Inventory
Guide · 14 min
The variance report: finding the 3% that never becomes revenue
Theoretical versus actual consumption, the four causes behind most gaps, and healthy ranges by format.
2 Sep 2026Read
Resources
How restaurants actually use their numbers.
Working guides, teardowns and product notes from the floor — written for operators, not for search engines.
Featured guideInventory · 14 min read
The variance report: finding the 3% of stock that never becomes revenue
Theoretical consumption versus actual, line by line — how to read the number, what a healthy range looks like by format, and the four causes behind almost every gap.
- Read variance per ingredient, not per invoice
- Separate theft, over-portioning, yield loss and data error
- Set an alert threshold your outlet can actually hold
- Worked example on a 3-outlet biryani menu
Healthy variance by format
QSR1.5–2.5%
Cafés1.0–2.0%
Casual dining2.0–3.5%
Cloud kitchens2.5–4.0%
Resource library
9 pieces in this view
Restaurant Intelligence
Teardown · 11 min
Why "sales were down" is never the answer
Walking a real 8% dip down to lunch traffic at one outlet — and what the owner changed the next day.
28 Aug 2026Read
AI & ML
Explainer · 9 min
What a confidence band actually tells your kitchen
Why a forecast of 198–232 covers is more useful than a forecast of 214, and how to prep against a range.
21 Aug 2026Read
Restaurant Operations
Playbook · 16 min
Cutting eleven seconds off every bill at peak
The billing-screen audit: taps per order, modifier depth, and the three settings that move the number.
14 Aug 2026Read
Customer Intelligence
Playbook · 12 min
Win back a regular without a discount
Visit-gap thresholds by format, what to say at each tier, and why the offer matters more than the amount.
7 Aug 2026Read
Industry Trends
Analysis · 10 min
Aggregator commission is a menu problem, not a pricing problem
How channel-specific menus and combo design change contribution margin more than a blanket price rise.
31 Jul 2026Read
Case Studies
Case study · 8 min
Three outlets, one definition of a good Friday
A multi-outlet group moving from three reporting standards to one, and what the comparison revealed.
24 Jul 2026Read
Guides
Guide · 18 min
Recipe costing from scratch, including the sub-preps
Yields, wastage factors and nested preparations — the setup that makes plate cost trustworthy.
17 Jul 2026Read
Product Updates
Release note · 4 min
Ingredient forecasts now write straight into purchase orders
Forecast to draft PO in one approval, with vendor price comparison attached.
10 Jul 2026Read
Product updates
What shipped recently
10 Jul 26
Forecast → purchase order in one approval
Ingredient forecasts now draft POs with vendor price comparison attached.
2 Jul 26
Copilot answers now cite their source screens
Every number in an answer links to the report it came from.
24 Jun 26
15-minute demand blocks for QSR
Peak shape is modelled per block rather than per day.
30 May 26
Group P&L across mixed formats
A café, a QSR and a dine-in room normalised into one statement.
Operator benchmarks
Ranges worth measuring against
Working ranges we use with operators during setup — targets to aim at, not industry claims.
Bill time at peak (QSR)under 30s
Stock variance1.5–3.5%
Table turn (casual dining)55–75 min
Repeat rate, 30 days (café)45–60%
Aggregator commission drag16–22%
Demand forecast error5–9%
Get the operator brief every second Tuesday.
One teardown, one number worth watching, one thing we shipped. No newsletter padding.