But why is that? Do search engines rank better when a text is longer? Or is it that with more words there's a higher possibility that one will match with what the user is searching? And doesn't it all just make users waste more time searching for something concrete?

Maybe there's space for search engines that can summarize a page if it's long and give the user what they want.

It's not really longer/shorter, the core issue is that people don't search for witty titles, they are usually trying to get something done.

In SEO terms the "get something done" is described as "user intent".

Imagine you have written an article describing how to change the oil in a Ford Mustang.

What title better fits user intent:

"How to change the oil on a Ford Mustang"

or

"An afternoon of tomfoolery, broken wrenches and dirty hands with an American classic"

If you are at all interested in people finding you on Google you need to have a title that's more like the former than the latter.

On the one hand, that sounds like a good thing, rather than a bad thing. On the other hand, that doesn't match up with my experience at all. There's far too much nonsense out there that sounds like the latter and exists to waste your time.

yeah, many studies out there proving that longer text equal to better ranking; which also comes with another benefit of hitting long tail keywords.

> Unfortunately SEO ruins great, witty and brief marketing texts.

SEO content marketing is basically a descendent of direct marketing -- the stuff you used to get in the mail, in two-page spreads in Reader's Digest and so forth.

Direct marketers have known for a very long time that long-form copy sells better. They had the numbers to prove it long before "analytics" was a common idea -- the very term A/B testing refers to the "A" and "B" sides of a newspaper print roller being able to print two versions of an advertisement.

Source: Tested Advertising Methods by Caples. A fascinating, if depressing, read.

> the very term A/B testing refers to the "A" and "B" sides of a newspaper print roller being able to print two versions of an advertisement.

Really? That's interesting. But how could they tell which version converted better? Maybe distributing each version in a different area?

That was the beauty of it. As the newsprint passes through, one paper will get A, the next will get B, the next A and so on. The plates are otherwise identical for each page being printed.

This perfect 50/50 split pretty much controls for all the plausible variables. Location, time of day, you name it, the odds of a reader seeing either A or B was exactly 50%.

No it doesn't it has nothing to do with that. However, if you think an article should be a paragraph long then you probably didn't have anything worth creating the page for anyway.

Plenty of ecommerce sites ranks with very little text on their pages. What you are saying is just wrong.