The myth: Strong program metrics, high scan rates with active members spending more than non-members, are evidence that the program is driving good business performance.
By Tim Tyler, Published 3rd November 2026
The numbers from Australian retail loyalty programs lately have been remarkable. My Dan's, the Dan Murphy's program, reports 5.4 million active members, an 83% scan rate, and members spending 80% more per transaction than non-members. Myer One added 706,000 new members in a single year, bringing active members to 4.4 million and a tag rate of 77.2%. Super Retail Group reports 11.5 million program members, who account for 77% of group sales. Michael Hill grew its membership fourfold to 2.5 million, with members delivering 80% of sales at an average transaction value 80% higher than non-members.
Every one of those figures was reported alongside deteriorating company financials.
Endeavour Group — the parent of Dan Murphy's and BWS — posted a 15.8% fall in net profit after tax for FY2025, with EBIT down 7.3% and sales essentially flat at $12.1 billion. In the first seven weeks of FY2026, combined sales at Dan Murphy's and BWS fell by a further 1.3%. Their response was to invest in lower shelf prices and targeted promotions, which lifted Q2 retail sales 2.2% — but compressed gross margins in the process, with first-half pre-tax profit forecast well below the prior year.
Myer's loyalty program had an outstanding year. Myer, the company, saw gross profit fall by 2.5%, total sales down by 2.9%, and stores closed. Michael Hill grew its program membership by a factor of four. Michael Hill's EBIT fell 76%.
What's going on?
Two things, working together.
The first is self-selection. Customers already disposed to buy from a brand join its loyalty program at higher rates and bring more of their spend with them. The program captures committed buyers; it doesn't create them. When the program reports that members spend 80% more than non-members, it's largely describing who joined, not what the program did to their behaviour.
The second is a statistical artefact of how brand health moves. Brands grow by selling to light, occasional buyers, the large population of low-frequency customers who make up most of any category. When trading conditions soften, those light buyers pull back first. What remains is a concentrated core of loyal, high-value members, now representing a larger share of a smaller customer base. Program metrics look better, even as the business shrinks.
As Warren Buffett put it: Only when the tide goes out do you learn who's been swimming naked. Rising program scan rates while sales fall is what a falling tide looks like in retail loyalty data.
None of this means loyalty programs have no value. "Members spend more than non-members" is not evidence that the program is generating that difference, and it especially isn't evidence that the business is healthy. The correct question is whether the program produces incremental behaviour change that wouldn't have occurred without it. That requires an appropriate attribution methodology that delivers an honest accounting for selection effects. Most program reporting does none of those things.
The practical implication is you should make sure your program delivers value to light, uncommitted buyers, too, not just to the high-value advocates who were going to find you anyway. A program optimised entirely for its most loyal members is optimised for the customers it is least likely to influence.
Program metrics that look best when the business is struggling should prompt a question, not a press release.
We are Ellipsis, The Loyalty Experts. We help you measure, manage and grow customer loyalty. We're here to help, please get in touch.
Related Articles
