How this list was built
SF SEO scores the websites of funded startups in San Francisco for SEO.
Which companies are listed
The list comes from a company search in Context. It includes active companies headquartered in San Francisco, founded in 2005 or later, whose funding stage is seed, Series A or Series B. Company facts such as stage, funding, employees and founding year come from that same source.
Each company was then given a category, a short description, a rating for how much its buyers rely on search, and six keywords a buyer might type. These were generated from the company description by an AI model.
What we measured
- The site itself. We loaded the homepage and read its title, meta description, H1 headings, canonical tag, Open Graph tags, structured data and image alt text. We read robots.txt and llms.txt. We read the XML sitemap and counted pages by type, such as blog posts, docs, comparison pages and customer stories.
- Search traffic. Estimated monthly organic visits from US Google search and the number of ranking keywords come from DataForSEO.
- Keywords. Monthly US search volume and keyword difficulty for each suggested buyer keyword also come from DataForSEO.
How the scores work
Every score runs from 0 to 100.
Strength
What the site earns from search today. Estimated organic traffic counts for most of it. The number of ranking keywords, the number of content pages in the sitemap and the technical score make up the rest. Traffic, keywords and pages are counted on a log scale, so going from 100 to 1,000 visits matters as much as going from 1,000 to 10,000. Each stops counting at a ceiling: 100,000 visits a month, 50,000 keywords and 2,000 content pages.
Technical
Starts at 100 and loses points for each problem found. A critical problem costs the most, a warning less, and a minor note a little. A site we could not read gets no technical score: one that refused our automated check, one whose domain redirects to another site, and one that did not load or returned an error. Strength counts this part at a middle value for them.
Demand
Whether there is search demand the company could capture. Most of it comes from the monthly search volume of the suggested buyer keywords. Each keyword is capped so one very broad term cannot carry the score, and keywords with buying intent count for more than informational ones. The rest comes from the traffic earned by the top 10% of sites in the same category. The result is scaled by search fit: a product whose buyers rarely start with a search keeps only a small part of its demand score.
Ease
100 minus the average keyword difficulty of the buyer keywords that have measurable search volume. If none of them have difficulty data, ease is set to a middle value.
Opportunity
Demand that is not captured yet. It starts from the demand score, shrinks as strength rises, and shrinks further when the keywords are hard to win. It is also scaled by how ready the company looks to act on it: more funding raised and a recent round or founding date count for more. Companies whose site did not load or returned an error get no opportunity score.
Segments
Each company is placed in one group, checked in this order.
- No site data
- The site did not load or returned an error when we checked. We have search estimates for it but could not read the site.
- Leader
- Strength of 70 or more. Already earns meaningful traffic from search.
- Growing
- Strength from 45 to 69. Has some search traction and room to build on it.
- Untapped
- Strength under 45 and opportunity of 42 or more, on a site with real content that carries the company's name, at a company with three or more employees and two or more keywords people search for. Little search traffic today, but people search for what it sells and the keywords are within reach.
- Low search fit
- Strength under 45 and a low search-fit rating. Buyers of this kind of product rarely start with a search, so SEO matters less here.
- Early
- Everything else. Little search presence so far.
Fixes
The fixes on each company page come from simple checks on the homepage, robots.txt and sitemap. The suggested text is built from the company description.
Limits
- Traffic is a modeled estimate, not analytics. It is built from the keywords a site ranks for and how often people click each position. Real traffic can be much higher or lower, above all for small sites.
- Keywords are suggestions. A company may sell to buyers who search in very different words.
- Company data can be wrong. Stage, funding, category and even the website can be out of date or mismatched.
- The check is a snapshot. We looked at each site once, from one location, without running JavaScript. Some sites refuse automated requests or redirect to another domain. They keep their search scores and have no technical score or fixes.
- A low score is not a verdict on a company. Many strong businesses do not need search traffic at all.
Corrections and removal
If something about your company is wrong, or you want it removed from the list, email support@openseo.so with your company name and domain. We will fix or remove it.