Customer retention for small businesses: the loyalty window is now months

Half of the executives surveyed for the BoF-McKinsey State of Fashion 2026 named customer retention as a top priority for the year. Most retention budgets are still spent on discounts at the point of sale. This article brings together four pieces of research published between May and July 2026 which, read together, point to a different approach: loyalty is built between purchases, in the company of other customers, and it can be shared with a partner business to win new customers. The article ends with three numbers to check this week and a practical plan for a business in Dubai or the UK.

Loyalty is won and lost in months

The BoF-McKinsey report Face to Face With Luxury Clients, published in June 2026 and based on a survey of about 2,000 luxury customers in the US and China, says that "the window between earning a client's loyalty and losing it has compressed from years to months." Two more findings from the same report change how I would spend a retention budget. The relationship with a sales adviser was the weakest reason for paying full price, given by only 13 to 19 per cent of customers, well behind wanting the item immediately at about 50 per cent. And 59 per cent of established customers regularly buy second-hand, so the second purchase may not be from the brand at all.

These findings apply well beyond luxury. What shortened the window is choice and information. A customer in any category can now find an alternative, compare it and buy it in an afternoon, and AI assistants are starting to do the comparing for them. A relationship that a business used to build over years now has to be maintained continuously, mostly at times when there is nothing to sell. The survey covers two markets only, and the report notes that its market data predates recent disruption in the Middle East, so the Gulf reading is my own.

Listen to the customers who score you five out of five

Marcus Buckingham, writing in Harvard Business Review in the May to June 2026 issue, uses Gallup research covering more than 1.8 million employees and tens of millions of customers. The relationship between how good an experience is and how customers behave afterwards is not a straight line. Behaviour barely changes as an experience improves from poor to acceptable, then changes sharply once the experience becomes very good. A four and a five on a five-point scale are different states of mind, and "top-two-box" scores, which add the fours and fives together, hide the only group whose behaviour changes.

The consequence for budgets is awkward. Most customer experience spending goes on fixing the problems of customers who score one or two. On this evidence, that spending changes very little behaviour, and the money is better spent on whatever produces fives. Buckingham's example is Kroger's Cincinnati and Dayton division, facing a new competitor, Publix, entering the market. Ratings of staff friendliness rose 45 points, cleanliness 28.9 points and staff turnover fell 10.9 per cent, through small, physical changes such as two cleaning ambassadors per store in distinctive jackets, every customer-facing employee carrying groceries to cars, and staff trained as certified cheesemongers. It is one division, with no control stores, and a competitor arriving at the same time, so the figures are indicative rather than proof. The instruction still stands: take your last customer survey, separate out the customers who scored you five, and read what they wrote in their own words. Those comments tell you what to build more of over the next twelve months.

Community matters more than influencers

The State of Fashion 2026 consumer survey found that 89 per cent of consumers say being part of a like-minded community strengthens their connection to a brand, compared with 67 per cent for influencers. 62 per cent feel an emotional connection to the brands they buy most often. In the luxury report, interest in brand-hosted social gatherings rises to 40 per cent among established US customers, and half of them want cultural events from the brands they buy from.

Put those findings next to the shortened loyalty window and the plan is clear. If loyalty has to be maintained between purchases, the question for a business owner is how many times last quarter your best customers heard from you without being asked to buy something. If the answer is none, the retention budget is being spent at the wrong moment. A gathering, early access, a conversation, a useful piece of content, an introduction to another customer: these cost little, and the research says they are what produce the fives.

Using another brand's loyalty to win customers

A new paper in the Journal of Strategic Marketing by Daeun Chloe Shin, published on 25 July 2026, tested four ways for an online start-up to win a new customer: a personal recommendation from a close friend, a recommendation included in the marketing email of a brand the consumer is loyal to, the same from a brand they are not loyal to, and a paid social media advertisement. Recommendations from brands the consumer was loyal to worked as well as a friend's recommendation. Recommendations from brands they were not loyal to worked no better than paid social advertising. I have read the abstract only, so the size of the effect is unknown to me. It is a single experiment and the results are almost certainly what people said they would do rather than what they bought.

