What can a five-city study rank for foundation repair, and what can't it?
It can show which records ChatGPT read when it checked foundation companies, and where foundation repair stood apart from six other trades we checked the same week. It can't show that any idea brings calls or signed repairs, or rank one idea's return against another's. Ideas with the clearest standout come first.
The study is ours. In Atlanta, Austin, Charlotte, Houston and San Antonio we kept the top three on Google Maps and the local pack's listings (three in four cities, two in Atlanta) for "foundation repair," ran five ChatGPT runs per city that named no company, and collected 150 ChatGPT answers about named foundation repair companies. Everything was collected on October 18, 2026, through the OpenAI API with web search, and it's reported separately from our 50-metro study of eight other trades.
Its limits matter. Three of the five cities are in Texas, where expansive clay drives much of the work, so the government sites cited lean Texan. The numbers are small counts from one snapshot, not rates, and Reddit or the BBB sitting beside the companies an answer named doesn't prove either one got them named.
So the order is: the one record where foundation repair stood far apart, Reddit, then three ideas built on published BBB, federal and state program pages, widest reach first, then practice. The four foundation repair marketing pages we read on November 1, 2026, were agency service pages; none ranked its ideas, cited a source for weather timing, or mentioned the BBB or Reddit.