What a loyalty programme is really worth.

A New Zealand apparel retailer paid about $1m a year to belong to a national shared rewards programme, and could not say what it earned back. I was the retailer's CRM Manager, and I pulled and analysed the customer data for the project that answered the question. A loyalty consultant designed the approach.

Objective

Earlier attempts to value the membership had been inconclusive. The brief asked for an answer in dollars: is the return adequate, and if it is not, why not? Even if it was adequate, where could the programme be made worth more? A second aim was to leave the retailer's own staff able to repeat the analysis on its other market.

Method

The work compared customers who belonged to the programme with customers who did not, using the retailer's own customer database, and put a dollar value on each difference.

• Like with like. Members were more likely to be long-standing customers, and long-standing customers spend more, so a straight comparison flatters the programme. We adjusted for age and tenure.
• A cohort followed through time. Every customer on the books in the first year of the data was tracked through the years after and sorted into retained, reactivated, inactive or dormant.
• Customer research. New and existing customers were asked how much the programme mattered when they chose to shop there.
• A five-year model. Extra retention, extra spend, customers who came through the programme's channel and the other sources of value were priced, less the cost of the points and other costs. The result was taxed, discounted at 12% and tested for sensitivity.

What the analysis found

The raw gap was two and a half times the real one

Members spent 50% more than other customers. Allowing for age and tenure, that fell to 30%. Members were also twice as likely to be in the retailer's top-customer programme (20% of members against 10% of non-members), which brings extra marketing and extra sales. Taking that out left 20%. Only the 20% was valued.

Members bought more often

The extra spend came almost entirely from a higher order frequency, and the gap showed at every length of tenure.

Retention was higher

Annual retention was 84% for established members against 79% for non-members, and 78% against 70% for new customers.

Few people joined for the points

Among new customers who were members, 10% said the points were very important in choosing the retailer, and 40% had not known it offered the programme when they decided. In the research, 59% of members named the programme as the retailer's main loyalty programme, so 41% were earning points without recalling the benefit.

More bonus points did not mean more sales

Items on bonus points sold more, but they sold more by a similar amount to non-members, so the points may not have been the reason. Earlier trials pointed the same way: a minimal bonus was enough to get the likely maximum response.

The biggest number rested on an assumption

About $9m of the roughly $22m of benefits, in present-value terms, came from existing customers assumed to be at the retailer only because of the programme. The research could not ask people why they started shopping years earlier, so the model assumed that about 23% of customers had been influenced that way, the share seen among newer customers. The model's own sensitivity test shows how much hangs on an input like this: moving the assumed sales boost by 10 points moved the after-tax result by about $4m.

The study valued staying in, not the alternatives

It did not test what would happen if the retailer left, or whether the $1m could earn more somewhere else.

Outcome

The base case came to a present value of about $15m before tax and about $10m after tax over five years, after the cost of the points and other programme costs. The advice was to stay in the programme and make more of it:

• Show the points at every purchase and in every promotion.
• Build a plan for customers who are not motivated by the programme. If they matched the retention and spend of members, that would be worth about $3m a year before tax.
• Stop new-customer promotions reaching existing customers.
• Move towards offers aimed at individual customers, because a blanket points scheme also pays customers who say it does not motivate them.
• Measure it again each year, using internal data and a small customer survey.

Questions to ask of your own programme

• After allowing for how long members have been customers, do they still spend more than non-members?
• Do members order more often, or spend more per order?
• How much of the claimed value was measured, and how much was assumed?
• Do members know they are in it?

The customers page has more on where customer value slips away. The free 45-minute chat ends with a one-page readback of where I think the money is most likely being lost in your business. Ask for it here.

More on where I look