A busy promotion can look successful while leaving a shop unsure what it actually gained. Discounted cups went out, the queue grew, and a campaign code appeared on many receipts. None of those observations answers the important question: did the offer bring customers who wanted to come back for the drink at its usual price?
For a small bubble tea shop, a useful promotion review separates the first purchase from the later visit. The aim is not to invent a universal conversion target. It is to decide what evidence would justify running your particular offer again.
Choose the outcome before choosing the discount
Start with one commercial question. Are you introducing a fruit tea to existing milk tea customers, attracting first-time visitors during a quiet afternoon, or bringing former regulars back? These are different experiments. A campaign that moves orders from Saturday afternoon to Saturday evening may not add much business, even if redemptions are high.
Write down the eligible drink, dates, channel, normal price and offer price. Avoid changing the recipe halfway through the test. For a fruit-tea offer, select from your existing fruit jam range and validate one recipe first, rather than buying several unfamiliar ingredients just for the promotion.
Keep a simple campaign record
Use a unique campaign code in the till and a short daily record of eligible orders, redeemed offers, refunds, remakes and additional preparation time. Separate genuinely new customers from returning ones only where your existing system can do so reliably. Anonymous cash sales may not support that distinction; record the limitation rather than estimating a precise customer count from receipts.
If you use an existing opt-in loyalty program, review repeat visits through that program. Do not make unnecessary personal information a condition of buying a drink. Aggregate reporting is often enough for a small operational trial.
Define a full-price return consistently
Choose a review window before the offer starts, such as the following four weeks. This is an example, not a recommended industry benchmark. Define a return as an eligible customer buying again during that window without the campaign discount. State whether another unrelated loyalty reward excludes the transaction from your full-price measure.
For illustration, suppose 80 identifiable customers redeem an offer and 12 return at full price within the chosen window. The observed return rate is 12 divided by 80, or 15%. That is not proof that the offer caused all 12 visits. Some customers might have returned anyway, and unidentifiable customers are absent from the calculation. Keep those caveats beside the result.
Count costs beyond the discount
Compare the amount collected with the ingredient, packaging, payment and other directly attributable costs of the offer. Then consider additional staffing, advertising and waste separately. A free topping still consumes stock and handling time. An offer that causes a long queue may also affect customers who were willing to pay the regular price.
The U.S. Small Business Administration’s marketing guidance recommends comparing marketing costs with the revenue generated. In this shop-level trial, go a step further operationally: record the costs and service changes that the till total alone cannot explain.
Compare like with like before repeating
Compare similar weekdays and service periods, noting holidays, weather, nearby events and stock availability. These are possible explanations for a sales change, not reasons to discard inconvenient results. If practical, test one offer at a time so a second campaign does not obscure the first.
At the review date, make a short decision: repeat unchanged, revise one feature, or stop. Include redemptions, observed full-price returns, the contribution calculation and one service lesson. A modest offer that brings repeat demand without disrupting drink quality can be more useful than a large redemption total that never leads to another visit.
Featured image: AI-generated illustration, not a photograph of Runming products or facilities.
