A Vending Forecast Counts Every Employee as a Buyer: Build a More Useful Demand Range
The number of employees attached to a site is a starting point for a vending discussion, not a forecast of daily customers. Some people work remotely, bring refreshments or use another service. Others may buy more than once. A useful demand range makes those assumptions visible instead of treating every name on a staff list as one sale.
Begin with the decision the forecast needs to support: equipment selection, stock planning or a commercial proposal. The appropriate level of detail depends on that decision. A rough early estimate can be useful as long as it is labelled provisional and replaced with observations when the service begins.
Estimate attendance separately from headcount
Ask how many people are normally present on different days and shifts. Distinguish regular attendance from visitors and occasional events. Avoid counting the same person twice because they appear in two departmental totals or split their week between buildings. Record the source and date of the attendance estimate.
For a vending business plan, explain the uncertainty rather than hiding it inside one apparently precise figure. An occupied-site count is still not a customer count, but it gives the next assumptions a more relevant base than the organisation’s total workforce.
Separate participation from purchase frequency
Participation means the proportion of people who make a purchase during the period. Frequency describes how many purchases those buyers make. Keep both distinct from the number of items in a transaction. Otherwise a basket containing two products can be accidentally counted as two customers.
As a purely hypothetical illustration, 100 people attending a site with 20 buying once would produce 20 purchasing occasions. If ten of those buyers returned once, there would be 30 occasions but still only 20 buyers. These numbers are not market benchmarks or a prediction for your site; they simply show why the definitions matter.
Build scenarios around observable differences
Create a cautious, central and stronger-demand case using clearly stated assumptions. Tie the differences to attendance, competing refreshment options, access hours and the proposed range. Do not choose a high participation rate solely to make the equipment proposal look attractive. Ask which assumption would most change the decision if it proved wrong.
Where a trial is possible, decide beforehand what will be recorded and for how long. Include availability and unusual events alongside sales. A busy launch day may reflect curiosity, while a quiet week with stock gaps may understate interest. Neither should silently become the permanent daily forecast.
Update the range with real service evidence
After launch, compare actual transactions and items with the forecast definitions. Keep attendance changes visible so a lower-sales holiday week is not interpreted as a sudden loss of interest. If the system does not report unique buyers reliably, do not claim that a transaction total measures participation directly.
Use the findings to adjust stock and service assumptions with the operator. Retain the earlier forecast and explain what changed rather than overwriting it without a trace. The objective is not to prove the original estimate right; it is to make the next decision better informed. A transparent range is more useful than a single confident number built on the assumption that everyone will buy.
Related guidance: Vending Trial Results: Separate First-Day Curiosity from Repeat Demand.
For help assessing your site, contact Vending-Machines.ie.
