Skip to content
ShyftKick

Daily briefings · in pilot

Every morning: what changed, and what to check.

A short email for restaurants with one to ten locations. It reads yesterday’s sales and labor, compares them with days that are actually comparable, and tells you what moved, the numbers behind it, and the one thing worth looking at. It is being built and tested inside a working restaurant, and the pilot list is open.

Join the pilot

A morning, for example

Sample — a fictional restaurant and fictional numbers

Copper Fork Kitchen

Friday, September 18

Complete — POS import reconciled at 5:12 AM

Net sales
$3,569−15%
vs a median of $4,200 over 8 Fridays
Checks
205−15%
vs a median of 241
Sales per labor hour
$57.56−16%
vs a median of $68.52
Direct labor
29.5%
wages only; excludes salary and payroll taxes
  1. Priority

    Lunch sales fell 29% below the Friday baseline

    • Lunch (11:00–2:00) net sales $927 vs a median of $1,306 across 8 comparable Fridays
    • Dinner held within its normal range — the drop is concentrated in lunch
    • 36 fewer checks than usual; average check roughly flat

    Next stepReview lunch staffing and any nearby lunch-traffic changes with the GM; confirm whether the patio closure affected seating.

    The drop in traffic is measured; the cause is not. This is a prompt to look, not a conclusion.

  2. Priority

    Beverage units are 29% below baseline

    • 132 beverage units vs a median of 186.5 across 8 comparable Fridays
    • Category mapping is stable across the comparison window

    Next stepCheck beverage availability and whether servers are offering drinks at greeting; compare against the last menu change.

    The data cannot tell a stockout from a change in how drinks are offered. Check on the floor.

  3. Watch

    The whole day landed below baseline, driven by lunch

    • Net sales $3,569 vs a median of $4,200
    • Lunch accounts for most of the shortfall

    Next stepNothing separate. It resolves with the lunch finding above.

    Same evidence as the lunch finding, not counted twice.

Every figure is worked out by code from the imported data, then checked. The writing only ever explains numbers that were calculated, and each finding says what the data cannot tell you.

How it works

  1. Connect it, read-only

    A read-only key from your own POS account, with instructions on screen for each system. On Square it takes about five minutes; Toast issues one on request. It reads sales and labor, and never changes a schedule, a menu, payroll or anything in your POS.

  2. Tell it what matters

    Your hours, your dayparts, and the two or three things you are worried about, in your own words. “Lunch labor keeps running high” becomes the checks it runs every morning.

  3. Read it with your coffee

    Each morning, yesterday against comparable days: what moved, the numbers behind it, and one thing to check. On a quiet day it says so and stays short.

What it reads

Where each system sits in the pilot today. Nothing on this list is live for customers yet; this is the order they come online.

First in the pilot

Square · Toast

Connects directly

Clover · Timecards in the POS · 7shifts

Through a daily report

NCR Aloha · SpotOn · Oracle Micros

If your plan includes API access

HotSchedules · Homebase · When I Work · Deputy

Join the pilot

Tell me what you run and I’ll tell you where you fit. When your system is ready to connect, I’ll get in touch and set it up with you. Joining is free and commits you to nothing; you’ll see the price before anything goes live.

Optional. For example: why lunch labor keeps running high, or which store slipped yesterday.

Free to join. No card, and nothing is billed.

Reading this as