With digital ads, it’s not the quantity, it’s the… well, it’s not the quantity, anyway.

Last year JP Morgan Chase decided to rethink the idea of advertising on 400,000 different web sites every month, especially when it had no idea what sort of content it was sponsoring.

It dropped back to a mere 5,000 channels all vetted by hand, eye, and human brain and, surprise surprise, discovered that the drastic reduction had very little effect on their anticipated metrics.

P&G similarly slashed its own digital ad buy by a whopping US$100 million and in a result that made hundreds of its digital analysts knock their heads against their cubicle walls, saw their sales actually increase by 2%.

It’s all a little too early and anecdotal to start claiming that up is down and plaid is in again, but it’s sure looking more and more like not even the big boys know what they’re doing.

The problem here is that for the last couple of decades one of the keenest selling points of online media has been the notion of highly targeted marketing – getting a niche message to a niche customer.

Only a really interested party is supposed to receive the ad and in an appropriate location that will make sense to the prospect.

What could possibly go wrong? Yes, right.

There have been many instances of programmatic ad platforms placing a mainstream brand next to a video of a terrorist barking out violent, hateful speech creating…awkward…unintended affiliations and sponsored support situations.

Maybe they should be looking at the sites with people that at least have a human mind who can make a proper decision. But then there’s still the issue of ignorance.

Even people with no lives can’t be expected to be experts on thousands, tens of thousands, hundreds of thousands of sites… what they publish, what their views are in general, and how good a fit they’d be with various advertisers.

As it stands, fewer than 5 percent of the many trillions of ad spaces in daily inventories of digital exchanges are sites run by “big media.” The rest are tiny independents that have every size audience imaginable.

Then there’s Google’s AdWords with their own two million+ smaller sites and Google’s YouTube with several million more opportunities for buying media space.

We have all of these measurement tools and all of these options, but without studying the behaviors of each unique audience group for each outlet, whether a site, YouTube channel, or an app sponsorship, advertisers and their advising agencies are just guessing.

If the P&G and JP Morgan results turn out to be consistently repeatable and amongst a wider group of brands, then it may not matter, is the moral of this story.

It doesn’t bode well to run niche content to try and catch some of that long tail magic if sales aren’t coming.

If I were running a DSP (demand-side platform) or other programmatic ad buying service, it would scare me and, sadly, my investors, to learn that 95% of my inventory is incapable of moving the needle.

Maybe it’s time to start investing in ideas instead of places to hang calls-to-action and discounts.