Data Integration

SEO data integration, joined page by page

Most SEO integrations stop at a chart. Vee3.ai joins the data on the unit you actually work with, the page. During a site scan it reads your Search Console page and query rows, GA4 organic metrics and the keywords SEO data providers see your site ranking for, together. Where Search Console and a provider both report the same page and query, Search Console's first-party evidence wins and the provider keeps what Google does not report, such as monthly search volume.

Units stay honest: Search Console clicks are never presented as estimated traffic, and GA4 metrics are Organic Search only. Tracking parameters such as utm_ tags are stripped before the join, so one page stays one row, and the Performance Sources column names what contributed.

Google Search Console and GA4, read-only

Vee3.ai asks Google for two read-only scopes, Search Console and Analytics, and nothing that can change your Google properties. Properties are matched to your domain, not their names. A Search Console domain property can cover your subdomains; a URL-prefix property must match the protocol, host and path. Unverified properties are left out. For GA4, Vee3.ai reads each property's web streams and picks the one whose URL matches your site, preferring an exact origin. If nothing matches, or several do, you choose from the list, and GA4 can come from a different Google account than Search Console.

Before a long scan starts, the GA4 connection is tested with a real reporting request, so a disabled API or a missing Viewer permission shows up while you are connecting, not halfway through a crawl. Tokens are encrypted with AES-256-GCM. Disconnecting deletes the stored credentials and revokes access at Google, and the data already imported to your roadmap stays.

What analytics integrations support SEO tracking?

Search Console gives the scan clicks, impressions, CTR and average position for each page and query over a 30-day window ending two days ago, plus the 30 days before it for the change columns. GA4 adds Organic Search sessions, key events, key event rate and revenue for each page, in your property's currency. On the SEO dashboard the same connections are queried directly for the last 30 days, 3, 6 or 12 months, with an optional previous-period comparison and query, page and country detail. Dashboard reads are cached for 15 minutes, and the stored scan snapshot stays on screen while a refresh runs.

Sitemap, crawl and SEO data providers

The scan finds pages from the sitemaps listed in robots.txt and the usual sitemap locations, following sitemap indexes to up to 10,000 URLs. The homepage is always read. When a site has no useful sitemap, a bounded internal-link crawl finds up to 500 pages instead. Each page is read for its status, title, meta description, canonical, robots directives, headings, schema types and links, which feed internal link counts and the technical checklist without a second crawl.

SEO data providers fill in what a crawl cannot see: search volume and difficulty, the organic keywords a domain ranks for, the live top 10 for a keyword, backlinks and referring domains per page, domain authority and traffic estimates for competitors you monitor, and which ChatGPT and Google AI Overview answers mention a keyword or cite a domain in the AI visibility tool. None of it needs a key from you.

CMS connections: what is read and what is written

Each brand holds one CMS connection to WordPress, Webflow or Shopify. Vee3.ai reads only what publishing needs: a check that the credentials work, existing entries with the same slug so a refresh updates the right page, categories and your media library. It writes the page, its SEO plugin fields, its featured image and, where needed, a category. The full picture, including drafts, refreshes and duplicate safety, is on CMS publishing.

Imports and exports: CSV in, CSV and PDF out

Bring an existing plan with you. The roadmap import takes keywords, page URLs, mixed lists or CSV, up to 1,000 rows per import. Recognised headers such as Keyword, Page URL, Status, Volume, KD or GSC Clicks map to roadmap columns, and the rest are listed as ignored. Values you supply win, saved roadmap and scan evidence fill the blanks, and providers are asked only for what is still missing. Drafts come into the editor by paste, by URL or as a plain text, Markdown or HTML file.

Going the other way, the roadmap, the crawl, the technical checklist and the publishing calendar export to CSV built for Excel, with accented keywords intact and formula-like cells kept as text, and keyword and domain comparisons download as CSV too. The dashboard and the Atlas export to PDF, and a public, read-only link to the topical authority map lets a client explore the result without an account.