Skip to content
WEBPRO International

A WEBPRO White Paper / The Sequel

Wolf. Wolf. Wolf. Wolf.

One false alarm was a coincidence. Four of them, arriving on a loop, is a business model. Here are the other notices that scare healthy sites, and the same quiet incentive humming behind every one.

We told you about the first cry. A client got an email that said “blocked by robots.txt” in a font that meant emergency, and the emergency turned out to be a lock doing its job. One email, one false alarm, one tidy little lesson. You could almost write it off as a fluke of automated messaging.

Except it isn’t one email. The scary notice about robots.txt has siblings, and they all share the family resemblance. Each one takes a normal, healthy, entirely expected fact about a website and dresses it up as a crisis. Each one lands in the inbox of the person least able to tell the difference. And each one nudges, gently, toward the same exit. The paid one.

So no, we’re not going to number these Wolf 1, Wolf 2, Wolf 3. That would miss the point. The power isn’t in any single notice. It’s in the loop. Four different alarms, four different triggers, one identical purpose. Let’s put them side by side and let the pattern speak.

The site that “just” started getting impressions

In April 2026, a wave of site owners opened their inbox to a message that made no sense. Google, it said, had started collecting Search impressions for their site as of a date a few days prior. For a business that had been ranking and reporting data for a decade, that reads like a death notice. It means one of two things, and both are catastrophic. Either your entire history just vanished, or Google is telling you it only just discovered you exist.

Google Search Console · Automated Notification

Search Console started collecting Search impressions

We have started collecting Google Search impressions for your property as of a recent date.

Impressions collected from: [recent date]

Explore your latest performance data in the Search Console performance report.

Neither thing happened. Nothing vanished. Nothing was discovered. It was a backend messaging glitch, and Google’s own John Mueller said as much in public, plainly. No penalty, no indexing issue, no ranking drop, just a broken notification firing off to millions of people at once.

Here’s the part that makes the timing so telling. That glitch landed right on the heels of Google admitting it had been over-reporting impressions in Search Console for nearly a year, from May 2025 into early 2026. So the same numbers people had been staring at, trusting, and reporting to their bosses were inflated the whole time. And the correction to that arrived wearing the costume of “we just started counting.” A house that has been standing for ten years does not get a letter saying the foundation was poured yesterday. Unless the letter is broken. It was broken.

The Truth

A messaging bug. Your data is intact, your history is intact, and the site never went anywhere. The only thing that moved was your heart rate.

The “not indexed” pages you never wanted indexed

Open the Pages report in Search Console and you will find a table titled “Why pages aren’t indexed.” It lists URLs, sometimes hundreds of them, under stern-sounding labels. To an untrained eye, that table looks like a list of failures. A tally of everything broken on the site. People screenshot it in a panic and send it over asking why half their website is “not working.”

Almost none of it is broken. That table is where a well-run site parks all the things it deliberately keeps out of the index. Pages you marked noindex on purpose. Old URLs you redirected on purpose. Parameter versions of a page, the sort ecommerce filters spit out by the dozen, that were never meant to rank on their own. Google even lists a page as “excluded” when it obeyed your own instruction to exclude it, then files that obedience under a heading that reads like a warning.

Google Search Console · Page Indexing Report

Why pages aren't indexed

Excluded by ‘noindex’ tag - 214 pages

Page with redirect - 89 pages

Alternate page with proper canonical tag - 156 pages

These pages are not indexed and will not appear on Google

Google has actually said the quiet part out loud on this one. Their own people have gone on record that many of the reports in that section are not real problems, and that a lot of what gets flagged needs no action at all. Which raises a fair question. If they know it isn’t a problem, why file it under a heading built to look like one. A “noindex” that is working is a success. Reporting it in red next to genuine failures is a choice about how the message should feel.

The Truth

Intentional exclusions are the mark of a controlled, deliberate site. The report mixes your good decisions in with actual errors and lets you sort out which panic is real.

The canonical Google “chose” instead of yours

This one has a sharper edge, because the wording is almost adversarial. “Duplicate, Google chose different canonical than user.” Read as a human, that sentence says Google looked at the URL you designated as the real one and overruled you. It took your authority and handed it somewhere else. For anyone who has spent money and years building a page, that lands like a theft.

Google Search Console · Page Indexing Report

Duplicate, Google chose different canonical than user

User-declared canonical: /product/blue-widget

Google-selected canonical: /product/widget

This page is not indexed

Here’s what that scary label is not telling you plainly. A canonical tag was never a command. Google treats it as a hint, and always has, and says so in its own documentation. So when two pages are near-identical, which is normal on any ecommerce catalog with color and size variants, Google picks one to represent the set and consolidates the signals onto it. That is the system doing exactly what it is supposed to do with duplicate content. It is not stealing your authority. It is deciding which of two nearly identical doors to put the welcome mat in front of.

