Why us

Three things that are genuinely different.

01

Platform-side intelligence

We spent years working directly with restaurant operators from inside Uber Eats, across hundreds of independent businesses. That means knowing how promotional incentive structures are actually designed, which levers an account manager can genuinely pull and which are theatre. It is the difference between reading the platform's report and checking the platform's homework.

02

Behavioural KPI design

Most restaurant analysis optimises for covers and revenue — both outputs of customer behaviour rather than measures of it. We optimise for the rhythm of customer return: Days-Between-Visits, measured at brand level, channel-agnostic, detectable within weeks, and directly predictive of lifetime value.

03

AI-native execution speed

We produce what has traditionally required a team of three — data analyst, strategist, designer — as a single operation. Analysis runs in Python, so it re-runs on refreshed data instead of being rebuilt. You receive the findings, not the scripts — but the foundation is reproducible, and that is why the turnaround is what it is.

How we think

We measure the rhythm of return.

Days-Between-Visits is the proprietary KPI at the centre of our retention work: the median number of calendar days between a customer's consecutive completed visits, measured at brand level across all your venues. Covers and revenue are outputs of customer behaviour; DBV measures the behaviour directly.

The implicationIf your Regular customers currently visit every 90 days, a programme that compresses that to 60 days increases their annual visit frequency by 50%. That compounds across the full segment with no increase in acquisition cost.

Work

Built from real engagements.

CRM & retention strategy — 3-venue restaurant group, London

3 months · Stage 1 complete and Stage 2 delivered
64,000+ reservation records13,448 customers segmented3 data sources integrated
Headline finding

50.6% of Uber Eats orders were carrying a restaurant-funded discount. Promoted baskets were larger — a median of £37 against £25.50 — but cash to merchant per order was only fractionally higher. The promotions were subsidising baseline volume, not driving incremental orders.

Scope

What we don't do.

We are not an agency. We do not run your campaigns, manage your social media, or manage your platform accounts. We do not resell software or take commission from any vendor we recommend. We design the logic, the sequence and the success conditions; your team runs it — a framework your people own outlasts a retainer.

If these are the problems you're working on, let's talk.

We take on a small number of engagements at a time. The first conversation is a scoping call — bring a month of exports if you want something concrete to look at.