Corpus Plan

Learn

What Claude finds when a small business asks for a web designer in Columbus, Ohio

Claude API with web search, 3 runs, 2026-11-03 21:47–21:52 UTC. A snapshot, not a ranking.

We asked Claude the same question three times on November 3, 2026, through the Claude API with web search turned on. It searched all three times, with the same search words, and got back the same 8 domains each time. Two of them were web design agencies' own homepages. Both agencies were named in every answer, in a list Claude said came from DesignRush.

  • Question, verbatim: "Who is a good web design company for a small business in Columbus, Ohio?"
  • Model and settings: claude-sonnet-5-5 through the Claude API, effort low, web search tool web_search_20260209 limited to 1 search per answer, dynamic filtering on (the default), max_tokens 4000, no system prompt.
  • Run times (UTC, 2026-11-03): 21:47:05, 21:48:50, 21:50:37.
  • Search Claude ran, all 3 runs: "best web design companies small business Columbus Ohio"

Which websites did Claude's search return for web designers in Columbus?

The same 8 domains in all 3 runs. Three were named in every answer: the DesignRush directory and two agency homepages, splitreef.com and marcy.com. The other five were directories, or agencies' own "best web design companies in Columbus" posts, and Claude did not name them.

DomainPage title, as the search returned itIn the search results (runs of 3)Named in the answer (of those runs)
designrush.comTop Columbus Web Design Companies; Top Ohio Website Design Companies (two pages)33
splitreef.comColumbus Web Design Company - Website Development33
marcy.comColumbus Ohio Website Design from Marcy Design. Call Us!33
loungelizard.comWeb Design Companies Columbus30
goodfirms.coTop 30+ Ohio Web Design Companies - 2026 Reviews30
columbuswebseo.comBest Web Design Companies in Columbus, Ohio (2026): An Honest Guide30
exaltodesign.comTop Web Design Companies in Columbus, Ohio in 202630
expertise.com20 Best Columbus, OH Web Designers30

"Named in the answer" means the site's or business's name appears in Claude's answer text. The API returned no citation data in these runs, so neither column measures citations.

In its answers, Claude said its list came from a DesignRush directory listing. Its code printed only the first 2 search results, both DesignRush pages, and every agency it named, Split Reef and Marcy Design Group included, appears in that printed DesignRush text. Across the three runs it also named Think Web, Studio 68 Web Design, The Media Captain and Red Gecko. We list them as a record of what the answer said. They are not our recommendations.

What do the listed pages have in common?

Every title names the place: all 9 say Columbus, Ohio or OH. 6 of the 9 say "best" or "top", and 3 carry the year 2026. Of the four questions Claude searched, this was the only one where a business's own homepage was in the search results. The two homepages, splitreef.com and marcy.com, both put "Columbus" and "Web Design" or "Website Design" in the title.

The other pages are lists. Three are directories: DesignRush, GoodFirms and Expertise. Three are posts in which an agency lists Columbus web design companies: loungelizard.com, columbuswebseo.com and exaltodesign.com. All three posts were in the search results in all 3 runs. None of the three agencies was named in an answer.

What could we measure on those pages?

6 of the 9 URLs served us the page. 5 of those 6 had JSON-LD structured data, and 5 had at least one heading phrased as a question. Both DesignRush pages and the GoodFirms page answered our server with HTTP 403 and are not measured. We fetched each URL from our server on 2026-11-03 between 21:53 and 21:56 UTC, with a Chrome user agent, and read the HTML as served before JavaScript ran.

DomainWhat our server gotJSON-LD typesQuestion H2/H3robots.txt: Claude-SearchBotrobots.txt: ClaudeBot
designrush.com (2 URLs)HTTP 403; not measurednot measurednot measuredallowedallowed
splitreef.compage (200)none18allowedallowed
marcy.compage (200)Organization0allowedallowed
loungelizard.compage (200)Article, FAQPage, Organization1allowedallowed
goodfirms.coHTTP 403; not measurednot measurednot measuredallowedallowed
columbuswebseo.compage (200)Article, FAQPage, Organization1allowedallowed
exaltodesign.compage (200)Article, Organization3allowedallowed
expertise.compage (200)AggregateRating, LocalBusiness, Organization, Review20alloweddisallows /

The homepages of the two agencies named in every answer differ on both measures. splitreef.com had no JSON-LD and 18 question headings. marcy.com had Organization JSON-LD and no question headings. robots.txt allowed Claude-SearchBot on all 9 URLs. expertise.com's robots.txt disallows / for ClaudeBot. Anthropic says a ClaudeBot restriction signals that a site's future materials should be excluded from its model training data.

Did Claude search the web every time?

For this question, yes: Claude searched in 3 of 3 runs. That does not hold for every question. Anthropic's web search documentation says "Claude determines when to search based on the prompt." When we asked who is a good local SEO company for a small business in Columbus, Claude searched in 0 of 3 runs, named no agency, and pointed to Clutch and UpCity.

Across all 10 buyer questions in this study, Claude searched in 10 of 30 runs. The questions it did not search are on the hub page.

Does your services page answer the question a buyer asks?

Check whether a page on your site has a heading phrased as a buyer's question. For example: "Do you build websites for small businesses in Columbus?" Then check that the first sentence under it gives the answer in 40 words or fewer. 5 of the 6 pages we could read in these results had at least one heading phrased as a question.

We did not test whether this changes what Claude finds. It is the shape the free Corpus Plan audit's answerability check looks for: H2/H3 questions plus FAQPage entries, each opening with a first sentence of 40 words or fewer. The audit covers five areas: SEO, AI citability, speed, trust and content. Its AI-crawler rules separate answer crawlers such as Claude-SearchBot from training crawlers such as ClaudeBot. The checks never ask a model.

Free preview with no account · free account for the full report · up to 20 pages per audit.

Run the free audit on your site · How the audit works

What this does and does not show

  • One question, one city, one model, one day: three runs between 21:47 and 21:52 UTC on 2026-11-03. Ask again and the results may change.
  • "In the search results" means the URLs the search returned. Anthropic's documentation says that with dynamic filtering, Claude's code filters the search results before they reach its context window. In all 3 runs, the code tried to print 6 results, hit an error, and then printed the first 3,000 characters of the results. That text held only the first 2 results, both DesignRush pages. The record does not show Claude reading the other 7 pages in the table.
  • The API returned no citation data, so nothing here measures citations.
  • This is the Claude API. We did not test the Claude.ai app.
  • We measured a state, not a cause. Nothing here shows that a homepage title, a JSON-LD type or a question heading makes a site more likely to be found or named.
  • 3 of the 9 URLs answered our server with HTTP 403 and are not measured.
  • It is not a ranking of agencies or of websites.

Same study, other trades: roofers · landscapers · plumbers · all 10 questions

Related: What ChatGPT and Google cite: three measured studies · Does my website show up in ChatGPT?