Turn your Looker dashboard into a Monday leadership review

Every Monday at 7am, build a narrative weekly business review from my main Looker dashboard, archive it in Notion, and email a short summary to my leadership list.

Start in Looker. Use Search Dashboards to find my main dashboard by title, then Get Dashboard to read its metadata. Do not work from a hardcoded list of tiles. Use Get All DashboardElements to discover every tile currently on the dashboard, and Get DashboardElement when you need more detail about a specific one. Then run each tile the right way for its type: use Run Look for tiles backed by a saved Look, and Run Query for tiles backed by a query defined on the dashboard itself. Discovering the tiles at run time is the point, because it means a metric someone adds to the dashboard next month shows up in the review automatically instead of being silently missed.

Pull enough history to talk about direction rather than level. For each tile, get the current week and the preceding weeks so you can describe a trend, say whether a move is inside normal week to week noise, and compare against the recent average instead of only against last week.

Before writing anything, use Query a Data Source in Notion to read the most recent review pages from my weekly review database, sorted newest first. Use them for continuity: what we said last week, whether what we predicted actually happened, and which metrics were already flagged as behind target. The archive is the memory for this workflow, so read it before you add to it.

Write the review as prose for executives, not as a metric dump. For each meaningful movement, say what moved, by how much, and why it most likely moved. Ground the explanation in what the data can actually support: look for a metric that moved in step with it, an effect isolated to one segment or channel, or a change in the mix underneath a flat headline number. Where you are inferring rather than reading a cause directly, say so plainly instead of asserting it. Skip metrics that did nothing. A metric that was flat and is meant to be flat does not need a paragraph.

Explicitly flag any metric that has missed its target two weeks running, using the previous reviews from Notion to establish the repeat. Call these out prominently rather than burying them in the body, and note whether the metric is getting worse, holding steady, or recovering while still short. A repeat miss is a different conversation from a first miss and the review should make that obvious.

Close the review with the three things genuinely worth discussing in the leadership meeting. Choose them on the basis of what needs a decision, what is trending badly enough to matter soon, or what nobody currently owns. Do not simply pick the three largest percentage changes. For each one, state the question leadership actually needs to answer, not just the observation behind it.

Save the full review as a new page in my Notion weekly review database using Create a Page, titled with the week ending date so the archive stays sortable. Put the narrative in the page body and set whatever properties the database has, such as the week, the headline summary, and the metrics currently behind target.

Then use Send a Message in Gmail to email my leadership list. Keep the email short: a two or three sentence read on the week, the repeat target misses, the three discussion items as a list, and a link to the Notion page for anyone who wants the detail. The email is the briefing people read on their phone before the meeting, so it should stand on its own without the link being opened.

If a tile fails to run or returns no data, say so in the review and carry on with the rest. A partial review that names its gaps is far more useful than a failed run, and silently dropping a metric is the one outcome to avoid.