For the full working ledger — the complete plan, hypothesis backlog, and decision log — see Search Strategy (full).
How we get found by homeowners who don't already know us — in three parts: the plan, what we're testing right now, and the work that never stops.
Search engines — and now AI tools like ChatGPT — used to reward whichever page had the most words on it. That stopped working.
Today they reward pages that prove real, local expertise: actual job photos from that city, reviews from people who live there, answers to the exact questions a homeowner there would ask.
So that's the plan: rebuild our city pages to that standard, one city and one service at a time — and prove it actually works in a handful of cities before we finish the rest.
Three efforts, running at the same time, each answering a different question.
Gave 5 cities a full local-detail rebuild — real photos, named neighborhoods, real Q&A — and left 7 similar cities untouched, as a comparison.
If the 5 rebuilt cities climb more than the 7 untouched ones, the rebuild is what's working. Started Aug 19; rankings take about two months to settle, so we won't call it before Oct 14, 2026.
Still moving through the same rebuild, just on their own timeline — outside this specific comparison.
Instead of writing about whatever sounds good, we now check what people actually search for — and whether we can realistically win that search — before writing anything. First test: everything concentrated on one area, basement remodeling, plus the basement half of the test above. So far: one new post built around a competitor's gap (shipped), two existing posts sharpened, and — since that same rival's basement guide is out-ranking us in a couple of specific spots — two more angles queued, one already scheduled and one waiting its turn, plus testing whether a "best basement companies in the area" comparison post works the way it has for at least one other local competitor.
Shipped Aug 25–26. We'll check each page's search position and clicks against its numbers from right before, about 3–4 weeks later. We're deliberately not piling on more basement posts before those results come in — the point is to prove the method works on one focused area before spreading it to kitchens and bathrooms, not to chase this one topic indefinitely.
Combined Lakeville's four separate pages into one — real neighborhoods, FAQs, and photos in one place. Live since Aug 30.
A different kind of test than the one above — just Lakeville's own numbers, before vs. after. We'll compare visits and clicks about 2–3 weeks post-launch. If it holds or improves, the same format rolls out to the next 5 cities.
Lakeville sits outside the test above on purpose — its rebuild changed too much at once to isolate what caused any lift.
Regardless of how the tests turn out, this work doesn't stop:
People who already know Honey-Doers find us easily. This whole strategy is about the people who don't — real, local pages instead of generic ones, tested before we commit to the rest, backed by content that never stops publishing.