Site SEO

AI answers and search rankings share a lot of plumbing: the engines that answer questions still find pages through search indexes, and a page Google cannot crawl is a page few assistants will quote. Site SEO covers that classic side. It crawls your site for on-page and site-wide issues, tracks where you rank on Google for the keywords you choose, and — on plans with SEO data — researches the keywords people actually search for.

#Where to find it

In the sidebar, open Site › Site SEO. It has three tabs: Crawl, Rank tracking and Keyword research.

#Crawl

#Start a crawl

1. On the Crawl tab, press Crawl the site.
2. The crawl starts from your brand's domain and runs in the background. Refresh the page after a minute or two; a large site can take a few minutes.

The crawler reads your robots.txt and obeys it, including any crawl delay. It starts at the home page, follows internal links breadth-first, and also visits the URLs in your sitemap, so pages that nothing links to are still found. One crawl runs per workspace at a time. If a crawl never finished, the page says so and you can simply start a new one.

How many pages one crawl may fetch depends on the plan:

See Plans and limits for the full comparison.

#Read the results

Four tiles sum up the latest crawl: SEO score (with when it ran), Pages crawled, Errors and Warnings.

The SEO score runs from 0 to 100. Every issue costs points according to its severity (an error weighs 5, a warning 2, a notice 1) multiplied by the number of pages it touches; the total is divided by the number of pages crawled, so a large site is not punished just for being large.

The Issues list is sorted by severity and then by the number of pages affected. Each line names the issue, how many pages it touches, and the first few addresses (hover to see them all). Issues the crawler looks for:

#Core Web Vitals

For a few representative pages, Rixe fetches Google's PageSpeed Insights results: Performance, LCP, CLS and INP. The last column says whether the numbers come from real users (Chrome field data, when the page has enough visitors) or from the lab. If PageSpeed could not be reached or its quota ran out, a note says so; the rest of the crawl is unaffected.

#From crawl to action plan

The latest finished crawl feeds the action plan. Each issue becomes one item in the We prepare it, you apply it pile, weighted by how much it affects ranking — a dead page or a missing title outranks an image without alt text — and by how many pages it touches.

#Rank tracking

#Add keywords

On the Rank tracking tab, type keywords — one per line or separated by commas — and press Add. You can also add them from Keyword research (below). The number of keywords a workspace can track depends on the plan.

#Where positions come from

• Search Console (free). If Google Search Console is connected, the position for each keyword you already rank on comes from Google's own data: the average position over real impressions, day by day. Connect Google under Connections — see Integrations.
• SERP check (paid). On Leader and agency plans, each keyword has a Check (paid) button. It looks up the live Google results for your market and language and returns your position in the top 20, plus the domains ranking above you. A keyword can be checked at most once a day; a second press the same day returns "Already checked today." The cost of each check is shown after it runs.

Each row shows the latest position and its source (Search Console or SERP check), the trend over the last 90 days (▲ is an improvement, ▼ a drop), and up to two competing domains. not in top 20 means a check ran and your site was not found; no data yet means no position has been recorded.

#From rankings to action plan

Keywords where you rank between 11 and 30 — close enough to page one that improving the page usually moves it — become items in the action plan. Positions 11–15 get a higher impact than 16–30.

#Keyword research

Available on the Leader plan and on agency plans.

1. On the Keyword research tab, type a seed keyword (for example hair transplant) and press Search (paid).
2. You get up to 100 related keywords with Monthly searches, Difficulty (green up to 30, amber up to 60, red above) and Competition.
3. Tick the keywords you want and press Track N keywords to add them to rank tracking.

Results are for your workspace's market and language. The answer for a seed is kept for a month, so searching the same seed again within that month is free — the page says "From the last month's cached answer — no cost." If the provider returns nothing, try a broader seed or one in the market's own language.

#Common questions

Why is the crawl showing fewer pages than my site has? The crawl stops at your plan's page limit, skips anything robots.txt disallows, and only follows pages on your own domain.

Search Console is connected but a keyword shows no data. Search Console only reports queries where your site already appeared. For a keyword you do not rank on yet, use the paid SERP check.

Does the crawl replace the Site health audit? No. Site health checks whether AI crawlers can access the site at all; Site SEO looks at the quality of every page for search engines. Run both.

#Related pages

• Site health
• Action plan
• Integrations
• Plans and limits