The implication is still useful, and I have seen it work in practice. Your customers' loyalty is an asset you can use to win new customers. You can offer it to a complementary business whose customers you want, and use theirs in return. The test for a partner is whether their customers trust them as they would trust a friend. If the answer is no, the recommendation is worth no more than an advert. Measured against your own paid social cost per customer, defined properly as I set out in marketing return on investment, a partner recommendation is one of the few ways of winning customers whose cost falls as your retention improves.

What a loyalty programme is for

Most loyalty programmes I review offer a discount and collect a database, and do little else. The ones that work give the customer a reason to stay in contact and give the business a reason to know them. Leading The Dubai Mall and Emirates Skywards programme from its 2018 launch taught me that the currency has to be one people already value, and that the return sits in the data: every receipt-linked transaction showed who was spending, on what and how often, and who could be brought back with the right message. What kept members engaged was recognition: being known, hearing about things first, being asked for their opinion.

The mistake I see most often in the UAE is a programme launched to match a competitor's, paid for out of margin and measured on sign-ups. Sign-ups are the easiest number in marketing to produce. The discipline I apply to any retail brief is the same: know who the customer is, give them a reason to come back that is not a price, and measure the return visit rather than the enrolment. The Bloomingdale's and Dubai Mall case studies show the scale at which it operates. If you have no programme, a well-kept customer list and one event a quarter will often do the same job for a tenth of the cost, and that is where I would start. My longer piece on the future of customer loyalty sets out the model.

Three numbers to check this week

• Second purchase rate. Of the customers who bought for the first time in a given month, what share bought again within ninety days. This shows the loyalty window directly, and it is the number that moves first when retention work is working.
• Time between purchases. The typical gap between purchases for repeat customers, tracked by the month they first bought. When it shortens, the relationship is getting stronger. When it lengthens, you are about to lose them and you still have time to act.
• Share of revenue from customers who first bought more than a year ago. If this is falling while total revenue holds steady, growth is being bought with acquisition spending and the existing customer base is shrinking. A board should ask for this number before it approves another acquisition budget.

Put those three numbers next to the comments from your five-out-of-five customers and a list of the customers who would recommend you, and you have a retention plan on one page. It follows the same rule as the three-number board report I describe in the board-level KPI dashboard: fixed definitions, reported every month, never changed.

Partnerships that work in the UAE

The partner recommendation idea works well in Dubai because the customer base is mobile, well connected and concentrated. The obvious pairs are everywhere. Property and interiors. Fitness and nutrition. Schools and family services. Business setup and banking, accounting or insurance. Clinics and wellness. Car dealers and car care. Each side already holds a customer list the other wants, and a recommendation added to an existing email or WhatsApp broadcast costs almost nothing to test. Agree the terms in writing, track the code or link, and compare the result with your paid social cost per customer after ninety days.

Community works here for the same reason. Expatriate populations form networks quickly and value the businesses that bring them together. A quarterly gathering for existing customers, with no sales pitch, is one of the highest-return activities available to a business owner in this market and one of the least used.

Where to start

Check the three numbers. Read the comments from the fives. Build one thing for existing customers this quarter that does not ask them to buy. Agree one partner recommendation and measure it against paid social. If you have a loyalty programme, measure it on return visits rather than enrolments. If you do not have one, wait until the customer list and the quarterly gathering are working before you launch one.

The limits of the evidence: the luxury data covers two markets, the Gallup research is largely from before 2016, and the recommendation study is one experiment I have read in abstract. All three point in the same direction, and none of what they suggest needs a new platform. This is the kind of growth and performance work that pays back fastest, because the customers are already yours.

Sources. The Business of Fashion and McKinsey, The State of Fashion: Face to Face With Luxury Clients, June 2026 (about 2,000 luxury customers, US and China). Marcus Buckingham, "What Companies Can Learn from Their Biggest Fans", Harvard Business Review, May to June 2026. Daeun Chloe Shin, "Cross-brand referrals as a customer acquisition strategy: how brand loyalty shapes consumer response", Journal of Strategic Marketing, 25 July 2026, DOI 10.1080/0965254X.2026.2708135. The Business of Fashion and McKinsey, The State of Fashion 2026, November 2025. Author's own work on The Dubai Mall and Emirates Skywards programme, Bloomingdale's and Dubai Mall.