📘 This use case demonstrates producing the same figures in the same order every cycle, with the verification steps that keep a repeated report from drifting.
Best for: Search leads and teams reporting across several brands, regions, or markets · Uses: Search Console and AI Visibility, via your AI assistant
Overview
Most Botify MCP questions end at an answer on screen. This one ends at a document that goes to someone else, on a cycle. The MCP supplies the Search Console and AI Visibility figures, and your AI assistant assembles them into the same report structure each period.
Prerequisites
Botify MCP connected to your AI assistant, such as Claude or ChatGPT. See Connecting to the Botify MCP for setup steps.
Access to the Botify projects the report covers.
A connected Search Console property, for performance figures.
AI Visibility configured, if the report covers brand mentions and citations.
A decision about which figures come from Botify and which must come from a source the MCP does not reach.
Description of the problem
A report that goes out on a set cadence carries two constraints a one-off answer does not. It has to present the same figures in the same order every period, and a report that reinvents itself each month cannot show a trend. Since the numbers leave your team, a mistake is found by the reader rather than by you.
Application
Settle your sources first. List the figures the report needs and mark which ones the MCP can supply from your Botify projects and which have to be pulled from elsewhere. Request anything in the second group at the start of the cycle, because the Botify sections can be compiled while you wait.
Pull the Botify figures in one request:
For [project], pull Search Console clicks, impressions, and average position for [date range], compared with the same range in the prior period. Add the pages and queries that moved most, plus AI Visibility brand mention and citation results for the same window. Return the same figures in the same order every time I ask.
Why this prompt works:Specifying the exact figures, the comparison window, and the instruction to "return the same figures in the same order every time" is what makes a report reproducible from cycle to cycle, rather than a slightly different answer each time the same question is asked.
Check any figure you derived rather than read directly. Before totaling a column of period-over-period differences, count how many cells are blank. If a meaningful share are empty, do not build the prior-period total by summing row by row: most tools drop blanks silently, which understates the total and can reverse the direction of a trend. Use the reported total from the source instead.
Compare every computed aggregate against a total you have already seen from that source before it goes into the document.
Assemble the report in the same structure as last period, then read the rendered output before it goes out.
Follow-up prompts
"Show me only what changed by more than [threshold] from last period." — surfaces what's actually worth calling out in the cycle's narrative.
"Pull the same figures for [additional brand or market]." — extends one cycle's report to cover another market without redefining the structure.
The verification habits in steps 3–4 are what make this safe to repeat: run them every cycle, not just the first time, since a blank-cell or total-mismatch error is exactly as easy to make on cycle ten as on cycle one.
Result
The same figures in the same order every cycle, each traceable to where it came from, gathered in one pass instead of assembled from separate exports. Because the Botify figures come from a stored snapshot rather than a live query, rerunning the prompt for a past period returns the same numbers, so the report you sent last month still reproduces.
Read next
Related topics