BotPass — GEO Experiments for WordPress

01 / SOURCE DETAILS

About this tool

BotPass is a GEO platform that proves which on-page changes make AI assistants cite your pages — using a real experiment with a control group. Most GEO advice is correlational: someone changed twenty things, citations went up, and nobody knows why. BotPass built the control group. How it works Pick — choose one change from a closed catalogue (never a free-text guess). Split — your eligible pages are drawn at random into a treated half and a control half. Apply — the plugin changes the treated pages only, always reversibly. Nothing touches the control. Measure — BotPass queries four AI providers directly and logs every crawler that visits your pages. Compare — after eight weeks, the difference between the two groups, with its confidence interval, is your result. Changes that clear the statistical bar enter a shared playbook and are applied automatically to every site with that page type — including yours, whether or not you ran that experiment. What the plugin does The plugin is the WordPress half of the BotPass system. The server decides which pages get each change and measures the result; the plugin executes the assignment on your site and reports back. On your WordPress site, the plugin: Pulls experiment assignments from the BotPass server and applies approved changes to the right pages. Applies every change in the render layer (not in the database), so reverting means deleting one record — nothing to undo in post content. Logs every crawler visit, identifies the AI provider by user-agent and IP, and sends the signal to the measurement server. Emits a human-referral event when a visitor arrives from an AI-referred link, which is the denominator of the extraction ratio. Never modifies your templates, your styles, or your SEO plugin’s output. Works alongside Yoast, RankMath, WPML, Polylang, WooCommerce, and any standard caching layer. What you see in the dashboard Status — a live readiness check: plugin connected, parity verified, baseline window complete, pages inventoried. The checklist is computed from the database on every load, not from stored flags, so it stays honest even after a theme update or a caching plugin change. Experiments — the experiment catalogue (every change BotPass can test, expand or collapse, with a link into the BotPass app for this site), plus all experiments on your site: which are in progress, which are awaiting the AI crawl, and which have a result. Each one shows what it is waiting for and who, if anyone, needs to act. Settings — API key, the opt-in/opt-out toggle for network waves, the risk ceiling for automatic changes, and the IP logging preference. The extraction ratio diagnostic Before you run an experiment, BotPass can show you your extraction ratio: how many pages each AI provider reads for every visit it sends back. The diagnostic is free, runs over 28 days, and requires no payment card. It is the first number you need to decide whether running an experiment is worth it. Extraction ratio by provider (industry, July 2026) Mistral: 3,389 pages read per referred visit Anthropic: 2,237 : 1 Perplexity: 225 : 1 OpenAI: 217 : 1 Copilot: 35 : 1 Google: 4.6 : 1 Your site’s number will be different. Finding it costs nothing. Safety and cloaking BotPass serves identical content to people and to AI machines. After each experiment wave, the server issues a parity report by URL and date that your SEO team and legal team can inspect. Cloaking means showing machines something different from what people see; BotPass does not do that, and the parity report is the evidence. Every change in the catalogue is reversible with one click. A record of what was applied, to which pages, and when is always available. Plans Network — $39/month Everything already proven applied to your site. You join the network waves that test new changes. Extraction ratio, always on. Program — $569/month One experiment inside your own site, up to 300 pages, with your own control group and a result with a confidence interval. Three-month minimum, because one experiment takes eight weeks. Program Plus — $2,199/month Three concurrent experiments, up to 300 pages each, with four AI providers measured. Agency and Enterprise — custom Domain portfolio view, white-label reports, REST API, SSO. Priced on request. The diagnostic is free for any plan, no card required. Compatibility WordPress 6.0+, PHP 8.0+ Compatible with Yoast SEO, RankMath, All in One SEO Compatible with WooCommerce, WPML, Polylang Compatible with WP Rocket, W3 Total Cache, LiteSpeed Cache, and other page caches The plugin flags itself as inactive if it cannot verify parity with the active cache layer What BotPass is not Not a visibility tool. Visibility tools measure what the AI says; BotPass changes your site and measures the difference. Not a GEO content agency. Nobody from our team touches your site. Not a guarantee of citations. Nobody controls what a model answers. Anyone who guarantees it is lying. External services This plugin connects to the BotPass API at https://api.botpass.io (operated by BotPass). Without that connection the Status, Experiments and measurement features cannot run. Account signup and the web app live on https://botpass.io. What leaves your WordPress site All calls use HTTPS. The plugin authenticates with the API key you paste in BotPass → Settings (sent as a Bearer token). No WordPress AUTH_KEY / salts, user passwords, or wp-config secrets are transmitted. Data sent to https://api.botpass.io includes: Site identity: home URL / domain, and (after connect) the remote site id returned by BotPass. Page inventory: counts and eligible page lists for experiment sizing (URLs, titles, post types — not full post bodies unless an audit/render request needs HTML for a single URL you triggered). Experiment sync: assignment acknowledgements, which interventions were applied or reverted, and readiness / parity signals used by the Status screen. Crawler and referral events: public URL visited, bot slug / type / provider (from user-agent matching), HTTP method, status code, optional Accept / referer / variant / cache metadata, token estimates when available. Visitor IP is included only if you enable IP logging in Settings; otherwise it is omitted or hashed locally per your setting. On-demand admin actions you start: GEO / entity / fan-out audits for a URL, Markdown render requests, llms.txt generation, and similar API reads tied to screens you open. What stays on your site Post content edits are not pushed as a bulk dump. Render-time intervention flags live in post meta on your database. Analytics rows in {prefix}botpass_analytics stay local until a corresponding event is also sent for measurement. Account requirement Create a BotPass account at https://botpass.io, obtain an API key, and paste it under BotPass → Settings. By using the plugin you agree to the BotPass Terms of Service and Privacy Policy. Service provider: BotPass — https://botpass.io Terms of Service: https://botpass.io/legal/terms Privacy Policy: https://botpass.io/legal/privacy

