Hacker News Reader

Comment by dkrich | Hacker News Reader

This is a well thought out bull case for AI paying off at Meta. I just wonder what the upper limit is on ad spend in this economy from that long tail of advertisers, especially to Meta users which seem to be mostly focused on consumer goods. Surely Meta is already tapping into an enormous portion of that limit.

View on HN

If AI can help an advertiser get the outcome they want in 1000 ad impressions instead of 5000 ad impressions (with more precise targeting or more appealing creative), Meta can 5x the price of those 1000 ad impressions, keep the advertiser happy with the same ROI they had before, and sell 4000 additional impressions to someone else as incremental profit.

Or they could 4.9x the price to show the current advertiser a better ROI and encourage higher spend from them, while still having the 4000 extra impressions to sell.

This is precisely why I’m skeptical. It assumes that that first advertiser can sell more of the same products just with better copy or more targeted placement and that more advertisers can backfill the previously used inventory that the first advertiser no longer needs and enjoy better sales too.

But doesn’t that then mean that many more product sales are taking place as opposed to the same number of sales spread across more impressions? I don’t see how many more sales can take place without users spending more money or the number of users going up and my original point was that meta is surely tapping into that addressable market as it is. There are only so many consumer dollars to go around.

It's way higher. I experimented with targeted search ads to try and sell some copies of Conveyor (a developer tool for desktop app devs, see bio). It was a waste of time. Google took a pile of cash and delivered zero conversions. It turns out that the conventional wisdom is now that search ads are too expensive for small companies. It's also extremely complicated. Even with a lot of work on targeting, the traffic you get is junk and it's hard to track conversions. I suspect most devs just block ads and never saw the ads at all, even if I was charged for them.

If Meta can use AI to make niche B2B advertising work better then there's still a lot of room for growth. Like, if there was a way to get AI to advertise my product to people asking frustrated questions about how to deploy their desktop apps, I'd be very interested in an offer like that.

They're also thinking about tapping into small-scale "mom and pop" shops, with lots of "AI"-augmented services, like customer support, translation, seo, marketing copy and so on. If they find a good price-point for that, they could bring in a lot of revenue from that side as well. They control lots of entry points for a lot of users, between fb, insta, whatsapp and so on.

B2B is kind of weird in advertising - you are still just selling to a person. So even Facebook advertising is used in b2b because targeting by job title/interest area is strong there.

IMO there isn’t much of an upper limit, as long as roas is decent the limit exists elsewhere (manufacturing, people, etc) and any gains in those areas means more advertising spend is viable.

Hyper targeting can grow the advertising pie. Eventually meta can become basically "the ultimate salesman" hand delivering users to 3rd parties that completely fulfill a problem the user has. The amount they could charge to completely take care of all user acquisition(especially if theres an implied threat of "we'll take all your customers to your competitor if you fuck around") is astounding.

Reply on news.ycombinator.com