iProv ×
Client Strategy Report

Texan Heating & Air · Fort Worth, TX

Built to be found by service, not just by name.

iProv's local-search and location-page strategy for Texan Heating & Air. THA already has the trust, referrals and reviews. What it doesn't have yet is a website that shows up when a Fort Worth homeowner searches for the job, not the name.

Prepared by iProv · August 2026 · Internal case study, identifying details of a referenced comparison client have been anonymized throughout

4,180
Monthly searches, Fort Worth HVAC terms
16
Towns in THA's service area
60%
Minimum uniqueness score to publish, machine-checked
92
City + service pages that earned their place

The problem with the obvious approach

The standard playbook for a multi-town service business is a page for every service in every city. It looks thorough in a sales deck. It usually doesn't work, and when it fails it fails quietly, the pages sit in Google's index racking up impressions from bots and rank-trackers while never earning a real customer.

Before recommending anything to THA, we tested that assumption against 16 months of real performance data from a comparable multi-location build. The result: single, well-built pages consistently outperformed templated location pages by roughly eight to one, per page. Over half the templated pages had never earned a single click.

So building 90+ pages was still the right call, once we changed how

The lesson wasn't "build fewer pages." It was "don't build a page that has nothing true and specific to say." Every town in THA's service area has a different housing stock, a different heating-fuel mix, a different build era, all public record. We built a scoring system that uses that data to decide, cell by cell, whether a city-and-service combination earns a real page or gets left as a line item. That turns the same page count everyone else gets wrong into something that actually holds up.

How the plan came together

01

Audit

Pulled real performance data before recommending anything. Crawled THA's current site and found the foundational gaps holding every page back.

02

Score

Built a per-town, per-service scoring grid from Census housing data and gas utility records, so every page built has something genuinely local to say.

03

Sequence

Fort Worth first, heating pages before cooling pages to catch the February peak, everything gated on an automated uniqueness checker before it goes live.

Read the plan

Start here
Fort Worth First
The client-facing narrative: what we found, what we're building first, and why. The best single document if you're reading one.
 The audit
Location SEO Plan
Where the plan started. 16 months of real Search Console data from a comparable build, a full crawl of THA's current site, and the case for why thin location pages fail.
 The method
Page Qualification Grid
The scoring system. 16 towns by 8 services, scored against build era, heating fuel and permit data, deciding which of 128 possible pages actually earn their place.
 The architecture
Matrix Architecture
The full technical case for the structure: URL shape, internal linking rules, and the real cost, in pages and unique words, of building it properly.
 The sitemap
Site Structure Map
Visual map of THA's site today against the rebuilt structure: what's kept, what's rebuilt, what's new.
 The rollout
Full Location Map
All 16 towns, phased by search demand and evidence, with every scored page and every deliberately skipped one shown.
 The build
Fort Worth Build Brief
The tactical layer: title tags, meta descriptions, target keywords and required local facts for every page in phase one, ready to write from.
Where it stands

Phase one is scoped and sequenced. The first pages built against this plan scored 97 to 100% unique on the automated checker before publishing. Full build-out proceeds in phases, measured at each step rather than committed all at once.

Prepared by iProv for internal and prospective-client reference. Search volume and seasonality from DataForSEO. Heating fuel and median build year from US Census ACS 2024 5-year estimates. Performance figures describing what doesn't work are drawn from a real, comparable multi-location build; that client's identity and site have been withheld and all identifying details anonymized to protect their confidentiality.