In fact, probably the most expensive campaigns are used to consolidate market shares, brand positioning and notoriety.
People just seem to forget that advertising existed before the internet, and it worked without the bloat of metrics.
All winning campaigns reflect in dollars. The question is when. That's why it's imperative to separate markets entirely.
Radio ads for a "branding/awareness" strategy. Search ads for a direct response strategy. The majority of people viewing/hearing one should never hear/view the other.
If it takes 6 months to properly test a branding campaign, decide ahead of time whether your organization has the patience to wait. If not, just stick with direct response.
Also, Scientific Advertising was written in 1923.
There's no space for brand positioning campaigns, or brand awareness, or anything that it's not tracked into a direct sale/lead.
Like you said, branding takes time, with huge investments - and usually it's almost a leap of faith because of the sheer amount of budget and pressure it requires, namely in terms of GRPs.
That's why testing used to be done in focus groups, and other forms of market research, to support a brand strategy.
Maybe branding is a "luxury" SMEs can't afford, but the thing they don't get is: it's a long term investment.
Brand advertising has a much higher ROAS when done well. The challenge is doing it well, and knowing if you did. My experience has been that SMEs rarely have the appreciation for art necessary to do branding well, and without strong branding, brand campaigns are a waste of money.
And it's not just a lack money. They'd rather spend the money on ads than on people. In some industries, that's a better strategy, but for most they would benefit greatly from recognizing that all customers, both B2B and B2C, are in fact humans.
Founders have an advantage here, because they can operate on a longer time scale.
(Side note, GM spent $3140000000 on ads in 2018.)
The engineering bounties would be for open source hardware, software, and tooling challenges, with a heavy focus on optimizing robots and social computer vision problems.
The races each weekend would alternate between full size trucks in a strong man style competition, manned quadcopters doing search and rescue challenges, and autonomous cars in an urban race series.
I would spend the remaining $10,000,000 from the week's budget on vocational scholarships and micro-loans for low income single parents.
If anyone from GM is reading this, let's talk.
Most people consider the latter, because it’s hard to accept that the spend needs to compete with every other way of spending that money, not with other ad media.
Analysis paralysis.
If so, what does it mean for all of those dollars to be more effective?
Does that mean that today's average net user is now subscribing to three different car insurance plans, because of how effectively they've been advertised to?
For the majority of businesses, the opportunity cost of properly testing all available options is higher than any potential gains.
It's like asking a developer if their code really "worked", even though they didn't write it and run tests in every popular and obscure programming language.
Should more thought be given to alternative ad options? Probably.
Does that mean that the ads didn't work? No.
Outside of the VC backed world, people play with real money and the market decided what works and what doesn't.
Unfortunately, VC money distorts the market for everyone.
The questions you pose, though completely valid - and should be asked - are done in different levels.
If you set a marketing budget, it's because there was a decision to achieve some goal that is bound to marketing. The same goes for R&D, Acquisitions, Human Resources, etc.
That's why, in a classical corporate sense, at the beginning of the fiscal year, each department fights for a cut of the big budget to be invested.
Within a marketing budget you then have to split it again, and again, and by this time you should be at the advertising level. Within this level you have Advertising (the creative part, copy and art) & Media Buying. Etc.
But getting back to your question: being ROI positive can be quite deceiving, because you can have negative ROI yet get positive long term results. Example:
You, as a manager, decided that Amazon is a viable channel for sales and distribution. You evaluate the competition and you realize that for you to have some relevance on organic results, you need to improve your ranking through Amazon PPC and external traffic. You factor in Price Promotion, to present a competitive offer and to help your product being established.
You run that campaign at loss (negative ROI), for 2 months, but you manage to get yourself to the Top 10 listings for organic results.
Question - assuming it's a healthy market: is this a bad campaign?
(The obvious answer is: it depends. But assuming the short term, ROI negative, campaign can set you up to generate more sales in the future, it's a good campaign. Then arguably you could present other solutions, at lower cost, to produce the same outcomes.)
Metrics did exist, though more rudimentary. X subscribers, X viewers, etc.