Three claimed listings.
One of them is doing all the work.

All three of your Google listings are verified, described, photo-stocked, and pointed at their own city landing page. This is a deliberately managed three-location setup, not a case of forgotten duplicates — and that matters, because it means the fix is positioning, not cleanup.

Claim status, addresses and phones fetched live 2026-07-20. Top-3 cells from grid scan bb_20260717_0351, 2026-07-17 — 1,544 map points searched against 10 keywords, 1,538,368 result rows. A "top-3 cell" is one map point where the listing lands in Google's top three. It is the same measure the live map paints green, so these reconcile with it exactly.

The likely structural reason

Two of your three addresses are shared, character for character, with another remodeling firm. Google reads a shared suite string as a signal about which business belongs at that address, and the listing with more corroboration behind it tends to win the spot.

Both address lists are exact-string matches from the 2026-07-17 scan. Five Star's 115 G9 is the tell — suite 115 is itself subdivided into lettered units, which is the signature of a shared workspace rather than a private office. These are address strings as Google publishes them, not a verified site visit.

Google uses the phone number as one of its main ways to confirm that a listing and a website are the same business. Right now every number you own is a single-source number — it appears either on the site or on a listing, never both.

Page counts from the 2026-07-16 crawl of 331 pages. Listing numbers fetched live 2026-07-20.

The site also runs CallRail, so the number a visitor actually sees is swapped dynamically by traffic source. That is a legitimate attribution setup and worth keeping — it is why a hand-check can turn up a number that exists in no page source at all. The issue is only that none of the three listing numbers has a second home anywhere.

Directory listings are the cheapest corroboration there is: each one is another independent source telling Google that BB-Builders Pro is a real contractor at a real address with a real phone number. You already hold 11 of them. 8 are unclaimed.

Checked 2026-07-16 via directory SERP lookups. This ties directly back to sheet 02: every listing you claim is one more place your address and phone corroborate each other, which is the signal Bellevue and Seattle are thinnest on.

Your domain carries 430 backlinks from 141 referring domains, at domain rank 211. A slice of that profile has the fingerprint of a bought link package — almost certainly a past vendor's shortcut rather than anything anyone chose deliberately.

33 of the top 100 referring domains carry a spam score of 20 or higher. Measured 2026-07-16. Shortlist for review, not a finished disavow file — each one gets eyeballed before it goes in.

21 of the inbound links are broken, pointing at URLs that no longer resolve. Those are recoverable value: redirect them and the equity lands somewhere useful instead of a dead end.

All from the 2026-07-16 crawl of 331 pages.

The single cheapest fix on this page

664 internal links point at a page that isn't there — and 660 of those are a single URL, /expert-home-extension-services/, sitting in a sitewide template so it repeats on essentially every page. (That is the 660 figure from the letter; the remaining 4 point at a second dead page.) Both are still listed in your sitemap. Every one of those links currently spends your site's own authority on a dead end. Repointing one template link recovers 660 of them in a single edit.

The content is working. It just isn't landing anywhere that sells.

Your blog earns 291,320 monthly searches worth of keyword visibility across 262 posts — 262 published since 2024 and still going. Your 60 sales pages earn 6,890, and most of that is your own brand name.

That gap is the whole story of the site. Someone is genuinely doing the work — publishing consistently, earning reviews week after week. Earning reviews and publishing articles is one trade; wiring a website so the traffic reaches a page that sells a remodel is another. The effort is real. The plumbing it flows into was never built to catch it — and of the two, the plumbing is far cheaper to fix.

The city pages haven't been given their own voices yet

134 page pairs are at least 60% identical. Redmond, Tukwila and Federal Way run 99.4% the same body text; /edmonds/ and /location/edmonds/ are 100% identical. Google has to pick one page to rank for a given city and, faced with near-copies, tends to rank none of them. Giving each city page real local content is what unlocks them.

Your best salesman is sitting backstage

20 pages can't be reached from your homepage at all — including the entire /projects/ portfolio. For a remodeler that is the most persuasive asset on the whole site, and right now neither a visitor nor Google can navigate to it.

Across the whole Puget Sound remodeling grid you are #12 of 949 contractors holding any top-3 map presence, on 499 top-3 cells. That is a genuinely strong position and it is not an accident. It is also concentrated in a way that's worth seeing plainly.

Sorted by market value (searches x cost per click). Grid scan bb_20260717_0351, 2026-07-17. Volume and cost-per-click from keyword research, 2026-07-16. Coverage is your top-3 cells as a share of all 1,544 map points.

One thing this grid can't tell you: it never measured decks or ADUs. Those weren't in the ten keywords scanned, yet they're 15-plus pages of your site and you already rank organically for TimberTech and DADU terms. Whatever the map says today, it is silent on two things you visibly sell.

This is not a neglected business or a neglected site. Nearly everything on the previous six sheets is wiring, and wiring is the cheap part. Here is what is already built and can't be bought.

Reviews pulled 2026-07-23; blog counts and technical scores from 2026-07-16.

Technical hygiene is genuinely good: no missing H1s, schema on virtually every page, valid sitemap and robots, SSL and HTTP/2, a median of 1,325 words per page. The stack is maintained. A reputation of 108 reviews at 4.9 stars is the single hardest thing to acquire in this trade, and you have it.

What we are not claiming. Nothing on this page says anything about your leads, your calls, or your revenue — we have no access to your analytics, your CRM, or your call logs, so conversion simply isn't visible from outside. What is measured is narrower and more useful: three assets that should be producing search visibility, and the specific places each one stops short of a page where a remodel gets sold. If CallRail is being monitored, you already hold the data that would show whether Bellevue and Seattle generate a single call — and that would be the first thing worth looking at together.

1. Repoint the dead template link

   Recovers 660 internal links in a single edit, and takes the dead URL out of your sitemap. Hours, not days — and it is the highest ratio of result to effort anywhere on this list.

2. Put each listing's phone on its matching city page

   Three lines of markup. Gives Bellevue and Seattle back the corroborating signal they are missing, without touching your CallRail attribution.

3. Claim the 8 open directories

   Plus claiming the BuildZoom listing you already appear in. Each one is another independent source confirming the same name, address and phone — which is precisely what the two quiet listings are short of.

4. Give the city pages their own content

   134 page pairs are 60%+ identical, which is why none of them rank. Real local content per city, starting with the cities where you actually want the work.

5. Surface the portfolio

   Link /projects/ from the homepage and the navigation. 20 pages currently can't be reached at all, and your best sales asset is among them.

6. Rebuild the service pages against terms you can actually win

   59 keywords already sit at positions 11–30 with 300+ monthly searches — one refresh from page one. That is where the 54 silent sales pages come back.

7. Then, and only then, the map

   Re-scan against keywords that include decks and ADUs, and revisit the satellite addresses before any more money goes into them. This one waits deliberately — it is the most expensive item and the least urgent.

Ordered by result-per-hour, not by size. The first three are days of work between them and they address the listing problem directly.

Nothing here is an estimate. Each figure was produced by a measurement on a stated date, and the older ones say so.

Site-health figures are from 2026-07-16. Listing, address and phone figures were re-verified live on 2026-07-20. The field size of 949 contractors comes from putting all 6,624 businesses in the scan through a classifier rather than trusting Google's category label — counting every business type instead, the field is 1,001 and the position is the same, #12.