We did the social media. It went fine, which was the problem.

The Grand Union build, part 2 of 8. Season 1: The Diagnosis.

TL;DR

Before we changed anything, we did the thing he asked for. More listings, more reels, more presence. The real estate content performed the way that content performs: some of it did nothing, some of it did fine, one or two did better than expected.

Fine is the worst possible result. A clear failure gets investigated. A mild one gets repeated, because there is always a reason to think the next batch will land.

The thing we should have measured was not engagement. It was whether anybody could tell who had posted it. Cover the name on any of those posts, put a competitor's logo in its place, and nothing about the post would have looked wrong.

Read next: Part 1: Nothing you do makes you stand out · Grand Union, the whole build · The referral you can't repeat

We actually ran the play

It would be a tidier story if I had refused the brief. I did not.

He wanted more social presence, that is a legitimate thing to want, and I had no evidence yet that it was the wrong lever. So we did it properly rather than half heartedly. New listings posted well. Walkthroughs. Neighborhood shots that looked like the neighborhood rather than like stock. Market notes that were accurate.

None of it was lazy work and none of it was wrong.

The results were mild

Some posts did nothing. Most did a little. A couple did better than the rest, for reasons that never repeated.

If you squint at a month of that, you can construct a story where it is working and just needs time. Everybody does. The numbers are never zero, so there is always something to point at, and the fix is always the same fix, which is more.

That is what mild results do. They fund their own continuation.

The test we were not running

Here is the check that would have saved us a few months, and I now run it on every piece of real estate content that crosses my desk.

Take the post. Cover the name. Put a competitor's logo on it. Does anything look wrong?

Not worse. Wrong. Out of place. Like it could not possibly have come from them.

Every single one of his posts passed that swap without a flicker. A photograph of a house he had listed, and a sentence about the house, and the neighborhood tagged. Any of four hundred agents in the metro could have published it exactly as written.

Which means the content was fine and the ownership was zero

The posts were not bad. They were unattributable.

That is a different failure from bad, and it does not show up in engagement numbers at all, because a stranger scrolling past does not need to know who you are to like a nice kitchen. The like is real. It just attaches to the house rather than to him.

So the account accumulated activity and did not accumulate a reputation. Twelve months of that produces a business exactly as distinguishable as it was on day one, with a content habit and a slightly tired owner.

What we should have asked first

Not is this good? Everything was good.

Is this ours? And if we stopped tomorrow, would anybody notice a gap shaped like us, or just a gap.

That question does not have a content answer. Which is why the next thing we did was stop making content.