Short answer: we read twelve real small business websites against the five-pattern framework from our first piece on AI slop and scored each one Very Low to High. The businesses are anonymized. The finding is not: budget did not predict the score. Specificity did.
The method
Twelve small businesses across Eastern Kentucky, spanning tourism, food service, hospitality, retail, professional services, and the arts. Each homepage was scored against the same five patterns from our first piece:
- Generic, templated phrasing
- Vague or unattributed claims
- No named human behind the words
- Repeated structure within the same page
- Absence of verifiable, checkable detail
A site that failed none of the five scored Very Low. A site that failed most or all of them scored High. We also read one additional site, a named client of ours, as a benchmark for contrast rather than as part of the sample itself.
What separated high and low scores
Not budget. Two of the least slop-affected sites in the entire sample had no professional design at all: one was a small marina running plain, functional copy with real operational details, and the other was a regional media arts nonprofit with a distinct voice and specific, checkable facts about its own history. Meanwhile, one of the more polished, professionally built sites in the sample scored among the worst, filled with unattributed superlatives and a paragraph repeated twice, word for word, further down the same homepage.
The pattern held across the whole sample. The businesses that scored well were not the ones that spent more. They were the ones that said something only they could say.
The "nestled in" problem
A version of the phrase "nestled in the heart of the mountains" or "nestled in the mountains" appeared on three unrelated homepages in this sample, a food business, a lodging business, and a tourism board, each in a completely different part of the audit. None of the three businesses have any relationship to each other. It is the closest thing this sample had to a signature phrase, and it is a near-perfect example of the core definition from our first piece: language that could describe almost any business in a mountain region, attached to none of them specifically.
In at least one case, the business sitting behind that generic opening line had a real, specific, documented founding story, an immigrant founder with a traceable personal history, that never made it onto the homepage at all. The real story existed. The website just did not tell it.
What the worst-scoring site got wrong
The single highest-scoring site for slop in this sample repeated the exact same paragraph twice on its own homepage, once near the top and again further down, word for word. It also carried an unattributed claim along the lines of being the regional best, with no source, no year, no named authority behind it. Between the repetition and the unsupported claim, it failed four of the five criteria on its own.
What the best-scoring sites got right
The two lowest-scoring sites for slop shared nothing in terms of budget or industry, but they shared a habit: they named real things. Real names, either of a founder or of a specific person tied to the business. Real numbers, whether a year, a headcount, or an operational detail a competitor could not casually claim. And a voice that read like one specific person had written it, not a template built to fit any business in the category.
One of the two had a documented family history behind the business going back multiple generations, with real names and real photographs tied to a specific historical event. The other operated out of unglamorous, specific circumstances, described plainly, with no attempt to dress the location up into something it was not.
Where our own client fits
We also read the homepage of a named client of ours, Powell Construction, a custom mountain home builder on the Highlands-Cashiers Plateau in North Carolina, as a benchmark rather than as part of the twelve-business sample. It scored at the very low end alongside the best performers above: real pricing figures, a real described process, and no filler language standing in for substance. We are not neutral about that site since we built it, which is exactly why it is presented separately from the anonymized sample rather than folded into it.
What surprised us
Going in, the working assumption was that professional design and marketing spend would track with lower slop scores. That assumption did not hold. The strongest predictor in this sample was not who built the site or what it cost. It was whether the business was willing to be specific about itself, on its own homepage, in its own words.
What to do with this
Read your own homepage against the same five patterns. Ask one question of every sentence: could this describe a competitor just as easily? If yes, it is slop, and the fix has nothing to do with budget. It has to do with naming the real, specific, checkable things that only your business can claim.
Frequently asked questions
Why did you anonymize the businesses in this audit?
To keep the focus on the pattern, not on any single business. Every site in this piece is described by category and pattern, not by name.
Does a bigger marketing budget prevent AI slop?
No. In this sample, budget did not predict the score. Two of the least slop-affected sites had no professional design at all, and one of the more polished, professionally built sites in the sample scored among the worst.
What was the single most common slop pattern in this audit?
A version of the phrase "nestled in the mountains" or "nestled in the heart of" appeared on three unrelated businesses' homepages, each in a completely different industry. It is the closest thing this sample had to a signature phrase.
Can I get my own website audited this way?
Yes. The five-pattern framework from our first piece on AI slop is the same rubric we used here, and it applies to any homepage. Read your own copy against it and see how it scores.