Thin Content Isn't a Word Count Problem

Length was never the measure

Most owners hear "thin content" and think it means short. Add a few hundred words, round the number up, move on. That habit misses what actually happens on the page.

Picture two versions of the same page, "Signs Your Tree Needs Removal." One runs two thousand words and covers leaning trunks, exposed roots, storm damage, and pest activity, all in general terms any tree service anywhere could have written. The other runs four hundred words and describes one call: a leaning oak a crew assessed, why they recommended removal over cabling, and what that decision protected on the property next door. The first page is longer. The second page is the one with something in it.

The same split shows up on location pages. A city page that only swaps in the city name around identical paragraphs is thin at any length. A city page that names the neighborhoods actually served, the drive times between them, and a call the crew took in that part of town is not, even if it barely fills a screen.

Word count cannot tell those pages apart. Building pages that pass this test, long or short, is the actual work of content marketing, not a word-count exercise.

What a thin page is missing

Four things separate a page with something to say from a page that only sounds like it does. A page missing all four is thin no matter how long it runs. A page with even two or three of them usually is not. None of the four require a longer page. They require a writer who actually did or witnessed the work being described.

The first of these, first-hand detail, is the same signal behind E-E-A-T for local service content. A generic paragraph can be well written and still have none of it.

A summary of nothing has nothing to show

An AI Overview is a summary. Summarize a page built from general statements and the summary is just as general, which means it is not worth showing over a page that has something specific to say. Summarize the page with the leaning oak, the named part, the stated opinion, and there is something to quote. Summarize the page built from filler and there is nothing left to lift out of it.

That is the whole test. Not whether a page mentions the keyword enough times. Whether there is a sentence on it a summary could not have written on its own. Pages built with that in mind hold up as AI citation content, and the same bar applies across the wider work covered under AI search optimization.

The strip test

Take any page, yours or a competitor's, and try this. Go paragraph by paragraph and delete every sentence that would still be true if you swapped in a different company's name. Delete the throat-clearing ("in today's competitive market"), the reassurance ("we care about quality"), and the advice so generic it applies to every business in the category ("regular maintenance saves money").

What is left is the page's actual content. On a thin page, that is often one or two sentences out of twenty. On a substantial page, most of it survives, because most of it was specific to begin with.

A sentence like "Our team is committed to providing quality service for every customer" does not survive. Swap in any other company's name and it is still true. A sentence like "We check the flapper first because it is the cheapest part that causes a running toilet, and it is usually the actual fix" survives. That sentence names a part, gives a reason, and could be checked by a plumber who disagrees with it.

1. Copy the page's body text into a blank document.
2. Delete every sentence that would still be true with a competitor's name swapped in.
3. Delete every sentence that gives no number, no name, no timeframe, and no opinion.
4. Read what is left. That is the page.

A page that survives the strip test at length is not thin, whatever its word count says. A page that survives at two sentences is thin, whatever its word count says.

Where the padding hides

Padding rarely announces itself. It shows up as a few patterns that read fine sentence by sentence and add up to a thin page.

An opening paragraph that restates the title in longer words before saying anything new. A middle section of advice that would fit any business in the category, not just this one. A closing paragraph that asks a question instead of taking a position. Adjectives standing in for detail: "high quality," "trusted," "professional," none of which a reader can check. A list of benefits with not one drawback, tradeoff, or exception anywhere in it.

None of these sentences are wrong. They are just not doing anything. A page can be free of errors and still have almost nothing survive once the sentences that would fit any competitor are stripped out. Fixing that takes writing that could only have come from actually doing the work, serving the actual area, and being willing to say what you think, the kind of sentence that gives a summary something to find.

If you want a read on whether your own pages would survive the strip test, a free SEO audit looks at what's actually on your highest-traffic pages before anything else gets discussed.

Frequently asked questions