How to size a UK loyalty scheme with ONS income, retail sales and inflation data

What to take away

Which ONS datasets size a UK loyalty scheme

A UK loyalty scheme business case rests on four ONS datasets. Each answers a different question about how much members earn, spend, and trust.

The ONS average household income dataset provides the baseline for reward economics. It reports median and mean disposable income by household type, region and decile. The exact release is the ONS Household disposable income and inequality bulletin. Use the median, not the mean, because a small number of very high incomes skew the average.

For a scheme targeting Greater London, the median is higher than in Wales or the North East. That changes what a £5 reward is worth to a member.

The ONS retail sales bulletin gives monthly volumes and values for Great Britain. It shows how much is spent in stores and online, seasonally adjusted. Use it to estimate the size of the transaction pool your scheme can capture. A scheme in South East England can size its addressable spend from the retail sales value for the relevant sector.

The exact release is the ONS Retail sales, Great Britain bulletin.

The ONS consumer price inflation bulletin supplies the CPI and CPIH indices. These tell you how fast reward costs are rising. The exact release is the ONS Consumer price inflation, UK bulletin.

The ONS personal well-being dataset measures life satisfaction, anxiety and trust. It matters because scheme design that feels exploitative damages brand trust. A wellbeing-aware scheme uses rewards that members value, not just discounts. The exact release is the ONS Personal well-being in the UK bulletin.

The ONS annual survey of hours and earnings gives wage benchmarks by occupation and region. Use it to check whether your target member can afford the spend threshold you set. The exact release is the ONS Annual Survey of Hours and Earnings (ASHE) bulletin. For full details, see the average household income data.

Turning household spending into reward economics

A reward economics model converts household spending into points liability. Start with the ONS average household income dataset. Take median disposable income for your target region and household type. Deduct essential spending such as housing, food and utilities. What remains is discretionary income. That is the pool your scheme competes for.

Assume a realistic share of discretionary spend goes through your scheme. If a household has £400 a month discretionary and you capture 5%, that is £20 of scheme spend. At a 1% reward rate, the member earns 20p in value. That is too low to change behaviour.

At 5%, they earn £1. The reward must be worth more than the effort of joining.

Now build the cost side. For every £1 of member spend, you issue points worth a set percentage. If the reward rate is 2%, you owe 2p per £1. If 30% of points are never redeemed, breakage reduces the liability. But breakage is not free money. It is a liability that may return.

Use the return on investment needs a cost model to avoid treating breakage as profit. A reward economics model should show gross reward cost, breakage, redemption cost and VAT. Points are a supply of services for VAT purposes in the UK. If you issue points on which VAT is due, the liability includes VAT.

Worked example: a regional coffee chain

A coffee chain with 40 shops in the North West wants a points scheme. Median household disposable income in the region is around £32,000 a year. The chain's average transaction is £4.50. A member visits twice a week.

Annual member spend is £468. At a 3% reward rate, the annual reward cost is £14.04. If 25% of points break, the expected cost falls to £10.53. The chain needs a gross margin above 60% to absorb that. The model shows the scheme is viable only if visit frequency rises by at least 8%.

Using retail sales and inflation figures together

Retail sales volumes show how many transactions happen. Inflation shows what those transactions cost. Used together, they size the loyalty-driven activity you can win.

The ONS retail sales bulletin reports values and volumes. If values rise but volumes fall, inflation is driving the numbers. Your scheme should not assume that higher spend means more loyal customers. It may just mean higher prices.

The ONS consumer price inflation bulletin gives the CPI. Use it to index your reward budget. If CPI is 3% and your reward budget is fixed, your real reward value falls by 3% a year. Members notice. A scheme that does not index rewards loses its appeal.

Combine the two: take the retail sales value for your sector, adjust for inflation, and you have a real growth figure. That figure is the realistic ceiling for scheme-driven sales growth.

If real retail sales are flat, a scheme promising 20% growth is promising to take share from competitors, not to grow the market. That is possible, but it needs a bigger budget.

Use the budget template without guesswork to link these figures to a monthly cost plan. A common error is to set a reward rate from last year's CPI. Check the latest consumer price inflation data before you sign off.

Reading the two bulletins side by side

Here is a worked calculation. Suppose retail sales values in your sector rise by 4% over a year while volumes fall by 1%. The ONS Consumer price inflation, UK: April 2024 release, published 22 May 2024, reported CPI inflation of 2.3%.

Deflate the value growth by CPI: 4% minus 2.3% gives about 1.7% real value growth. Falling volumes mean fewer transactions. A scheme promising 10% sales growth is mostly share gain, not market growth.

Segmenting members with income and wellbeing data

Not all members respond to the same reward. Income and wellbeing data let you segment without guessing.

The ONS average household income dataset breaks income down by decile. The top decile has several times the disposable income of the bottom. A £10 voucher means little to a high earner and a lot to a low earner. A flat reward rate is regressive: it gives more value to those who need it least.

The ONS personal well-being dataset adds a second dimension. It measures anxiety, life satisfaction and feelings of worth. Members with low wellbeing may respond better to experiential rewards or charitable donations than to discounts. Members with high wellbeing may value convenience and status.

The ONS annual survey of hours and earnings gives occupation-level pay. Use it to build personas. A nurse in Scotland and a software developer in Greater London have different pay and different reward preferences. The key metrics without vanity numbers helps you track segment-level performance rather than a single average.

Segment by region too. Median income in the South East is higher than in the West Midlands or Yorkshire and the Humber. A national scheme with one reward rate will overpay in some regions and underpay in others.

Regional reward tiers are legal under UK consumer law if the terms are clear and not misleading. The CMA expects transparent terms under the Consumer Protection from Unfair Trading Regulations.

Wellbeing indicators that matter for scheme design

Use the personal well-being data to test reward concepts before launch. A reward that scores well on satisfaction but poorly on anxiety may still fail.

Building scheme sizing figures you can defend

Scheme sizing figures are estimates. They must be defensible to a finance director. Build them from ONS data, state your assumptions, and test them.

Each step needs a source. Population from ONS. Frequency from retail sales. Reward cost from your margin. If a step has no source, mark it as an assumption and test it.

Use the ONS annual survey of hours and earnings to cross-check affordability. If your target member earns £25,000 and your scheme requires £500 annual spend to reach a reward, that is 2% of gross pay. That may be too high.

The table below is an illustrative end-to-end sizing calculation. It uses round assumptions and is not a forecast.

Checklist for a defensible model

For wage benchmarks, see the employee earnings data.

Where loyalty software assumptions break the model

Loyalty software platforms come with default assumptions. Some are wrong for the UK.

Default reward rates are often set for US retail, where margins and sales taxes differ. A 1% reward rate that works in the US may be too low in the UK, where VAT is included in displayed prices. Check the default against your own margin.

Default breakage assumptions are often too high. Platforms may assume 30% of points never redeem. In the UK, the Consumer Rights Act and CMA guidance on unfair terms make it harder to expire points without clear notice. If your terms are unfair, breakage may be zero.

Default data consent flows often ignore PECR and UK GDPR. Loyalty software must collect consent for marketing emails and texts separately from scheme terms. The ICO enforces this. A platform that bundles consent into sign-up is a compliance risk.

Default reporting often shows points issued and redeemed, not incremental profit. Define incremental sales and reward cost per member in your own reporting. A high redemption rate is not a success if the reward cost exceeds the margin.

Finally, default assumptions about member income are often too high. If your software assumes an average income above local ONS income levels for your region, your reward tiers will be too generous. Check the ONS average household income dataset for your target area before you launch.

Common software assumptions and UK corrections

Common questions