Skip to content

Weekly Restaurant Performance Summary

Example prompt: "Every Monday morning at 8am, pull last week's covers from our bookings sheet and the takings from the POS export we get emailed on Sunday night, plus any one or two-star reviews from the week and the no-show list, and send me a short summary email with the numbers, the comparison against the previous week, and a couple of lines reading what stood out."

The Problem

The owner of an independent restaurant wants the same four numbers every Monday morning — how many covers we did, what we took, where the reviews landed, who did not turn up — and assembling them takes thirty to forty-five minutes of clicking around the bookings sheet, the POS export, the Google Business Profile reviews, and the manager's WhatsApp notes about the no-shows. The job is small enough that nobody schedules an hour for it and large enough that it slips, which means the Monday-morning sit-down with the manager either does not happen or happens off half-numbers nobody has time to cross-check. The week-to-week reading is what catches the slow Tuesday drift early and the new starter on the section who is consistently behind on the pass — and it only catches them if the numbers are in front of the owner every week without fail.

How GloriaMundo Solves It

We build a workflow that fires every Monday at 8am and assembles the weekly summary from the four sources. An integration step reads last week's covers from the 'Reservations' tab in the bookings Google Sheet, broken down by day and service. A second integration step opens the POS export the supplier emails on Sunday night, parses the CSV attachment, and pulls the takings totals, the average spend per cover, and the food-cost percentage if the report includes it. A third integration step reads the 'Review Replies' log for the week and picks out the one and two-star reviews along with the highest-rated. An LLM step assembles the lot into a one-screen email — covers and takings up top with the comparison to the previous week, the no-show count and the names if any, the review highlights and lowlights with the link to each, and two or three sentences reading what stood out. The email lands in the owner's inbox at 8am Monday so the manager-meeting at 10am happens off the same numbers both sides have already read. Glass Box preview shows the assembled summary before it sends so the first run can be sense-checked against the bookings sheet by hand.

Example Workflow Steps

  1. Trigger (scheduled): Every Monday at 8am.
  2. Step 1 (integration): Read last week's bookings from the 'Reservations' tab in the bookings Google Sheet, broken down by day and service, including the no-show flag and the cover count per booking.
  3. Step 2 (integration): Open the most recent POS export email from the till supplier received between Sunday 8pm and Monday 6am, parse the CSV attachment, and pull total takings, covers, average spend per cover, and the food-cost percentage where the report includes it.
  4. Step 3 (integration): Read the 'Review Replies' tab for the week and pull the one and two-star reviews and the highest-rated review with the platform and the link to the original.
  5. Step 4 (code): Compute the week-on-week comparison for covers, takings, and average spend, and compute the no-show rate as a percentage of total bookings.
  6. Step 5 (llm): Assemble the summary in a one-screen format — covers and takings up top with the week-on-week deltas, no-show count and names, review highlights and lowlights with the links, and two or three sentences reading what stood out across the four sources.
  7. Step 6 (integration): Send the summary as a Gmail email to the owner with a subject line of 'Weekly summary — week ending [date]'.
  8. Step 7 (integration): Post a short Slack DM to the owner with the headline numbers — covers, takings, no-show rate, review tone — so the email is read against the right expectation.

Integrations Used

  • Google Sheets — the 'Reservations' tab for the covers and no-show count, and the 'Review Replies' tab for the review highlights and lowlights
  • Gmail — both the source of the POS export CSV (parsed for the takings numbers) and the destination of the Monday-morning summary email
  • Slack — the owner's DM with the headline numbers so the email is read against the right expectation

Who This Is For

Owner-operators of independent restaurants, bistros, and gastropubs where the bookings sheet lives in Google Sheets and the POS supplier emails a weekly CSV report on Sunday night — anyone who currently spends thirty to forty-five minutes assembling the Monday numbers and would prefer them already sitting in the inbox at 8am so the manager-meeting at 10am runs off the same read.

Time & Cost Saved

The assembly is thirty to forty-five minutes of an owner's time every Monday, which is two and a half to four hours a month — the kind of job that gets done some weeks and not others, and the weeks it does not get done are the weeks the slow drift on the Tuesday lunch sitting is not caught early. The workflow does not replace the reading — the owner still reads the summary and decides what to do with the numbers — but it removes the assembly and standardises the format so the week-to-week comparison is real rather than half-eyeballed. The Monday-morning meeting with the manager runs better off the same numbers in both heads, and the slow drift is caught two weeks earlier on average.