Word count is the wrong measure
The "minimum 1,500 words" rule is folklore. It survives because it's easy to check and easy to bill for. A page answering "what is a 301 redirect" in 180 words, correctly, is a good page. Padding it to 1,500 makes it worse and slower.
The real question is comparative: against the pages currently ranking for this query, does mine give the searcher something they'd otherwise miss? Original data, first-hand experience, real prices, a clearer explanation, a decision they can act on. If the honest answer is "nothing, it's a rearrangement of the top three results", the page is thin, and length won't change that.
It's also why AI-drafted content isn't automatically thin and human writing isn't automatically fine. A human on their fortieth paraphrase of a competitor's blog post is producing thin content by hand.
Scaled content abuse, in Google's own words
In March 2024 Google rewrote its spam policies, replacing "spammy automatically-generated content" with scaled content abuse. The old wording could be argued around by saying a human touched the page. The new one targets generating many pages primarily to manipulate rankings rather than to help people — by automation, by humans, or by both.
That moved the line. Under the old wording an agency could produce 200 templated pages with a writer filling blanks and claim compliance. Now that's the exact behaviour described. Enforcement is rarely a dramatic manual penalty either — most sites see something quieter: pages dropping out of the index, crawl attention shrinking, good pages losing ground alongside the thin ones.
The 200 city pages problem, which is also the doorway problem
Here's the version you'll actually be sold. Landing pages for every city you serve: "SEO services in Pune", "SEO services in Nashik", on to two hundred. One template, the city name swapped in six places, sometimes a landmark name-dropped. It's cheap to produce, the deliverable looks enormous on an invoice, and it occasionally works for a few months — which is why it keeps getting sold. Then a spam update lands and the whole folder goes.
Google's older doorway policy describes the same thing from another angle: a page built to catch a query variation and funnel the visitor elsewhere. The test is what happens after the click. If fifty location pages all say "we serve [city], contact us" and push to one form, the reader gained nothing by landing on the city version.
We run city pages ourselves, so let's be straight about the line. A city page earns its URL when it carries something true and specific that couldn't appear on any other city's page — the real competitive picture there, local search behaviour, pricing that reflects what businesses in that market actually pay. If we can't write that, we don't publish it. A local SEO page that helps someone in Kochi decide is a destination; the same page with "Kochi" swapped for "Kollam" is a doorway.
Same test for service permutations: "affordable SEO services", "best SEO services" and "professional SEO services" as three URLs is one page and two cannibalisation problems.
Fixing it: consolidate, improve, or cut
The remedy is the same however the thin pages got there. Inventory every URL, decide one of three things about each, and act in a single release.
- Consolidate. Merge near-duplicates into one strong page and 301 the rest to it. Four thin service variants become one page that ranks for all four.
- Improve. Keep pages where you can genuinely add something — real prices, real local detail, real experience. Rewrite them properly or don't keep them.
- Cut. Remove pages with no traffic, no links and no purpose. Return 410 for dead URLs; 301 only where a real successor exists.
- Change the commissioning rule. Noindex is a holding move, not a fix. If nothing changes about how pages get approved, you'll rebuild the same folder within a year.