Home / Case studies / Building SEO from zero for a US home services marketplace

Building SEO from zero for a US home services marketplace

At a glance

• Company: HomeBuddy, a US marketplace that connects homeowners with local contractors. Part of QuinStreet since 2026.
• My role: I lead SEO, from strategy and site architecture to specs, launches and analytics.
• Period: February 2024 to now.
• Result: 6 page types built from zero, about 6,000 SEO pages that bring traffic, leads and revenue. Google impressions grew 1,500% in the first 9 months.

The challenge

When I started, the site had no SEO at all. The only landing pages were built for paid traffic. They ran as A/B tests, were closed from the index and could not be changed beyond meta tags. There were no SEO pages, no content process for search and no technical base for local pages. The data we needed was limited too. We had cost data, but for contractors we had little beyond names and services. Our competitors had far larger contractor databases.

The task was to build an organic channel for a national marketplace from this starting point, without breaking the paid funnel that already worked.

What I did

Research and architecture. I ran the technical audit and studied the niche to see which page types exist in it. Based on this, I designed the site structure for thousands of keywords and defined the page types we needed.

Launch order by readiness. I chose the order of page types by what we could actually build and support, not only by traffic potential:

1. Cost guides came first. We already had cost data, and they needed the least development.
2. Contractor profiles came next. They are the foundation for directories.
3. Geo directories were built on top of the profiles.
4. Near-me landing pages were reworked from existing pages instead of being built from zero.
5. Three cost calculators launched in early 2026.
6. The Cost Index launched in July 2026.

Running the process. SEO at HomeBuddy is a cross-team process, and I run it. I did all the analysis and wrote all the specs for development, design and content. I set the priorities, split the tasks and drove each page type through design and development to launch. I also built the content process from scratch and managed three copywriters working on my briefs.

Quality over volume. The first rollout of geo directories created about 20,000–25,000 pages. Many of them had only one or two contractors and looked almost the same. I set a rule: no directory page for a service and city with fewer than three contractors. The site went down to about 6,000 pages, and each remaining page has real content behind it.

Adapting to search changes. By 2024–2025, Google results for near-me queries became local. A homeowner in Texas who searches for "walk-in shower near me" sees local directory pages and Texas service pages. Our general near-me pages looked strong in Ahrefs and Semrush, but they brought far less traffic than we expected, because homeowners search from their own cities. This shaped the next step: geo directories that also work as local landing pages.

First-party data. There is a lot of public information on what a walk-in shower costs. I wanted to show our own numbers. I built it from our own request data, comparing it across states and by the answers users give. The result is the HomeBuddy Cost Index, based on 9,171 real project requests from 49 states. Without any keyword optimization, it became the 14th most cited page of the site in Bing AI search soon after launch.

Results

• 6 page types built from zero. All of them bring traffic, leads and revenue. Cost guides lead in traffic, near-me pages lead in conversion.
• About 6,000 SEO pages, after removing thin pages.
• Google impressions grew 1,500% in the first 9 months.
• SEO pages work for paid traffic too. 35% of leads and revenue on SEO pages come from organic and direct visits. The other 65% are assisted conversions: users who first landed on paid pages, did not convert there, kept exploring the site and converted on SEO pages.
• The site is cited in Microsoft Copilot and appears in Google AI features, and this visibility keeps growing.

What's next

• Rebuilding geo directories into fewer, richer pages that are half landing page and half directory. Each will show our own data: what a service costs in this city, which options homeowners choose and why.
• More research based on first-party data.
• More focus on AI search: external sources, reputation and customer feedback.

What I learned

Match the number of pages to the data you have. Fewer pages with real, unique information work better than thousands of similar ones.