As long as companies can measure a return on ad spend (ROAS) of >1, their ads are delivering value and they'll keep spending. The book you're naming seems to be saying that they can't actually measure ROAS. But that just seems... out of touch? Clicks are tracked from the ad, to the storefront, to the buy button, to the final exchange of money; it's pretty measurable. A ton of internet-only brands get almost all of their sales just from advertising -- are the dollars in their bank accounts lying to them?

Of course, there's the fantasy that events like IDFA deprecation will make the ad market crash. Spoiler: IDFA deprecation has been more of a blip than a bomb. Advertisers rapidly found ways around it (whether those ways are in TOS compliance or not). Then there are the brands coming into the space just looking for impressions -- think of your old school giants like Coca Cola. They're already comfortable with fuzzy, somewhat unknowable returns.

Haven't read the book but I have close contact with multiple people in the industry. And as I said, their results are right there in dollar signs. (For the record: I'd love to believe the ad market actually is threatened, I'm not arguing for it and personally I block ads.)

I think the problem with this idea (and to be clear I haven't read the book) is that advertisers are much better able to track the value advertising delivers than they were in decades past. Even post Apple privacy changes, I can still very clearly see exactly what I'm getting from my Facebook ads in terms of revenue and profit, which just isn't possible with things like TV, billboards, etc.

In fact, I'd argue that the internet ad market continues to boom because it has continually delivered highly measurable value for years.

It also seems very naive to me to say that large advertisers don't understand what they're doing. In many cases we're talking about large public companies whose expenses are scrutinized by major shareholders. If they were throwing money away on useless ads that didn't deliver value, there would be activists in there pushing for change.

Do you have data-based evidence supporting those claims?

My understanding of the field is quite the opposite. The only way to measure the true value of online advertising is to measure how much sales have been generated with ads enabled minus the sales that would have been generated without the ads.

To the best of my knowledge, this is unsolved, even by the giants who have plenty of talent and data to throw at the problem. And for a reason, this is a very hard causal problem to solve, with many (many) confounding variables.

> To the best of my knowledge, this is unsolved

Ideally, you could support this informed opinion with some studies? Preferably not just general anti-ads clickbait.

Sure, you can’t ever /really/ know if your “inception” in the case of branding campaigns worked, but product/seasonal campaigns are super easy to measure and prove.

Well, the onus is on people who make the claim that it works to put the scientific material forward, not the other way around.

And so, "to the best of my knowledge, this is unsolved" is not an opinion but a fact.

I am happy to be proven wrong with the relevant material.

And by material, I mean scientific papers, not assertions such as "product/seasonal campaigns are super easy to measure and prove."

I understand your sentiment, but surely you realise that the entire ads industry has a lot of proven outcomes? All the millions/billions spent by customers isn’t done so blindly. There’s quite a bit of research on this topic - googling will give you a ton of results, and then a lot of hyped “ads don’t work” material too.

I spent six years in ad tech. I definitely realize that the industry has proven outcomes. My point is that these outcomes are not about what is called "incrementality" in the field, but about attributed clicks/sales (which could very well be a zero-sum game).

And so, if you google-scholar (not google) for "online advertising incrementality", you are left with (granted, excellent) work that shows that nothing has been proven yet and that it is a very hard problem [1].

And when customers of Google, Facebook and other ad tech companies ask them about incrementality, their answer is surprisingly quiet.

[1] https://dl.acm.org/doi/abs/10.1145/3447548.3470819

PS: I don't make any hyped claim that ads don't work. I question this fact in a constructive way.

I’ve spent ~5 years in ad-tech, and I’m well aware that it is not irrefutably solved. But I would attribute that to measurement limitations, than actual conceptual problems. I’ve seen the impact ads can have for SMBs and MNCs alike across verticals.

Proving specific incrementality is rarely a concern from actual customers, unless they wanted to tighten their purse strings. Nobody questioned if actual incremental gains could be made - only whether there are diminishing returns for their increased spend.