How to tell if your loyalty program is working

A visit, everywhere on the dashboard, is one purchase recorded for a member, normally by staff at the Counter. If nobody records a member's purchase, that member has not visited as far as these figures can tell. Spend is narrower still. Home's Member value view says: "This is the qualifying amount recorded for members, not your total takings."

The Counter asks for an amount only when the card earns from what members spend: Points or Qualifying spend. On a card that counts only Stamps or Visits, staff record the purchase with no amount, so the ringgit figures have nothing to add up. That is settled at setup: a live program's earning rules cannot be changed.

The switch at the top of Home sets the period, 30 days or 90 days, and compares it with the period just before. Of the ten figures here, only Came back, Visits per visitor and From returning members follow it. The rest are fixed by their own definitions and do not move with it.

Came back. "Of members who visited in the 30 days before, the share who visited again." They visited in the earlier 30 days and again in the latest 30.

A fall is not always bad news. An example with made-up numbers: 100 members visit in a normal month and 55 come again the month after, so the figure reads 55%. Then a promotion brings in 60 new members who each buy once, and 160 members visit that month. If the usual 55 return and 9 of the newcomers do, the next reading is 64 of 160, or 40%. No regular has left. The line under the figure gives both counts. While the group is small, read the counts before the percentage.

The period can mislead too. A salon client who comes every six weeks can be entirely loyal and still miss a 30-day block. If your customers come less often than monthly, read this figure on 90 days.

Second visit, on Home's Retention view. "Of members who joined 2 to 13 weeks ago, the share who have visited twice or more." The group rolls forward, so the figure shifts when a large batch of joiners enters or leaves it. It also measures your counter. A member who joins while paying has made a first visit only if staff record that purchase after the member has joined. Where staff skip that step, the member needs two further visits to count. So a low figure can mean new members are not returning, or that joining purchases are going unrecorded. How to get new customers to come back a second time covers what to do between the first visit and the second.

The third is the stages. Home puts each member in one of five:

Watch the movement: how many slid from Active to At risk, and whether Never visited is growing. If it is, people are joining and no purchase is being recorded for them. You can open a stage to see who is in it. Only one of the five can be sent a notification on its own: New members, described as "Joined in the last 30 days.", is a choice under Who gets it. The other stages are not.

The trap is that the lines sit at 30 and 60 days for every business. Where regulars come every five or six weeks, some of your best customers are in At risk on any given day.

These move slowly, or for reasons built into their definitions. Read weekly, they are mostly noise.

Return window is your own clock. It is worked out from the gaps between each member's visits over the last 180 days: "Nine in ten return visits come within this many days. A regular away longer is overdue." Hold the stages against it. If it reads 12 days, a member in At risk is long overdue. If it reads 45, At risk is partly routine. Its limit is that it is built only from returns that happened. A member who never came back adds no gap, so it says how soon returning members return and nothing about how many do. It can lengthen for a good reason, too: occasional customers who start coming back bring long gaps with them.

Visits per visitor is "Average visits by the members who visited at least once in the last 30 days." Only members who came are in it. When occasional visitors stop coming they drop out of the average, and it rises. Read it beside the number of members who visited, at the top of Overview. More visits each, from fewer people, is a shrinking program.

Average member value and From returning members, on Member value, are the two that flatter.

The first is "Total spend since joining, divided by members who have spent at least once." A member's total only grows with time. So the figure rises as the program ages and dips when a wave of newcomers joins, whether or not anyone's habits change. Members with no recorded purchase are left out. It is spend so far at your prices. It is not profit and not a forecast.

The second is "The share of this period's spend from members who had visited before it began." Stop recruiting and it heads for 100%. A fall during a month of heavy joining is what growth looks like.

One study supports reading both with suspicion. It followed customers of a convenience store franchise. Those who were heavy buyers when its program began "were most likely to claim their qualified rewards, but the program did not prompt them to change their purchase behavior". Customers who started as light or moderate buyers "gradually purchased more and became more loyal to the firm". That is one franchise, read in the paper's abstract, which gives no sample size. So treat a high share from returning members, and the names under Top members, as a picture of who was loyal already. How to test whether your program makes a difference shows how to test the rest.

Have your card and Can get offers, on Audience, have to be multiplied. The first is "Members with an active Wallet Card. Confirmed on a phone means Apple or Google reported it added." It depends on whether a new member adds the card before walking away. The second is "Card holders with offer notifications on. Balance updates still reach every card." That is a share of card holders, not of members. An example with made-up numbers: of 400 members, 70% have the card (280) and half of those can get offers (140). An offer reaches at most 140 of your 400. A reward reminder or business news is not held to the second figure. If Can get offers reads zero, the program may not be asking. The setup step Offers and news, if you send them says: "Leave it blank and people are simply not asked."

Time to earn it is in the program: open it from Programs and choose Rewards. It is shown "For a typical returning member". It is a projection from how often returning members come and what a visit earns them. Members who came once are not in it, so it is the wait for someone who already has the habit. First claim, beside it, is the observed figure: "Typical time from joining". When Time to earn it passes eight weeks, the Rewards view marks the figure, the program's Overview lists the reward by name with a link, See why, and Home may suggest Add a closer reward. How soon should customers earn their first reward? explains why the distance matters.

Customers shows the first of these, Came back within 30 days, for all your members, worded as "the share who made a second visit within 30 days".

Home's figure covers everyone who visited in the earlier period, regulars included. So the three will not agree, and none is wrong. The Campaigns figure is the subject of Which sign, link or post brings you the best members?

Came back needs an earlier period with visits in it. Second visit needs members who joined at least two weeks ago. Return window reads "Needs more return visits to measure". Time to earn it needs 10 returning members. In a program's own Overview, Came back within 30 days needs 10 first visits, Days between visits 5 regulars, and Typical spend per visit 20 visits.