Google Made Rank Data Expensive to Collect. Your Tool Pays First.

Google confirmed on August 26 that it now routes every search result link through a redirect at google.com/goto, with the destination encoded instead of written out. Two things follow for you, and only one of them is about SEO tools.

The first is a bill you’ll see at renewal. The second is a number you should write down today, because if your organic traffic falls next month you now have two suspects instead of one, and the only thing that tells them apart is a baseline taken before anything moved.

Derek Perkins, founder of the rank-tracking firm Nozzle, measured the rollout at near 100% across several residential IP providers the same day Google confirmed it. Alex Greenland first spotted it on June 23, and Brodie Clark documented it across paid and organic listings on July 2. Two months from curiosity to default.

Here is the part that costs money. A results page used to be readable in one pass, because every link on it wrote out where it went. Now the destination is encoded, so anything reading that page has to follow each link to find out where it goes. Perkins gave the number:

“resolving all links for a single five-page ranking takes between 500 and 1,000 requests”

That’s one keyword. A tracker watching a thousand keywords used to make a manageable number of requests a day and now makes something closer to a million. Perkins added that rate limiting, not bandwidth or storage, becomes the binding constraint, which matters because rate limits are the one cost you can’t buy your way out of quickly.

Nobody sends you an invoice for this. It lands in your rank tracker’s cost of goods sold, and it reaches you as a price change, a slower refresh, or a quietly reduced keyword allowance at renewal.

Google’s entire explanation is one sentence

The company gave it through a spokesperson:

“We have a long history of deploying technical measures against evolving forms of abuse, and we regularly take steps to protect our services and users.”

No blog post, no Search Central entry, no help-page documentation. The statement names no target, and that vagueness is doing work: “abuse” covers scrapers reselling results, and it also covers the AI companies harvesting Google’s index to answer questions inside their own products. The mechanism doesn’t distinguish between them, and neither does the cost.

This is the second front in the same campaign. We covered the first one when Google refiled its scraping suit against SerpApi on August 10, narrowing the case to the licensed content sitting inside its results. That was a legal instrument, slow and contestable. This is a technical one, and it needed no court at all. One shipped a lawsuit against a single company. The other raised the price of the whole activity overnight.

What actually breaks, and what probably doesn’t

Two things are worth separating, because the coverage has been merging them all week.

Rank tracking gets expensive. That part is measured and confirmed. Every tool you pay for that reads a results page absorbs a multiple of its old cost, and that includes the AI-visibility trackers, which read Google’s AI Mode and AI Overviews the same way anyone else reads a results page. The scorecards that already disagreed with each other now have to disagree at a higher operating cost.

Your analytics probably survive. This is where people are getting ahead of the evidence. Digital Applied, writing before Google confirmed anything, was careful about it: depending on how the redirect passes headers, the referrer could survive, change, or drop entirely. If it drops, organic sessions land in GA4 as (direct) / (none) and your organic numbers fall off a cliff for no reason you did anything about. If it survives, nothing changes. Nobody has published a measured answer either way.

So don’t rewrite your attribution model this week. Check your own numbers instead, which is a smaller job and a real one.

The counter-case: this may be aimed past you entirely

The strongest argument against treating this as an SEO story is that Google isn’t thinking about your rank tracker at all.

The pressure Google is under right now comes from AI systems reading its results at industrial scale to answer questions elsewhere, which is the same pressure behind bots passing humans in web traffic this year. A redirect that makes bulk reading expensive hits that traffic hardest, and rank trackers are collateral. If that’s the real target, the tooling market adapts, the redirect gets decoded, prices settle, and in six months this is a footnote.

That reading is plausible and it doesn’t change what you should do, because collateral damage still costs the same as the intended kind.

Three moves

Ask your rank-tracking vendor, in writing, what this does to your renewal. Not whether they’ve noticed it. What it does to your price, your refresh frequency, and your keyword allowance. The ones who have already absorbed the cost will answer specifically. The ones who haven’t measured it yet will tell you they’re monitoring the situation, and that answer is the information.

Take a baseline of your organic share in GA4 this week, before anything moves. Write down organic sessions and (direct) / (none) sessions as a ratio, dated. If the referrer does start dropping, a drop looks identical to a ranking collapse, and the only thing that tells the two apart is a number you wrote down beforehand.

Stop treating rank position as a free number. It never was, and now the cost is visible. Decide which keywords genuinely need daily tracking, because that list is shorter than the one you’re paying to refresh, and the bill for the difference is about to get itemized whether you look at it or not.