On a healthy site this notice usually means one thing. You have similar pages, Google merged them sensibly, and the ranking power went where it was always going to go. The label “chose different canonical than user” is worded to sound like a demotion. Most of the time it is a filing decision on pages that were duplicates by design.

The Truth

A canonical is a suggestion, not a contract. On duplicate-by-design pages this is normal consolidation, not a seizure of your authority.

The overnight cliff that wasn’t a penalty

This is the one that empties the room. You open the performance report and the graph has fallen off a cliff. Impressions down a third overnight. Clicks maybe holding, maybe not. Every instinct built into a site owner screams the same word. Penalty. You got hit. The algorithm turned on you while you slept.

Google Search Console · Performance Report

Impressions ▼ 38% · last 7 days

Total impressions: sharp decline beginning [overnight]

Total clicks: roughly stable over same period

Average position: improved

Look at that third line. Impressions collapsed, but clicks held steady and average position actually got better. A real penalty does not work that way. When Google punishes a site, the clicks go with the impressions, because you’re being buried in the rankings where real people can’t find you. Here the real people kept clicking exactly as before. Only the phantom traffic disappeared.

Because that is what it was. In late 2025 Google stopped supporting a URL trick that let rank-tracking tools pull 100 results in one request. Overnight, a huge volume of bot and scraper “views” that had been silently padding everyone’s impression counts simply stopped registering. Impressions dropped across the entire industry at once. Rankings looked better because all those views from position 80 vanished. Clicks didn’t move because clicks were always human. The cliff was a bunch of robots leaving the building, and Google’s graph drew it in the exact shape of a catastrophe.

The Truth

Clicks steady plus position up equals no penalty. The drop was inflated bot traffic being stripped out, finally showing numbers closer to real humans.

Four cries, one wolf

Line them up. A glitch that says you’re brand new. A report that files your best decisions next to real errors. A label that says your authority was taken when it was merely consolidated. A graph that draws a bot exodus as a bloodbath. Four completely different triggers. Four completely different underlying facts. And every single one arrives with the same tone, the same urgency, and the same unspoken suggestion. Something is wrong with your free traffic. You should be worried. You should do something.

We said it in the first paper and it holds for all four. This is not a conspiracy in a boardroom. Nobody is writing individual scare emails about veteran operators who know the rules cold. It is quieter and more effective than that. Google’s money comes from paid search, not organic, and so the whole apparatus is tuned by incentive to make the free channel feel shaky. A confident site owner is a problem. A rattled one opens the ad account.

Cry wolf often enough about the free road and the toll road starts to look like safety.

The shepherd boy cried wolf and the village stopped believing him. That story had a moral about lying. This one has a different moral, because the boy here isn’t lying by accident. He’s crying wolf on purpose, over and over, on a schedule, and every time you come running he points you toward a road that costs money. The wolf was never at the door. The door was locked. It was always locked. That was the point.

The tragedy of the fable was the villagers who stopped listening. The tragedy here is the opposite. Most site owners never stop listening, because they never learn that the cry was hollow in the first place. They just keep running, and keep paying, and keep handing over confidence they never needed to lose. Unless somebody shows them the pattern. Which is the entire reason we write these down.

The bigger picture

Four emails. Still the same book.

The first paper dissected one notice. This one caught four more in the wild, all crying the same wolf, all pointing at the same paid exit. And this is still just the inbox. It doesn't touch the shifting rules, the moving goalposts, the updates that reset the board every few months, or the strategy that lets an ordinary business win the organic game in spite of all of it.

We wrote the whole thing down. The playbook, the incentives, the pattern behind the panic, and the plan for beating it. It is the book Google would have preferred stayed unwritten.

Scroogled

SEO Survival. How Google Killed the Internet, and How You Can Win.

More from SEO Insights

  • · White Paper

    When Google Cries Wolf

    A single automated email spooked a client whose site was doing everything right. Here is what really happened, and what that false alarm quietly reveals about the game being played.

  • · AI Search Measurement

    Certainty For Sale.

    I read all 38 pages of Similarweb's 2026 AI search report so you wouldn't have to. The summary doesn't match the charts, one headline number changed after publication with no note, and the only advice worth taking came from a guest. Here's what's actually in there.

  • · Content Strategy

    The 91% Lie

    A client wanted to bolt a borrowed statistic onto his homepage. It claims people pick businesses on culture, not cost. The reports it cites say the opposite, and repeating it now carries a penalty that didn't exist a year ago.