A revenue projection model turns assumptions about customers, volume, prices or capacity into a sales forecast. Start with the business drivers you can explain and check, rather than choosing a growth percentage because it produces the revenue target you want.
This guide compares four methods using retail examples. The numerical cases are fictional. They illustrate the mechanics rather than provide a forecast for Ulta Beauty or another actual retailer.
Choose a method that fits the evidence
| Method | Main calculation | Important limitation |
|---|---|---|
| Revenue growth | Prior revenue × (1 + growth) | Can hide changes in capacity, price and volume |
| Store productivity | Time-weighted stores × sales per store | New stores may ramp differently |
| Selling-area productivity | Time-weighted selling area × sales per area | Store size and mix can change |
| Existing and new cohorts | Existing-store sales + new-store sales | Requires explicit timing and cohort definitions |
These are alternative ways to explain the same business. Do not add the results of all four together. Different assumptions can produce different forecasts; reconcile those assumptions rather than forcing every method to match.
Project revenue from a growth rate
If prior-year revenue is 100 million and the forecast growth assumption is 8%, projected revenue is 108 million. In Excel, with prior revenue in B2 and growth in B3, use =B2*(1+B3).
The formula is simple. The work is explaining the 8% assumption. Separate price, customer demand, new capacity, acquisitions and changes in reporting scope where they matter. A historical average is a starting reference, not proof that the same growth will continue.
Use time-weighted stores
Assume 100 stores operate for the whole year and ten new stores each operate for half the year. The equivalent full-year store count is 100 + (10 × 0.5) = 105. If every equivalent store generates 1 million annually, the projection is 105 million.
The ending count is 110, but using 110 × 1 million would treat the new stores as open for the entire year. Averaging opening and closing counts can approximate an even opening pattern; use actual planned opening dates when available.
Model closures, relocations and partial trading periods as well. A new store may need a separate productivity ramp rather than immediately matching a mature store.
Project sales from selling area
Assume time-weighted selling area of 105,000 square feet and annual sales of 1,000 per square foot. The projection is 105 million.
Match the denominator to the revenue measure. Keep selling area distinct from total premises area if storage and other space are excluded. Check whether e-commerce revenue belongs in the numerator before using a physical-area measure.
A larger store is not automatically equally productive per square foot. Investigate location, format, product mix and the time taken to reach a normal trading level.
Separate existing stores from new stores
This version uses the same opening 100 stores, each with prior-year revenue of 1 million. Assume all remain in the comparison cohort, their revenue grows 4%, and there are no closures, acquisitions or online sales.
The ten new stores open halfway through the year. Assume their annualized productivity while open is 80% of a mature store’s forecast sales of 1.04 million.
| Component | Calculation | Forecast revenue |
|---|---|---|
| Existing stores | 100 × 1 million × 1.04 | 104 million |
| New stores | 10 × 0.5 year × 80% × 1.04 million | 4.16 million |
| Total | 104 + 4.16 | 108.16 million |
Total revenue growth is 8.16% against the prior 100 million. Of that increase, 4 million comes from the existing cohort and 4.16 million from new stores. The existing cohort’s revenue is 104 million, not merely its 4 million increase.
The simple 105-million store-count case assumed different productivity. The cohort case explicitly includes existing-store growth and a new-store ramp. That explains why the two outputs differ.
Check the company’s comparable-sales definition
“Comparable sales” or “same-store sales” is a defined comparison measure, not a universal rule that every store enters after exactly one year. Read the company’s definition, including treatment of online sales, closures, relocations and calendar differences.
For example, Ulta Beauty’s annual report for the year ended 31 January 2026 describes entry into its comparable-sales measure from the first day of the 14th month of operation for the relevant stores and e-commerce platforms. See the company’s filed annual report. This definition is a source example, not an assumption that the fictional case follows Ulta’s reporting.
Link the forecast to cash
Revenue is not automatically the cash collected in the same period. Collection terms, returns, discounts and recognition timing can create differences. Use the revenue forecast to drive the relevant receivables and cash schedule rather than treating a sales total as a bank balance.
The three-statement Excel example shows those links. The startup cash forecast focuses on payment timing and funding needs.
Test the forecast
In the cohort case, reduce existing-store growth to 2%, keep new-store productivity at 80% of mature forecast sales, and leave timing unchanged. Existing revenue becomes 102 million, new-store revenue becomes 4.08 million, and total revenue becomes 106.08 million.
Then test a delayed opening separately. Distinguish lower demand at existing stores from fewer months of trading at new stores. They may produce similar revenue totals but require different explanations.
Before using the forecast, reconcile store counts, opening dates, sales scope, units and the prior-year base. Record which assumptions are supported by evidence and which still need testing.
