What to Do About Poisoned SERPs: Takeaways From Our Women in Tech SEO AMA

Since mid-September, Google has been serving rank trackers poisoned SERPs: complete, normal-looking pages with results you probably wouldn't see if you searched yourself. Our Poisoned Google SERPs study measured how much and where. Now SEOs want to know what to do about it, starting with the reports they send every week.

On November 10, Women in Tech SEO invited Derek Perkins, Nozzle's founder and CEO, to a live AMA session to discuss this topic in their Slack. It was a busy AMA, which shouldn't surprise anyone who has seen all the chatter on LinkedIn and X about it over the past couple of months. This is one of the biggest changes to SEO data in years, and people are frustrated. Women in Tech SEO has published the full questions and answers. This post is the short version: the six changes we would make to SEO reporting today, in Derek's words, backed by numbers from the study.

1. Annotate everything since September 13

September 13 is when requests to Google started failing in large numbers. From September 20, many of the ones that got through came back poisoned, as many as 68.6% on the worst day so far, November 2, the high point on the chart below. Derek's advice:

Put big annotations around September 13 to today in any of your reports, rank tracking or GSC.

One thing we are seeing is that Google is pulling major domains out of a poisoned SERP and filling the gap with social platforms, forums and directories, so:

• A large brand should read any dip in that range as likely poisoning, not lost rankings.
• A smaller brand should be just as wary of a big jump. In Derek's words: "that might just be your site jumping up to replace those larger brands."

2. Don't fall back on Search Console alone

For many, the obvious response is to trust Search Console instead. Derek doesn't think it's exempt:

I don't believe that the vast number of teams inside Google are coordinating enough to make sure that poisoned data isn't being fed into GSC. Poisoned SERPs are what is actually being delivered, and thus it would take significant work to make sure that was filtered out.

Search Console has been shaky too. One person at the AMA saw impressions fall by double digits on September 17 without changing anything on their site, and Joy Hawkins called Search Console's data unusable just this week. It's one more data point, not the final say.

3. Use site analytics as your baseline, and spot-check the rest

Right now, your site analytics are the most stable reporting you have. If that isn't changing, it's likely that your SERPs haven't been upended.

When analytics do move, start with the landing pages that moved, then search the keywords those pages rank for yourself, logged in. We haven't seen any evidence of poisoning in logged-in results, so that's the closest thing to a clean check right now. If what you see looks nothing like what your rank tracker has reported for those keywords over the past few weeks, the tracker's latest numbers are probably poisoned.

4. Ask your rank tracker how it keeps poisoned SERPs out

The biggest key going forward is to make sure you understand how your data platforms are ensuring data integrity. As Google drives that cost up, there are going to be lots of temptations to cut corners.

When Google blocks a rank tracker outright, it sends back a CAPTCHA or an error page, and the tracker knows to try again. A poisoned SERP looks like a normal results page, so a tool that does nothing about it stores it as a real ranking. Four questions to put to any provider:

• How do you detect a poisoned SERP? It looks normal until you compare it with the keyword's usual results.
• What happens when you find one? Is it collected again, flagged, or kept?
• What about keywords you have no history for? A new keyword has no baseline to compare against.
• Can I see the SERP you collected? A screenshot or the raw HTML is the only way to check an anomaly yourself.

Our answers: we keep a rolling history of each keyword's usual organic sites and treat a SERP as poisoned when too many of those sites are missing from it. We then collect that keyword again, up to five times, through a different proxy network each time, until we get a clean SERP. Every SERP's raw HTML stays in the app, so you can check any of them yourself. New keywords are the hard case: until there's enough history, there's nothing to compare against.

Here's what that check looks like: the same search two days apart, both rendered from the HTML we stored.

5. Look hardest at "near me" and local-service keywords

Poisoning is not spread evenly. On page 1 from September 13 to November 3, the study found:

• "Near me" searches: 26.0% poisoned, nearly double the 14.6% rate for all other searches.
• Local services: legal (29.8%) and home services (25.2%) ran highest of any industry, against 14.0% of all US SERPs, and both topped three in five on their worst days.
• Brand searches: 8.1% for searches looking for a specific brand or site, the lowest of any group. Low, but not zero.

If your reporting leans on local services or "near me" searches, it needs the closest look.

6. Don't trust SEO tests or "who's winning" studies since mid-September

If you made changes to your site after mid-September, you may never be able to measure whether they worked. One SEO at the AMA had rewritten a client's titles on September 13, and a week later impressions were down about 35% while clicks held flat. Derek's verdict:

Honestly, I don't think there's any data available that can reliably help you answer if that change was effective. Between poisoned rank tracking and GSC data, plus algorithm changes over the period, it's probably a lost cause.

The same goes for the visibility studies published since mid-September that name the sites winning and losing in Google:

Every single one of those studies is based on poisoned data. We source SERP HTML from every major proxy network and provider, and I can promise that nobody is immune.

Studies that count something other than who ranks, such as how often AI Overviews appear, may still hold up.

Rank tracking isn't going away

I don't think this is an existential threat to rank tracking, as much as it has felt like it. While this is the most concentrated effort Google has ever put out to block us, there's only so far that they can go without impacting real users.

Google still drives most search traffic, and the SERP keeps changing, so knowing what's on it matters as much as ever. What's different now is how much work it takes a tracker to get a clean SERP, and how much you should ask yours to prove.

Thanks to Areej AbuAli and Erin Simmons for hosting, and to everyone in the Women in Tech SEO community who brought a question. Their write-up has every question and answer from the AMA, and our Poisoned Google SERPs study has all the charts and numbers if you want to dig in further.

And I highly recommend WTSFest. We exhibited in Portland and Philadelphia for the last two years and it's always a great time. Here's the three of us in Philadelphia doing our signature pose in front of the lights.