Zapier lost 1.5 million visits in one month. It's not alone.

Google visits to automation tools halved in a year, and even without Zapier the slide is steep.

Written by Miles Foster, Owner and tester, RankDude.18 October 2026 / 2 min read

Zapier had about 7.39 million visits from Google in August. In September it had about 5.90 million, RankDude data shows. That's a 20% drop in one month. But Zapier is only half the story.

The short version

-54.5%: visits from Google to these tools in a year.
Zapier: down 20% in September alone.
-42.3%: the yearly drop even with Zapier left out.
n8n: lost about two thirds of its visits.

WorldGoogle visits to AI workflow automation sites

Low 8.9MHigh 23MAverage 15M

Hover, tap, or use the arrow keys to inspect a month

Estimated monthly visits from Google, worldwide, one-off spikes smoothed. Source: RankDude analysis of DataForSEO data.

Zapier's bad month

Zapier is still the most visited site here, by a mile. It had about 5.90 million visits from Google in September.

That's about two thirds of every visit in the market.

It lost about 1.49 million visits in one month. That's bigger than the whole market's drop of about 1.44 million. Yes, really.

Over the year, Zapier explains about 79% of all lost visits. So is this just a Zapier problem? Not even close.

The numbers

Estimated visits from Google.

Whole market, Google visitsin 12 months: -54.5%
Zapier, in one monthAugust to September: -20.2%
n8n, Google visitsin 12 months: -67.3%

It's not just Zapier

Take Zapier out and the rest is still down 42.3% in a year. That's the number that tells the real story.

n8n got hit hardest of the big names. Its visits fell from about 2.39 million to about 780,000.

It held near 2.4 million a month until March 2026. Then it dropped to about 1.46 million in April and kept sliding.

And n8n is no tiny startup. The Berlin company raised $180 million in October 2025, Crowdfund Insider reported. That deal valued it at $2.5 billion.

Make kept its cool, at about 1.35 million visits. Smaller names fell hard. Bardeen lost about 81%, and Pabbly Connect lost about 70%.

So did anyone win?

The tiny winners

A few did. Albato more than tripled, from about 7,600 to about 25,000 visits.

Integrately rose 29%. Vellum rose about 12%.

September was calmer outside Zapier. Without it, the market actually rose 1.6% from August.

So where is everyone going?

Where did everyone go?

Fewer people are even typing the names. US searches for them fell 46% in a year, to about 345,000 a month.

Searches for n8n dropped from 246,000 a month in October 2025 to 90,500.

Here's the strange part. US visits from Google rose 27% over the year. So most of the drop is happening in other countries.

My guess? Some people now ask AI helpers to build automations for them. Others get answers without clicking. I can't prove that from this data.

The market sits near its two-year low of about 8.9 million visits in July. The peak was about 23 million in October 2025.

You can see every month on the AI workflow automation statistics page.

“Down 42% even without Zapier? That's not one company's bad month.”
Miles Foster

How I measured this

Visits are DataForSEO estimates of visits from Google search, worldwide unless I say US. They are estimates, not people, and name searches are US Google Ads figures.

More in my methodology and on the DataForSEO page.

Source: RankDude analysis of DataForSEO data, September 2026.

On this site

Sources

Crowdfund Insider, “Berlin Automation, AI Startup n8n Secures $180m in Series C Funding”, 13 October 2025.

Figures about the sites I track come from RankDude’s own monthly data (DataForSEO, Google Ads keyword data, Tranco, Google Chrome UX Report, Ahrefs, SerpApi and Google Trends); the figures here are for September 2026, published 15 October 2026, next update 1 November 2026. Each tool is explained on the testing tools pages. Reported facts about other companies are as published by the sources above.

Miles FosterOwner and tester, RankDude. Miles Foster owns RankDude and runs every hands-on test in person: one written procedure per category, the same sample sizes and scoring for every tool in it, with the measurements and screenshots published on each review's full test results.