When I was doing some work in this space for Travelocity many years ago the biggest problem was cannibalism of data and traffic. Yes, the traffic coming in from ads was shitty and generally costs more than it returned, but even still cannibalism was the bigger problem. We would have been better off taking all the ads off the site and spending money in other offline venues. It was so apparent that ad spend was a drug addiction. I still don’t understand how this was a mystery to anybody in management considering the strength of the brand and the success of their television ads.
Eventually the business imploded and the partner/affiliate segment became more valuable than the core business partially because it was absent online ad spend.
Optimization is part of the process to ensure wasteful spending is reduced as analytics and sales data is generated but unfortunately the biggest issue is the people who actually run these campaigns. Lots of politics, perverse incentives, lack of skills, and general apathy.
But the promise of performance marketing is exactly this immediate connection between ad spend and performance (i.e. sales).
Sure in broad strokes you are probably right that brand awareness has ROI positive in the long term. But then why would you buy targeted advertising vs a giant billboard on the main road?
To more specifically answer your question; I imagine they have some sort of strategy with advertising. A billboard campaign might work. I always see Barracuda ads in airports, but nowhere else. Soap operas are named because soap and consumer products advertised to homemakers during the day. Even when using a broad brush you look for the best use of your money and target your audience.
The problem with attribution is that it's entirely flawed. Just compare what Google Analytics reports for a campaign vs what Facebook is reporting for the same thing. They'll give you different results and in most cases you probably shouldn't trust either.
BUT, what they do is provide an easily accessible number that you can show your boss to show you had results. This is massively flawed but most marketers don't understand the mechanics of advertising or tracking well enough to understand why they're reporting garbage data.
Why is it an initial assumption that customer acquisition and ad spend are linked short or even medium term? That doesn't seem right.
When I see an ad for washing machines it doesn't make me run out and buy a washing machine. It gets me familiar with Samsung the brand so in 6-12 months when I buy a washing machine I am familiar with Samsung and buy their brand. Or it means that if my washing machine starts to sputter I'm more confident knowing where to go to buy a new one.
It doesn't seem enough to stop advertising for 6 months and claim no damage is being done. It seems well understood that repeated exposure to a brand has a powerful effect on humans.