AI Connector
Connect your location data to Claude and ChatGPT so you can ask questions about how you are doing in plain words.
The AI connector links your location data to Claude or ChatGPT so you can ask questions in plain words and get an answer back. It can read your data. It cannot change anything.
What it does
A dashboard answers the questions the person who built it thought of. Everything else means exporting to a spreadsheet, joining the tables by hand and working it out yourself. So most one off questions about local performance never get asked at all, because the asking costs more than the answer is worth.
That threshold matters more than it sounds. Which locations lost ground this month and also have reviews nobody has answered? That is a fair question with a useful answer, and almost nobody asks it, because getting to it means half an hour of exporting and sorting that nobody has spare.
Connecting your data to Claude or ChatGPT takes that cost away. Ask in plain words, get the answer, act on it. Questions that were never worth the effort become normal. It changes what your data gets used for, not just how you reach it.
It can only read. The connection cannot publish a post, edit a profile or change a setting, and that is deliberate. It also sits alongside your dashboards rather than replacing them. A dashboard is still better for the numbers you look at every week.
Everything you get
Ask in plain words
Ask about rankings, reviews, profiles and tasks without building a report or setting up a filter first.
Works with Claude and ChatGPT
Connect through a standard connector to whichever assistant your team already has open.
Read only by design
The connection can look at your data. It cannot publish, edit a profile or change a setting.
Scoped to your account
Only your locations are visible, and you can narrow that further per user or per client.
Questions that cross subjects
Ask about rankings, reviews and profile health in one question. Dashboards keep those on separate screens.
Compare by asking
Compare locations, months or clients in a sentence instead of building each comparison by hand.
Take the answer with you
Drop what you find straight into a client update or an internal note without rebuilding it.
Your data is not used for training
Your data is queried and the answer comes back. It is not fed into a model.
Every question logged
A record of what was asked and by whom, which starts to matter once a whole team is using it.
From setup to first result
- 1
Connect the account
Link your account to Claude or ChatGPT through the connector. It takes a few minutes and no code.
- 2
Set who can see what
Limit access per user or per client so nobody can ask about a location they do not look after.
- 3
Start with a real question
Ask something you would normally have exported a spreadsheet for. That is where the value shows up first.
- 4
Check it against the dashboard
Run two or three questions you already know the answer to. It tells you how far to trust the rest.
- 5
Ask the questions that cross subjects
Rank and reviews and profile health together. Those are the ones a dashboard cannot answer.
- 6
Keep the questions that work
The useful ones get asked again every month. Write them down.
- 7
Put the answers in the report
An answer that never leaves the chat window has not saved anybody any time.
What changes when you use it
- Expecting it to replace your dashboard
- Asking a vague question
- Never checking the working
- Leaving access wide open
- Leaving the answers in the chat
- Most questions never get asked
- Dashboards keep subjects apart
- Plain words beat filters
- Read only makes it safe
- Comparison gets cheap
Most questions never get asked
If getting an answer takes half an hour of exporting, the question quietly stops being worth it. That is where most useful analysis dies.
Dashboards keep subjects apart
Rank sits on one screen and reviews on another. Real questions cross both, and joining them by hand is the part nobody has time for.
Plain words beat filters
Not everyone on a team can build a report. Everyone can ask a question.
Read only makes it safe
The worst case is a wrong answer, not a wrong edit pushed to forty profiles.
Comparison gets cheap
Comparing two months, two locations or two clients stops being a task and becomes a sentence.
It fits how people already work
Your team is already in Claude or ChatGPT. This puts your own numbers where the questions are being asked.
What it measures
- Rankings by location and keyword
- Reviews, ratings and which ones are unanswered
- Profile completeness and what is missing
- Tasks open and tasks done
- Change between any two periods
- Anything else held in your account
Reports you can send
Every output can be exported and carries your branding, never ours.
A written answer
The response to your question in plain words.
A table you can copy
Numbers pulled out in a form you can paste into a report.
A comparison
Two locations, two clients or two periods lined up.
A shortlist
The locations that match what you asked about.
Saved questions
The useful ones kept so they can be run again next month.
A query log
A record of what was asked and by whom.
Who this is built for
Answer a client question in the meeting instead of promising to look into it.
Ask across the whole estate without waiting for somebody to build a report.
Let head office ask about any territory without opening thirty dashboards.
Ask a plain question about your own numbers and get a plain answer.
Get more out of it
- Start with a question you would normally have exported a spreadsheet for.
- Check two or three answers you already know, so the team trusts the rest.
- Ask questions that cross subjects. That is where a dashboard cannot help.
- Keep the good questions written down and run them every month.
- Scope access per client before the whole team starts using it.
- Use the dashboard for the weekly numbers and this for the one off questions.
Expecting it to replace your dashboard
A dashboard is better for the numbers you check every week. This is better for the question you only ask once.
Asking a vague question
Which locations are doing badly gets a vague answer. Which locations dropped more than three places since last month gets a useful one.
Never checking the working
Ask it something you already know. It is the fastest way to find out how far to trust it.
Leaving access wide open
Everyone able to ask about every client is a problem waiting to happen. Scope it first.
Leaving the answers in the chat
An answer nobody puts into a report has saved no time at all.
By hand, or with this
| Doing it manually | With AI Connector |
|---|---|
| Export, sort, pivot, then answer | Ask the question and read the answer |
| Only the questions the dashboard planned for | Any question your data can answer |
| Rank on one screen, reviews on another | Both in one question |
| Report building skill needed | Anyone on the team can ask |
| Comparisons built by hand every time | A comparison in one sentence |
| No record of what was asked | Every question logged |
- Connect your account to Claude or ChatGPT
- Scope access per user or per client
- Ask three questions you already know the answers to
- Write down the questions that turn out useful
- Ask one question that crosses rank and reviews
- Put one answer into the next client report
- Keep the dashboard for the weekly numbers
One of 42 tools on the same plan
Every tool below is on the same plan at the same price. Nothing here is an extra, an upgrade or a separate subscription.
Rank & Visibility
Manage Profiles
Posts & Automation
Reviews & Reputation
Reports & White-Label
Questions people ask
What does the AI connector do?
It links your location data to Claude or ChatGPT so you can ask questions in plain words. Which locations slipped this month, which profiles are missing services, which reviews are still unanswered. No report building first.
Can it change anything?
No. The connection is read only. It can look at your data and answer questions about it. It cannot publish a post, edit a profile or change a setting.
Is my data used to train a model?
No. Your data is queried and the answer comes back. It is not fed into training.
Which assistants does it work with?
Claude and ChatGPT, through a standard connector. You use whichever one your team already has open.
Does this replace my dashboard?
No, and it is not meant to. A dashboard is better for the numbers you check every week. This is better for the question you ask once and would otherwise skip.
Can I limit who sees what?
Yes. Access can be scoped per user or per client, so nobody can ask about a location they do not look after.
What kind of question works best?
Specific ones. Which locations dropped more than three places since last month works well. Which locations are doing badly does not.
Every tool. One price. Nothing extra to buy.
$5.33 per location per month gets you AI Connector and the other 41 tools.