02 / SOURCE DETAILS

Getting started

In your WordPress dashboard, go to Plugins > Add New. Search for “BotPass” and click Install Now, then Activate. Go to BotPass > Settings and paste your API key from botpass.io. The Status screen will walk you through the remaining readiness checks (baseline window, page inventory, parity confirmation). The initial baseline window is 14 days. BotPass measures your pages before touching anything, so the experiment has a clean starting point. For non-WordPress sites, BotPass connects over the REST API with no plugin required. See botpass.io for documentation.

03 / SOURCE DETAILS

Frequently asked questions

How is this different from the AI visibility tool I already use? Visibility tools measure from the outside: they query the models and report what came back. They tell you what happened. They cannot tell you why, because nothing on your site changed on purpose. BotPass changes a randomly drawn half of your pages and compares against the untouched half. That comparison is what makes a causal claim possible. Keep your visibility tool — BotPass can read its data as a measurement source. How long until the first result? Six to eight weeks from the start of the experiment. That is how long an AI takes to come back, re-read your changed pages and reflect them in its answers. Anyone promising results sooner is not measuring anything. Isn’t this cloaking? Could Google penalise my site? No. BotPass serves exactly the same content to people and to machines, and issues a parity report by URL and date that documents it. Cloaking is serving something different to machines; BotPass does not do that. What happens if an experiment comes back flat? You are told in the same detail as if it had worked, and the next hypothesis is on the list. Knowing something does not work saves you the budget you were about to spend on it. That is part of the product, not a failure. Who writes to my site, and with what access? Nobody from the BotPass team. The plugin you install, running on your own server with credentials you control, pulls a task list and applies changes only to the pages in the experiment. Every change comes from a closed catalogue and every change is reversible. Does the plugin write to my post content? No. Every change is applied in the render layer, from a flag in post meta. Reverting means deleting one database row; there is nothing to undo in the post editor. Your content is never modified. Is my data shared with other sites in the network? What is shared is the aggregate measured effect of each change: the intervention type, the page type, the effect size and its confidence interval. Never your content, your traffic, your keywords, your domain or your company name. It is specified in the Data Processing Agreement from day one. Does it work with my caching plugin? Yes. BotPass requires that each URL is served consistently — one version to people, one version to machines — and the plugin verifies this on setup. If it cannot confirm that the active cache layer varies responses by the right headers, it flags this in the Status screen and does not proceed until the cache is configured correctly. What does it cost? Network is $39 per month. Program is $569, with a three-month minimum because an experiment takes eight weeks. Program Plus is $2,199 for three concurrent experiments. Agency and Enterprise are priced on request. The extraction-ratio diagnostic is free, no card required. My site is not on WordPress. Can I use BotPass? Yes. For Shopify, Magento, PrestaShop, headless or custom builds, BotPass connects over the REST API. Your technical team integrates it once; after that it works the same way.

Explore related topics