Platform
Participating restaurants contribute trading data. RMI validates, standardises and anonymises it, aggregates it into eligible competitive sets, and returns market context that no single business can produce on its own.
What RMI does
The platform is deliberately narrow in V1. It measures a small number of universally understood restaurant metrics accurately, rather than a wide set imprecisely.
Validation
Submitted data is checked against historical ranges and internal reconciliation rules.
Standardisation
Revenue definitions, trading dates and service periods are normalised across operators.
Anonymisation
Restaurant-level records are separated from identity before aggregation.
Comp-set aggregation
Eligible competitive sets are built to confidentiality thresholds.
Benchmarking
Each restaurant is measured against its comp set on revenue, covers and average spend.
Reporting
Daily, weekly, monthly and quarterly market intelligence is returned to participants.
Measured
RMI reports the metrics operators already manage against. Where the underlying dataset does not support a measure for a given comp set, it is suppressed rather than estimated.
Net revenue, current period, prior period, prior year, comp-set movement and index.
Total covers served, with comp-set movement and Cover Index.
Revenue divided by covers, with comp-set movement and Spend Index.
Analytics
Every chart in RMI sits against the 100 line. Absolute movement, market movement and the relative gap are always shown together. The examples below use illustrative figures.
From signal to action
RMI separates measured evidence from suggested follow-up. It identifies where to look; operators decide what changed and what action is appropriate.
01
Latest gap +4.1pts
Identify which sites and markets are contributing most to the gain.
02
104.0 Revenue Index
Track the next reading before treating normal variation as a change in trend.
03
Covers vs market · £85.4k
Use the bridge to separate market movement from covers and spend before assigning ownership.
04
Edinburgh strongest · Leeds weakest
Compare the sites behind Leeds with the pattern in Edinburgh.
YoY %, lines · variance in points, bars · Illustrative
AnswersAre we gaining or losing ground, and by how many points each month?
Variance = Portfolio YoY − Comp Set YoY
£k · Illustrative
| Component | £k | % of PY |
|---|---|---|
| Prior-year revenue | 1,842.0 | — |
| Market | +36.8 | +2.0% |
| Covers vs market | +85.4 | +4.6% |
| Spend vs market | −9.8 | −0.5% |
| Current revenue | 1,954.4 | +6.1% |
Observation: covers accounted for the larger share of the movement relative to the market. The data does not indicate why.
AnswersHow much of the movement was the market, covers or spend?
PY + Market + Covers + Spend = Current
Mean 103.4 · σ 1.04 · Illustrative
AnswersIs the latest reading a trend or normal variation?
Band = mean ± 1 sample standard deviation
Points gained against comp set · Illustrative
AnswersHow much ground has the estate taken from, or given to, the market this year?
Cumulative = Σ monthly variance, pts
One point per restaurant · sized by revenue · Illustrative
AnswersIs a site winning on footfall, on spend, or both?
Axes cross at 100 = Comp Set
100 = Comp Set · Illustrative
| Mon | Tue | Wed | Thu | Fri | Sat | Sun | |
|---|---|---|---|---|---|---|---|
| Lunch | 97 | 99 | 100 | 102 | 104 | 101 | 106 |
| Dinner | 95 | 98 | 101 | 103 | 108 | 110 | 99 |
| Combined | 96 | 99 | 101 | 103 | 106 | 106 | 103 |
Weakest cell: Mon dinner (95). Strongest: Sat dinner (110). Cells with fewer than the minimum sample would be suppressed.
AnswersWhich trading patterns are ahead of or behind the market?
Index = own ÷ comp set, matched day and service
Min · Q1 · median · Q3 · max across your restaurants · Illustrative
AnswersIs the gap between our best and weakest site in a market widening?
Box = interquartile range; line = full range
Trailing 7 months · Illustrative
| Market | Rev YoY | Comp Set | Pts | Rev Idx | Cov Idx | Spd Idx | Trend | Confidence |
|---|---|---|---|---|---|---|---|---|
| Edinburgh | +9.4% | +3.0% | +6.4 | 106.2 | 108 | 100 | High | |
| Manchester | +6.8% | +2.6% | +4.2 | 104.1 | 104 | 101 | High | |
| Bristol | +3.1% | +1.3% | +1.8 | 101.8 | 101 | 101 | Medium | |
| London | +1.9% | +2.3% | −0.4 | 99.6 | 99 | 101 | High | |
| Leeds | −0.6% | +3.4% | −4.0 | 96.1 | 97 | 99 | Medium | |
| York | −2.4% | +1.1% | −3.5 | 96.5 | 96 | 98 | Developing |
AnswersWhere in the estate is the movement, and how reliable is the benchmark?
Revenue Index = 100 × (1 + own) ÷ (1 + comp set)
Getting data in
Most operators begin with a weekly CSV or a simple manual submission. EPOS, reservation and accounting integrations reduce effort over time, but they are not a prerequisite for participation.
Questions
If your question is not answered here, the Methodology page covers how comp sets are built and how the indices are calculated, or you can request a call.
RMI benchmarks your restaurants against a defined competitive set. You submit trading data; RMI validates, anonymises and aggregates it, and returns how you performed relative to comparable restaurants in your market on revenue, covers and average spend.
Six fields per trading day per restaurant: trading date, restaurant, net revenue, covers, food revenue and beverage revenue. Most operators begin with a weekly CSV or a simple manual submission — integrations are optional, not a prerequisite.
Nobody outside your organisation. Restaurant-level records are anonymised before aggregation, and only your own performance, aggregated comp-set performance and market indices are ever reported. An individual competitor's performance is never visible to any participant.
Competitive sets are built from geography, format, average spend, cuisine, seats, service style and trading pattern. RMI proposes the classification during onboarding and you confirm it.
Every comp set must meet minimum participation and ownership-diversity thresholds before it produces a benchmark. Where those thresholds are not met, the figure is suppressed rather than published.
An index of 100 means you performed exactly in line with your comp set. Above 100 means outperforming the market, below 100 means underperforming — measured in points of relative performance, not percentages.
Benchmarking is calculated daily. Participants receive a weekly performance report, with monthly, quarterly and annual market reviews produced from the same data.
Operator access is £149 per restaurant per month, with group and institutional access priced on application. Figures are indicative and confirmed during onboarding; Founding Partners receive preferential terms.
The demonstration environment contains a fictional multi-site operator with twelve months of coherent trading data across nine markets.