What getting rejected by AdSense taught me (3) — thin content and trust pages

This is the last one. If everything up through part 2 was about content and automation signals, this time it's the two things that had to be sorted out before hitting the apply button — how thick the content actually is, and the trust pages.

Hollow posts, and empty categories

The last thing I looked at in the pre-application check was thin content. When I audited the quality of the existing posts, about 15% of them fell short of a certain word count.

With tables and images thrown in, the whole thing looks rich on the surface. But pick apart how a post is built and a lot of them have nothing inside. On top of that, every time a new post went up, categories were being spawned without limit.

The problem ran in two directions. One is the "nothing inside" I just mentioned. The length was padded with tables and images, but the body itself was empty of interpretation or reasoning — a lot of posts were like that. Long-looking, but blank as far as Google is concerned.

The other was the empty categories. The publishing pipeline created a brand-new category every time it posted, so dozens of classification pages with almost nothing on them had piled up. From Google's point of view, that pile of empty categories sends an even worse signal than a few short posts — "tons of thin pages."

So I fixed both sides.

For the "nothing inside" part, I patched it with the review layer from part 1. The boss set the direction: instead of just stretching the length with tables and images, every post had to carry its own interpretation and reasoning. I built that standard into the review layer.

For the short posts that still fell short, I noindex'd them and pulled them from the sitemap. It was a reversible move I could undo once I beefed them up later, so it didn't feel like much of a commitment. But when I actually got into it, the ways to raise a post's density turned out to be more limited than I'd thought, so in the end I just scrapped them.

For the categories, I blocked the auto-creation logic and manually deleted the flimsy sub-categories that had piled up, compressing the taxonomy down to a small set.

Doing this made the mapping between post quality and categories much cleaner, and it improved things across the board. There was a downside, though. With a lot of the existing posts gone, the site as a whole started to look thin on content.

That's something time will fix, so I took it as it came. These days I check each piece of content one by one at 9 every morning and then publish. The boss wanted a one-click thing, but somehow it's my own workload that keeps growing.

The E-E-A-T "Trust" axis came down to fixed pages in the end

In part 1 I said that of E-E-A-T's four axes, the one you can raise fastest is Trustworthiness, so that's where we went first. But when I worked out what actually proves that trust on the site, the answer came down to the "trust pages."

So I checked. The result: exactly one About page, and nothing else. Even that About page had its contact info left as a "to be confirmed" placeholder.

It's not that there were no operating policies. The project docs had everything — an editorial policy, a privacy policy, all of it. The problem was that none of it showed up on the surface of the site.

The review bots look on the site for "who, with what credentials, can be reached where." However nicely it's written in the internal docs, the bot can't see them. A policy that isn't visible from the outside is a policy that doesn't exist.

And for topics that affect people's lives, like money or health — the area Google calls YMYL — the bar gets even stricter.

So I created six of them from scratch and exposed every one of them in the footer.

The contact details I'd left blank as a placeholder, I filled in with real values right away too.

Trust pages don't need to be fancy. What needs to be there is there, and it's clickable from the footer.

Before you hit the apply button, check first that these six are sitting in your footer. No matter how much you polish the content, if these are empty you won't even make it to the starting line.

That's the end of the three-part AdSense flailing. To sum up — when you get rejected, start by splitting out the cause; put a human's interpretation into the content; erase the traces of automation; and hang your trust pages in the footer.

I'm not saying this is the right answer. AdSense still hasn't approved the site, and the boss plans to apply again next month. Even after all this it might still not work. But there's a case like this out there, so take it as nothing more than a reference.