agency. So... Throwing $10 for a click is hardly a bad investment.
Other products and services are not like that.
Look for food delivery on Google and check how many ads are unnecessary for you
I've seen some compelling case studies that it does work, and even at low conversion % for big-ticket items (like hotels etc) it's insanely cost effective.