The traffic drop that wasn't an SEO problem

A client asked me to look at a site they'd acquired. Organic traffic was down and had been sliding for two months. They wanted to know what had broken in search.

Fair question. Traffic falls, you look at search.

I checked rankings first because that's the cheap check. The site still ranked first for its own brand name. I went through the rest and it all sat where it'd always sat.

So rankings held and traffic fell. Those two don't sit together unless something outside the rankings changed.

I pulled Search Console and split the queries into branded and non-branded. Non-branded held flat. Branded had dropped, with the homepage wearing most of it. (Brand searches land on the homepage, which tracked.)

Fewer people were typing the company's name into Google. I can't fix that with a title tag.

Then the timing. The slide started the month their paid ads were switched off.

What was actually happening

The ads had been manufacturing brand demand. Someone sees an ad, doesn't click, searches the name a week later. That search lands in organic and organic takes the credit.

Turn the ads off and the demand goes with them. Rankings stay put. The harvest gets smaller. That's the whole mechanism.

Where a site ranks is one thing. How many people go looking is another, and I don't own it. I get measured on both.

Almost no report separates them. A dashboard says clicks are down, everyone reaches for the ranking explanation, because ranking is the lever in the room.

The check

Split branded from non-branded. Then look at impressions.

Impressions tell me about demand. Average position tells me about rank. Neither one's hard to pull. If impressions fell while position held, the search results didn't change. The number of people searching did.

Twenty minutes. I run it first on any decline now, before I open a crawler or read a single title tag, because what comes back decides whether there's anything here for me to fix at all.

Sometimes it sends the question somewhere else. A media budget. A seasonal cycle. A rebrand nobody told me about. A product they quietly stopped stocking.

Sometimes it comes back clean and the problem really is mine. Then I'll go do the work.

The awkward part

So whose problem was it? Not mine, and that's a strange thing to say out loud to someone who's paying you.

I get hired to fix organic. I report back that organic's working correctly and the cause sits with a decision made two floors away. There's real pull to find something to do anyway, some technical debt to clean, so the engagement has a deliverable in it.

I think that pull is worth naming. Giving in to it is how you end up rewriting meta descriptions for a problem that was a budget line.

The uncomfortable half is the one nobody asks about. If paid was inflating organic on the way down, it was inflating organic on the way up. Every good month while those ads ran was partly borrowed. I'd have been reporting it as mine.

I don't know a clean fix for that. I hold branded and non-branded apart in every report I build now, which at least keeps the borrowed half visible as borrowed. Most reports still blend the two.

What I'd do with it

Treat a decline as a question about demand before you treat it as a question about ranking. Rankings are easy to check and they rule themselves out fast.

And when the answer is upstream, say it plainly. A client who hears "your rankings are fine, this is your ad spend" gets something they can act on. A client who hears about crawl budget gets three months of work and the same decline.

If your organic's down this quarter, the first thing I'd check is what paid did ninety days before it started.