A Vending Refund Rate Looks High on a Quiet Day: Show the Counts Behind the Percentage
A vending refund rate can look dramatic on a day with very few purchases. Always show the number of refunds and the number of relevant sales behind the percentage. The rate can help identify a question to investigate, but it cannot explain the cause of the refunds or the reliability of the service by itself.
Begin by defining what is counted. Refund requests, approved refunds and completed refund transactions are not necessarily the same measure. Likewise, a refund processed today may relate to a sale made earlier. Choose a consistent definition and keep the source records available before comparing days, machines or sites.
Show the counts beside the percentage
As a hypothetical example, one refunded purchase among ten relevant completed purchases gives a rate of 10%. One among one hundred gives 1%. The refund count is the same, while the denominator changes the percentage. These examples illustrate arithmetic and do not describe an acceptable industry rate or an actual site’s performance.
Use the formula that matches your definition: relevant refunded purchases divided by relevant completed purchases, multiplied by one hundred. If you instead measure refunded value as a share of revenue, label that separately. Our vending information covers the service context, but the explanation of a particular rate must come from the underlying transaction and support evidence.
Match the periods carefully
Check whether the report groups refunds by processing date or links them back to the original sale. Both can be useful for different purposes, but they should not be mixed without explanation. A quiet day’s payout report may include refunds for earlier purchases, making a same-day comparison misleading if described as that day’s customer experience.
Where possible, keep the original purchase reference and the refund outcome connected. Ask the provider to explain how the report handles partial refunds, multiple items and reversed transactions. Do not assume that each line is a separate affected customer. One case may produce several administrative entries that need reconciliation before counting.
Investigate the events, not only the rate
Read the relevant support records and identify the reported reasons. A product mismatch, a payment question and a goodwill resolution may require different follow-up. Do not label every refund as a mechanical failure unless the evidence establishes that. Similarly, a low percentage does not make an individual unresolved customer issue unimportant.
Check whether a short-term event affected the denominator, such as restricted opening hours or unusually low attendance. Record that context rather than deleting the quiet day because it looks inconvenient. If a longer period is used, explain why it provides a more useful view and still retain visibility of any specific incident requiring action.
Combine periods using their totals
Do not simply average daily percentages when the daily purchase counts differ. In the hypothetical two-day example above, two refunded purchases out of 110 relevant purchases give approximately 1.82% overall. Averaging 10% and 1% would give 5.5%, which does not represent the combined purchase count. Retain the totals so the calculation can be checked.
Present the result with the definition, dates, counts and unresolved limitations. Then identify the practical next step, such as reviewing a selection or clarifying a payment case. This gives the site a fair explanation without dismissing complaints, exaggerating a quiet day’s percentage or claiming that a single ratio proves the machine’s overall performance.
Related guidance: Vending Refunds in a Later Report: Keep the Original Sale Traceable.
For help assessing your site, contact Vending-Machines.ie.
