How AgentSiteScore Works
Our reports are generated by an automated scanning system that collects publicly available data about a real estate agent’s online presence and scores it against a fixed rubric. This page explains exactly what we measure, where the data comes from, and how we weight it. If you believe a report about you is wrong, the correction process is at the bottom.
What we measure
Every scan evaluates seven categories. Each is graded A–F independently, then rolled up into the overall 0–100 score.
1. Website fundamentals
- Page loads and responds to HTTP requests
- Contains a discoverable headline and bio
- Has contact information (phone, email, or form)
- Mobile-responsive layout
- HTTPS enabled, valid SSL
- Basic on-page structure (title tag, meta description)
Source: direct fetch of the public URL.
2. SEO & visibility
- Branded search ranking (does your name + city find you?)
- Market keyword ranking (do generic "real estate agent [city]" searches find you?)
- Title tag and meta description quality
- Presence of structured data (JSON-LD schema)
- Robots.txt and sitemap.xml reachability
Source: Google search results pages (live queries), direct fetch of robots.txt and sitemap.xml.
3. Online reputation
- Review counts and average star ratings on Google Business, Zillow, Realtor.com, and Yelp
- How many of those four platforms are verified as actively used
- Recency of the most recent review where obtainable
Sources: Google Places API (Google Business), public Zillow/Realtor.com/Yelp profile pages. Each platform is fetched independently; if one source is unreachable (blocked, rate-limited, or the profile can’t be located), it is marked "not found" rather than assumed absent. Review data may be stale by up to 30 days depending on when the source was last crawled.
4. AI search readiness
Can Perplexity, ChatGPT-with-search, and Google AI Overviews surface you when someone asks about real estate agents in your market? This is a forward-looking signal and relatively new.
- Presence of FAQ / Q&A content on the page
- Consistent name / phone / address across the site
- Structured data markers that AI crawlers rely on
- Real-estate-specific schema (RealEstateAgent, LocalBusiness)
Source: direct page analysis. We do not query AI models directly as part of scoring.
5. Social presence
- Linked profiles on Facebook, Instagram, LinkedIn, YouTube, TikTok, X (Twitter)
- Each link verified as active (not broken, not defunct)
Source: link extraction from the page + HTTP HEAD request on each target to detect dead links.
6. Competitive positioning
- Your position vs. the top agents in your local market
- Google Maps local pack presence
- Review count gap vs. the market leader
Source: live Google search for "real estate agent [detected market]" queries, Google Places nearby search.
7. Headline & bio
- Headline contains market (city/state)
- Headline has a value proposition beyond just name + role
- Bio is specific (numbers, credentials, specialty) vs. generic
- Social proof present (awards, designations, sales volume)
Source: direct page analysis against a rule-based rubric.
How the overall score is calculated
Each category produces a 0–100 sub-score. The overall 0–100 score is a weighted average:
- Website fundamentals: 15%
- SEO & visibility: 20%
- Online reputation: 25%
- AI search readiness: 10%
- Social presence: 10%
- Competitive positioning: 10%
- Headline & bio: 10%
Grade buckets: A (85–100), B (70–84), C (55–69), D (40–54), F (below 40).
The rubric is deliberately tough. Most agents score 45–65 on their first scan. That’s intentional — a score of B+ should actually mean something.
What the score is NOT
A low score is a signal that an agent’s public-facing digital presence has fixable gaps. It is not:
- A judgment of the agent’s competence, character, or business practices
- A statement about the agent’s actual sales volume, revenue, or client satisfaction
- A ranking of "best agents" in any market
- A substitute for hiring or referring decisions
Agents with excellent websites can be poor at their job. Agents with weak websites can be outstanding at their job. We measure websites, not people.
Known limitations
- Third-party data is only as fresh as its source. If a platform hasn’t been re-crawled recently, review counts may lag reality.
- Some sources occasionally block scanning. Review aggregators sometimes throttle or challenge our scanner. When a source is unreachable for a given scan, that platform is marked "not found" for that run rather than guessed at — re-scanning after a short wait usually resolves it.
- Market keyword ranking is single-query. We run one representative market query; agents with niche specialties may rank well for queries we don’t check.
- Score changes over time. Re-running a scan tomorrow may produce a different score as the rubric evolves or third-party data shifts. The date on every report shows when that particular snapshot was captured.
Requesting a correction
If you believe a specific finding about your site is wrong — not a matter of opinion about the rubric, but a factual error — we want to fix it. Email chris@agentsitescore.com with:
- Your name and the URL of the site that was scanned
- The specific finding you believe is incorrect
- What the correct information is (e.g. "I have 47 Google reviews, not 12" — include a link so we can verify)
We respond within 3 business days. If the correction is valid, we’ll re-scan your site and issue a corrected report. If the score is correct per our rubric but you disagree with the rubric itself, we’re always open to feedback — email us with your thinking and we’ll consider it for future rubric updates.
Full Terms and Privacy
For the legal framework — what you agree to by using the Service, how we handle your data, refund policy, and more — see our Terms of Service and Privacy Policy.