Skip to content

Connecting your data sources

The more a project knows, the better every surface gets — and every connection is read-only for data, honest about its status, and never publishes or writes anywhere without your explicit action. This guide covers what you can connect yourself and how each source behaves.

All of these are connected from the Integrations canvas (Data sources in the Command Centre), each on its own tab:

  • Google — one sign-in grants read-only Search Console and GA4, and the same connection covers Google Ads and Local Services reporting reads. See “Connecting Google” for mapping and failure states.
  • Bing Webmaster — read-only, with your API key. No URLs or sitemaps are ever submitted.
  • WordPress — your site URL plus an Application Password, so approved content can be published as drafts or live posts. A project links to the connected site it publishes to. Installing the WordPresto Connector plugin on the site turns that into a full two-way integration — lossless block-level import, Yoast and ACF fields, and an honest report of what the site can actually do. See “The WordPress plugins”.
  • Sanity — a real consent-screen sign-in, then per-project id and dataset mapping.
  • Site repository (GitHub) — connect and pick a repository, for site publishing by pull request.
  • Social channels — LinkedIn, X, Facebook and Instagram connect with an account sign-in; Bluesky connects with an app password (created in your Bluesky settings), since it is an open protocol with no platform app to configure. Connecting enables drafting and the inbox; posting always needs your explicit action.
  • Site Signals — WordPresto’s own first-party analytics. The card generates your unique install snippet from the project’s URL; paste it into your site and the card shows honestly whether events are arriving. On WordPress you can install the WordPresto Tag Injector plugin instead of editing your theme — it takes the Site ID and prints the tag for you (see “The WordPress plugins”). It tracks page views, AI-assistant referral sessions (ChatGPT, Perplexity, Claude, Gemini, Copilot), conversions, outbound clicks, rage clicks, scroll depth and dwell time. A separate server-side key covers AI crawlers, which run no JavaScript.

Keyword, SERP and competitor enrichment and the answer-engine checks run through providers WordPresto manages centrally — there is no personal key to add. You use them through explicit, credit-metered actions (keyword research, rank tracking, competitor analysis, AI visibility), and you always see the projected cost before a run.

A connection is only useful once it produces data. After connecting, use Populate project data (or wait for the automatic six-hourly sweep) to pull safe, read-only data from your sources into the project. Paid sources and large crawls always stay explicit. Each source reports its own honest status — connected, mapping required, no data yet, reconnect required — so you can always see what the project really knows and why.