When a founder or marketing lead asks me whether a loyalty pilot can pay back within 90 days using only till (point-of-sale) data, my first response is: yes — but you need a tight, pragmatic framework and a few realistic assumptions. I’ve run and modelled dozens of small pilots for retailers and hospitality operators where the only reliable data source was the tills. If you know what to measure and how to calculate incremental value, you can make a credible, evidence-based case in under a week.
Why focus on till data?
Small teams often don’t have a single customer view, CRM, or clean online tracking. Tills are the least sexy asset, but they’re consistent: transaction timestamps, basket totals, SKU or category codes, and—if you capture them—tender type and receipts with loyalty identifiers. That’s enough to estimate short-term revenue uplift, repeat behaviour and payback. Using tills keeps the pilot simple, low-cost and fast to prove.
The simple logic I use
My approach is built around three questions:
- How many incremental transactions will the pilot produce within 90 days?
- What is the average incremental margin on those transactions?
- What is the cost of the pilot (reward cost + setup + marketing) within 90 days?
If incremental margin > pilot cost within 90 days, the pilot pays back.
Key till-based metrics to extract
From the till data you’ll need:
- Total transactions (baseline): number of sales in a comparable pre-pilot period (e.g., prior 90 days).
- Transactions by customer cohort: if you can flag loyalty-enrolled receipts (barcode, phone number, email at till), separate enrolled vs non-enrolled.
- Average transaction value (AOV): average basket size for baseline and during pilot.
- Return frequency: proportion of customers who make a repeat purchase in the 90-day window.
- Product or category margins: to estimate incremental gross margin on extra spend.
Designing the pilot so tills can prove it
When I design pilots that will rely solely on tills, I insist on three operational requirements:
- Enable a simple loyalty identifier at tills (phone number, email, or loyalty card barcode) so you can tag enrolled customers’ transactions.
- Run the pilot for at least 90 days so behaviour changes (repeat visits) can show up in the data.
- Keep the proposition clean and measurable — e.g., "Join and get a 10% off next visit" or "Earn a £5 reward after three visits".
Example: a 90-day payback model using only till data
Below is a compact example I often walk clients through. Replace numbers with your tills’ actuals.
| Baseline 90-day transactions | 10,000 |
| Baseline AOV | £20 |
| Baseline repeat rate (90 days) | 20% |
| Target enrolment (pilot) | 1,000 customers |
| Expected conversion to another visit within 90 days (uplift) | +15 percentage points (from 20% to 35%) |
| Average incremental visits per enrolled customer (90 days) | 0.15 visits |
| AOV on incremental visits | £22 (assume slightly higher) |
| Gross margin | 45% |
| Reward cost per redeemed reward | £3 |
| Marketing/setup cost (90 days) | £1,500 |
Calculation steps:
- Incremental visits = enrolled customers × incremental visits per customer = 1,000 × 0.15 = 150 visits.
- Incremental revenue = 150 × £22 = £3,300.
- Incremental gross profit = £3,300 × 45% = £1,485.
- Reward redemptions: assume 60% of incremental visits redeem the reward = 90 redemptions × £3 = £270 cost.
- Total pilot cost = reward cost £270 + marketing/setup £1,500 = £1,770.
- Net incremental margin = £1,485 − £1,770 = −£285 (in this illustration the pilot slightly misses payback).
That example helps you immediately see the levers: increase enrolment, reduce reward cost, improve conversion to visits, or target higher-margin items.
How to measure uplift in practice using till reports
Run two simple till reports:
- Enrolled customers’ behaviour (90 days): transactions, AOV, and repeat visits for customers who enrolled during the pilot.
- Matched historical cohort: customers who joined in a comparable prior period (or non-enrolled customers with similar baseline spend) to estimate what enrolled customers would have done without the pilot.
Compare the proportion who made a second purchase within 90 days, average visits per customer, and AOV. The difference is your observed uplift attributable to the pilot (with the usual caveat about attribution in uncontrolled environments).
Practical tips to increase the chance of payback within 90 days
- Choose a short, high-frequency reward: "Spend £30 and get £5 off next visit" nudges faster return than long-term points schemes.
- Promote at-point-of-sale: staff prompts and till receipts drive immediate enrolment and next-visit intent.
- Target high-margin categories: if tills allow, push redemptions toward items with stronger margins (add-on coffees, sides, accessories).
- Use an A/B style test in-store: run the offer in half your stores or random weeks to control for seasonality using till-level splits.
- Monitor weekly: export till data weekly and watch enrolment, redemptions and repeat visit rate — small changes early can flip payback.
Common pitfalls I see
Relying on tills is powerful but imperfect. The main issues I help clients avoid are:
- Counting total revenue uplift without isolating enrolled customers — you need to attribute to the pilot group.
- Underestimating reward redemption timing — some rewards are redeemed after 90 days.
- Ignoring operational costs — staff training time, till stickers, and receipt messaging add to pilot cost.
- Using unrealistic baseline comparisons — always use the most comparable pre-period or a matched control.
If you want, I can provide a simple spreadsheet template that automates these calculations from your till exports (transaction file + customer identifier). It’s the same template I use when I audit a program — fast to populate and great for convincing finance and the founder that the pilot is either worth scaling or needs tuning.
