Why Merchant Repeat Rate Was the Key Metric
Daily deal platforms needed a continuous supply of merchant offers to operate. Without new deals, consumers had nothing to buy. Without consumer engagement, platforms had no reason to exist. The supply of deals came from merchants — and the only question that mattered was whether merchants who ran one deal would run another.
If repeat rates were high — say, 70% or more — the industry had a sustainable supply chain. Merchants were happy, economics were working, and the model could scale. If repeat rates were low, platforms were in a constant race to recruit new merchants to replace the ones that would not come back. Recruiting new merchants is expensive. It requires sales teams, education, and negotiation. At some threshold of low repeat rates, the acquisition cost of new merchants would exceed the revenue per deal.
That threshold was reached. The repeat rate data told the story before the bankruptcies did.
What the Research Found
Research into the daily deal model from 2011 to 2013 — when the category was at its peak and declining simultaneously — produced remarkably consistent findings. The numbers varied by platform, city, and industry, but the pattern was the same.
Restaurants had the lowest repeat rates. Food and beverage businesses ran the most deals but returned at the lowest rates. The margin compression was most severe in this category: a restaurant with a 15–20% food cost margin who offered a 50% discount and shared 50% with the platform often found themselves below break-even on the deal day. Repeat rates for restaurants were often below 30%.
Service businesses fared better, but not well enough. Spas, fitness studios, and entertainment venues had higher margins and sometimes better conversion of deal customers to regulars. Repeat rates in these categories were higher — sometimes 40–50% — but still well below what a sustainable supply model required.
The conversion narrative was not holding. The deal platforms' pitch to merchants was always about customer acquisition: run a deal, get new customers, convert them to regulars. Post-deal surveys of merchants consistently found that most deal customers did not return at full price. The conversion rate that would have justified the discounting — typically cited by platforms as 20–30% — was rarely achieved.
The Supply Problem This Created
If a platform had 1,000 active merchants and a 35% repeat rate, it needed to acquire approximately 650 new merchants every cycle to maintain its deal supply. At a sales cost of $200–400 per merchant acquired (including time spent by the sales team), the recruitment cost alone was $130,000–260,000 per cycle — before any operational costs.
This was not a sustainable model. As the pool of willing first-time merchants shrank in any given city, acquisition costs rose. Platforms expanded to new cities to find fresh merchant supply, which added operational overhead without fixing the underlying repeat problem.
Marketplaces are only as durable as their supply side's willingness to return. Platforms optimised for consumer experience but neglected the merchant experience — because the merchant was seen as the means to an end, not a party whose long-term interests mattered. Every sustainable marketplace that emerged after the daily deal era (Airbnb, Etsy at scale, modern gig platforms) invested heavily in making the supply side feel the economics were fair. Daily deal platforms never solved that problem.
What a Healthy Repeat Rate Would Have Required
For the daily deal model to work at sustainable acquisition costs, merchant repeat rates needed to be above 60–65%. Achieving that would have required one of three things:
Better deal economics for merchants. Lower platform commissions, smaller required discounts, or caps on voucher volume would have made the math more survivable. But reducing commissions reduced platform revenue, and merchants offering smaller discounts attracted fewer consumers. Every lever that helped merchants hurt the platform's financials.
Genuine customer conversion. If deal customers had converted to regular customers at meaningful rates, the merchant's total lifetime value from running the deal would have improved the math. This was the one outcome everyone wanted but no one could manufacture. Consumer behaviour was the variable no one could control.
Better merchant selection and vetting. Platforms that focused on business categories with better margin structures — services over restaurants, experiences over commodities — saw better merchant satisfaction. But the most popular deals were always food, and food was the category where the economics were worst.
What the Number Predicted
A 35% merchant repeat rate, combined with rising merchant acquisition costs and declining consumer novelty, created a trajectory that was visible in the data well before the bankruptcies arrived. The platforms that survived understood this and pivoted before the crisis forced their hand. Those that didn't — most of them — continued recruiting new merchants until they ran out of budget and goodwill simultaneously.
The data did not lie. The platforms chose not to read it carefully enough, or chose to believe their conversion numbers would improve. They did not.