Put Your AI Visibility Data in Data Studio: A Setup Guide for the Beket Connector
Eight steps to connect beket.ai to Google Data Studio (aka Looker Studio) — and how to tell a bad key from a bad plan from an empty audit.
On this page
Your AI visibility data sits in Beket. Your stakeholders work in Data Studio. Google renamed that tool back from Looker Studio in April 2026. This guide connects the two.
Setup takes five minutes. Three steps fail in ways that look the same on screen. Read those parts twice.
Prefer to watch? Here is the full walkthrough.
1. Create an API key
Go to app.beket.ai. Open Settings → Developer → API keys. Create a key.
Copy the bkt_… value now. It is shown once.
Your org must be on a paid plan. A free org gets a 402 from the API. On screen that looks like a broken connector. Check the plan first when something breaks.
2. Open the connector
Stay on the same page. Find the “Connector link” card. Copy the link. Open it in a browser window and click Authorize.
3. Paste the key
An error here means one of two things. The key is wrong, or the plan is free. Check both.
4. Pick a report
Three options: Visibility over time, Share of voice, Engine breakdown.
This screen is stepped. Only the report field shows. Choose one and hit Next.
5. Pick an audit
The audit dropdown appears. You get “All audits (organization)” plus one row per audit. Each row is labelled Name (domain).
This is the first real fetch of /audits. Your key already worked. This call tests your data, not your credentials.
“No audits with data yet” means the endpoint answered and returned nothing. That is a different failure from step 3. There the key was rejected. Here the key was accepted and there is nothing to chart. Run an audit. Then come back.
6. Set a date range
The date range is required. Pick a window you know has snapshots.
An empty range gives an empty report. That looks like a broken connector too.
7. Connect
Click CONNECT. The field list loads.
8. Create the report
Click CREATE REPORT. You get a table with rows.
From here it is normal Data Studio. Chart share of voice by engine. Trend visibility by week. Drop it in the deck your client already reads.
Put all three tables in one report
Each connector link gives you one report. One report is one table. For all three in one place, build three data sources and blend them.
- Open the connector link three times, once per report. That gives you three data sources.
- In your report, go to Resource → Manage blends.
- Join the tables on Date + Audit ID. Not Snapshot ID.
The join key matters. Share of voice is recomputed on full runs only. For the same date it can come from a different snapshot than the other two tables. Join on Snapshot ID and those rows drop out of the blend. A dropped row looks like missing data. Date + Audit ID keeps them.
Why share of voice looks sparse
Share of voice runs on the full-run schedule. That is about one day in seven. It is not daily.
Chart it and you get scattered points, not a line. Visibility over time updates far more often. Share of voice lands about once a week. Plot them together and the sparse series is the healthy one.
Three failures, three causes
Most tickets on this connector come down to one of three things.
Step 3 fails: the key is wrong, or the org is free. Same message either way.
Step 5 is empty: the key works and no audit has data.
Step 8 is empty: the date range has no snapshots.
Check them in that order. Each one rules out the next.
beket.ai scores how ChatGPT, Gemini, Perplexity, and Copilot describe your business. Then it shows you where they get it wrong. The Data Studio connector puts those numbers where your team already looks. Start at beket.ai.