Do what works. See the effect in numbers.
After every measurement, we suggest actions with an expected effect in points: reach out to a source, fill a content gap, open up your site to crawlers, recover sentiment. Every action has a step-by-step plan. Thirty days after 'done', we measure the real effect.
Tracking without action doesn't change what AI says about you.
A dashboard full of numbers tells you where you stand, not what to do next. And GEO is new for most businesses. Which source should you reach out to? Which page is missing? Which line in robots.txt is blocking AI crawlers? Without a concrete answer, you end up just watching.
So GeoReady turns every measurement into actions. Not generic advice, but what's actually happening in your answers, sources and site check: this prompt doesn't mention you, this site cites your competitor and not you, this crawler is blocked. Every action comes with the evidence, a step-by-step plan and an expected effect.
And because you want to know what it delivered, we measure that too. Thirty days after you mark an action done, we compare your visibility on the related prompts with the thirty days before. Over time, you learn which actions actually work for your business.
“Every action comes from your own data. Every effect gets measured.
Suggestions from your answers, sources and site
After every measurement, GeoReady looks at your answers, citations, competitors and latest site check, and only suggests actions where the signal is real. There are several types: reach out to a source, fill a content gap, make a technical fix, create or improve a page, add a prompt, or address a recurring sentiment theme.
Specialized checks also help businesses that are already visible: prompts where you're mentioned but rank third or lower, mentions where AI cites other sites instead of yours, cited pages that are over a year old or lack a date and author, titles without a service or location, and services from your business profile you aren't tracking yet.
If an action already exists, it gets updated instead of duplicated. Dismiss an action and it stays away for ninety days. And when the signal disappears, for example because a later site check finds your llms.txt, the action closes itself.
- See new actions after every measurement, labeled by type: source outreach, content gap, technical, create page, add prompt.
- Read “Why this matters” and the evidence from your own data for every action.
- Dismiss an action that doesn't fit. It won't come back for ninety days.
- Create your own action, with notes and an optional expected effect.
Know which sites to be on, and how to get there
AI models mention businesses that already appear on comparison sites, directories, review sites and in the media. When an AI model cites a site for your prompts that lists your competitors but not you, you get an action to reach out, with the number of citations and the AI models using it.
For that source, GeoReady looks up the sign-up path: the sign-up, claim or contact page on the site itself, two to four steps to get listed and, when known, what it costs. That saves you the digging and turns “get on that site” into a ten-minute task.
Source actions are ranked by expected effect. A comparison site cited by three AI models across ten prompts ranks above a blog that came up once.
| Domain | Type | Citations | You |
|---|---|---|---|
| heatpumpcompass.com | comparison | 184 | missing → action |
| northfieldheatingair.com | own site | 96 | mentioned |
| hvacfinderguide.com | directory | 71 | missing → action |
| summitclimateco.com | competitor | 58 | missing → action |
| trustpilot.com | reviews | 41 | mentioned |
- See the domain, citation count and AI models for every source action.
- Open the sign-up or claim page straight from the action.
- Follow the steps and see the cost, when it's known.
- Jump to the source (Source →) to see which prompts cite it.
Every action tells you what to do, with ready-to-use code
An action has three parts: why it matters for your AI visibility, what we found on your end, and “Here's how” with concrete steps. Technical actions include ready-to-use code filled in with your business profile: robots.txt rules that let AI crawlers in, a complete llms.txt, JSON-LD with your business details, FAQPage schema, AggregateRating for your reviews, or sameAs links to your profiles.
Content actions are just as specific. A content gap names the prompt where you're missing and who gets mentioned instead. A ranking action shows who holds the top spot and through which source. You also get actions per AI model, matched to how it works: ChatGPT searches through Bing, Gemini through Google, Perplexity through its own index, and Claude favors factual pages.
That means anyone on your team can pick up an action, even without GEO experience. Copy the code, paste it into the right file, and the next site check picks it up.
- Expand “Here's how” for the step-by-step plan.
- Copy code blocks (robots.txt, llms.txt, JSON-LD) in one click.
- See which prompts a content gap or AI model action is about.
- Go from a technical action to My Website to see the check.
An owner, a deadline and notes for every action
Actions are work, and work needs an owner. Assign each action to a teammate, give it a deadline, and use notes to track what was done, with whom and when. Actions move through open, in progress (“Pick up”), done or dismissed.
At the top you see the three open actions with the most expected effect, their combined expected effect, and whether the first one has an owner yet. In a weekly meeting, you know exactly where to start.
Completed actions stay visible with their date and, after thirty days, their measured effect. The top of the page also shows the average measured effect of your completed actions and your average lead time, so you know how fast your team ships.
- Assign an owner and set a deadline for every action.
- Keep notes: what was done, with whom, when.
- Switch between open, in progress, done and dismissed.
- See your average lead time per action.
- Reopen a done or dismissed action when you need to.
Expected effect up front, measured effect after
Every action has an expected effect in visibility points. It's based on the win-rate gap with your competitors, times the number of prompts involved, weighted by the AI model's share. It's an estimate, not a promise, but it's a fair way to rank your actions.
When you mark an action done, the expected effect shows up as a reward. Thirty days later, GeoReady measures the real effect: your visibility on the related prompts in the thirty days after, compared with the thirty days before. If the action has no prompts of its own, we look at the whole project.
When the effect is positive, you see “Your action worked” along with the proof. In a few months, you build your own record of what works for your business and what doesn't.
- Sort actions by expected effect in points.
- See “effect pending” on completed actions until the thirty days are up.
- Then read the measured effect in points after 30 days.
- Open the proof for an action that worked.
How GEO Recommendations works.
From twelve suggestions to three actions with measured results
After its first measurement, Northfield Heating & Air had twelve actions waiting. The team had one morning a week for GEO.
- 1Pick by effect
At the top: “Allow GPTBot and OAI-SearchBot” (+3.1 pt expected), followed by a source action for heatpumpcompass.com, where Summit Climate Co. and Redwood Home Services were already listed. Two actions about a blog that was cited once got dismissed.
- 2Ship it with the code
The web developer pasted the robots.txt rules from the action. The next site check saw both crawlers as allowed and closed the action on its own. For heatpumpcompass.com, the owner followed the sign-up path: claim the business profile, free, three steps.
- 3Fill a content gap
The action for “hybrid heat pump Alexandria” showed Summit Climate Co. mentioned in 70% of answers. The team wrote a page on hybrid heat pumps in Alexandria, with pricing and lead times, and marked the action done.
- 4The effect after thirty days
The source action delivered +2.4 points on its related prompts, the content page +1.8. Northfield went from being mentioned in 2 of 25 prompts to 6.
Every number traces back to the exact prompts.
Questions about GEO Recommendations.
Another question? Email info@geoready.nl.Can I add my own actions?+
Yes, with an expected effect. Those are measured after thirty days too.
What are points?+
The expected effect of an action on your GeoReady score. You get them as a reward when you mark it done; after thirty days you see the real effect.