The latest trading day at a glance, measured against how that weekday normally performs. Gross is what went through the till; net excludes VAT and is what the business actually keeps. Day net profit is that day after every cost, including one day\u2019s share of the monthly rent, overheads, loan and asset finance.
Last 14 days
Daily net sales with the breakeven line.
This weekday over time
How today compares with the same day in recent weeks.
Top sellers
Week to date
Monday to today, against the same point last week.
Overview
The whole trading position for the selected period: sales, margin, cost structure and where you sit against breakeven.
Net sales is the figure that matters. Every margin, target and breakeven number on this portal is measured against net sales, excluding VAT, because VAT is collected on HMRC's behalf and never belongs to the business. VAT is taken per transaction straight from the till rather than assumed at a flat rate, because much of the takeaway trade is zero-rated: the effective rate has ranged from 4.3% to 13.6% across individual days, with a median near 9.6%, moving with the eat-in and hot-food mix.
Loading live data from the connected Google Sheet…
Sales against breakeven
The line is the net sales per day needed to cover every cost including rent, loan and asset finance. Bars above the line are profitable periods.
Where every £1 of sales goes
Cost structure for the selected period.
Margin trend
Gross margin after COGS and labour, and net margin after everything.
Sales
Revenue detail by period and by day of week. Always read per trading day rather than monthly totals, because month lengths and closures distort comparisons.
Sales per trading day
The reliable trend measure.
Average basket
Falling basket with steady orders means people are buying less each visit.
Orders and average basket
Separates a footfall problem from a basket problem: falling orders is a footfall issue, a falling basket is a spend issue.
Day of week
Averages across the selected range.
Labour
Wage efficiency and hours. The loaded figure applies a 1.18x on-cost estimate for employer NIC, pension and holiday pay — replace this with your actual payroll on-cost when known.
Wage % and hours paid
Hours bars against the wage percentage line.
Sales per labour hour
Productivity measure. Harder to game than wage % when sales swing.
Hours and cost by team member
Whole period, 20 Apr to 17 Aug.
Sales and labour cost
Bars are sales, the line is labour cost. When the gap narrows, margin is under pressure.
Wage % of net sales
Base pay against the working target.
Cover
Effective orders per hour by day, from 19 July onward. Eat-in orders are weighted more heavily than takeaway because they carry crockery, table clearing and washing up. One person sustainably handles about 12 effective orders an hour, so any cell above that needs a second person.
How to read this. Each cell is takeaway orders plus eat-in orders multiplied by the eat-in weighting. Amber is approaching the solo ceiling, red is past it, and the width of a red block sets the second person’s shift length. Hover a cell to see the raw split. The weighting is an estimate, not a measurement — it is a setting in the Google Sheet, so raise it if eat-in involves table service and clearing, or lower it if it just means a mug instead of a cup.
Orders per hour heat map
Averages per trading day. Green is comfortable for one person.
Two-person windows
Profit & loss
Full trading position, monthly or weekly. COGS is theoretical, from recipe cost cards and Goodtill product costs. Labour is actual paid hours with employer on-costs applied. Overheads and financing are entered monthly.
Reading this. Theoretical COGS assumes no waste, over-pouring or staff drinks, so real cost of goods runs higher. A part-month shows full-month overheads and will therefore look worse than it finishes.
Profit & loss
Where the money goes
Each cost line as a share of net sales, latest full month.
Net profit by month
After all costs including loan and asset finance.
Breakeven
Net sales per trading day needed to cover every cost, at each month's cost base.
Menu margins
What each drink costs to make against what it sells for. Costs come from the recipe cards, ingredient invoices and packaging prices; selling prices are the actual average achieved.
These are theoretical costs. They assume no waste, no re-makes, no staff drinks and perfect portioning, so real margins run several points lower. Use them to compare drinks against each other rather than as an absolute figure. Ingredient prices are a snapshot and need revisiting when suppliers move.
Margin by drink
Gross margin percentage, lowest first.
Where the profit comes from
Total cash gross profit since opening, top ten.
Full cost cards
Sorted by cash contribution. Sold = units since opening.
Benchmark
The prior operator traded this same site from October 2024 to February 2026. Her figures are shown per trading day, because she opened fewer days over summer and monthly totals would mislead.
Read this with care. She served an older resident demographic with no student or young-professional trade, opened roughly 46 hours a week against our 73, and appears to have run the site solo with no payroll. Her seasonal shape transfers; her cost structure does not.
Revenue per trading day
Her 2025 line against our 2026 actuals.
What her data tells us
Her per-day trade held flat from May to July, dipped about 20% in August, then recovered to her strongest day rate of the year in September.
September was her second-best month across sixteen months of trading, consistent with an academic-calendar catchment.
She stopped opening Sundays entirely from July onward, and closed at 4pm on weekdays — her 7am and 4pm hours were under 3% of orders each.
She averaged 49 orders a day against our 90 to 100, and her busiest hour typically reached 11 orders, comfortably inside solo capacity.
She closed in February 2026 despite having no wage bill, which indicates the site cannot support a paid team at her volume.
How this portal works
A shared reference for all three partners. Everything here is computed from two exports.
Updating the data
Export transactions from Goodtill for the new period, and the timesheet detail from Deputy for the same dates.
Send both files across and the portal is rebuilt with the new period included.
The Excel KPI pack can alternatively be updated by hand — paste new daily rows into its Data tab and every other tab recalculates.
The measures that matter
Wage % of net sales — the core efficiency measure. About 30 to 32% is realistic at current trade; 25% is not achievable while keeping two people across the lunch peak.
Sales per labour hour — productivity, and more reliable than wage % when sales move sharply.
Sales per trading day — the only fair way to compare periods of different length.
Average basket — tells you whether a change in sales came from footfall or from spend per visit.
Assumptions in use
These drive every P&L figure. Edit them in the Settings tab of the Google Sheet and this portal picks them up on the next reload.
Known limitations
Actual hours ran roughly 250 above rostered plan across April to August, so any forecast built on the rota alone should carry an 8 to 9% upward variance.
The 1.18x on-cost multiplier is an estimate and materially affects the loaded wage percentage.
The franchise commission step-up at month 19 is not modelled here and must be added before medium-term planning.
Order preparation time is not captured by the till, so coverage figures are derived from order volume rather than measured service time.