Forecast a new product using comparisons you can explain

A new product has no sales history of its own, but it may share useful characteristics with earlier launches. Build the forecast from those similarities and differences explicitly. Selecting one successful launch because its revenue matches the desired business case creates a forecast with a persuasive story and a weak evidence base.

Define the outcome first: units sold, active customers, recognized revenue or collected cash over a stated launch period. Then identify comparable products with similar purchase cycles, distribution, price points and customer problems. Preserve unsuccessful and slow launches where they belong in the comparison population.

Forecasting: Principles and Practice discusses forecasting by analogy and other judgmental methods for new products. The process below is a practical way to document such a comparison. It does not turn subjective judgments into statistically established probabilities merely because they appear in a spreadsheet.

A product sold to an existing customer base has a different starting position from one that needs an entirely new audience. An enterprise product with procurement reviews has a different sales cycle from a self-service purchase. Choose comparisons that resemble the mechanism behind demand, not only the industry label or visual appearance of the product.

Record the launch channel, reachable audience, purchase frequency, implementation burden, capacity limits and commercial terms for each candidate. Mark facts that are unavailable. A competitor's public revenue number may cover a different product mix or period and may be unsuitable as a direct input.

Where internal launch data exists, use the original dated records. Later lifetime revenue should not be compared with the first three months of the new product. Align the observation window and distinguish launch-period sales from renewals, upgrades and revenue from products added later.

Consider an invented comparison. An earlier launch reached 10,000 eligible customers, converted 4% during its first quarter and sold at a net price of $250. It produced 400 sales and $100,000 of revenue under the simplified assumption that each sale belongs entirely in that quarter.

The proposed launch can reach 6,000 eligible customers at a $300 net price. If the 4% conversion assumption were retained, the forecast would be 240 sales and $72,000. A 2% conversion scenario gives 120 sales and $36,000; a 6% scenario gives 360 sales and $108,000. These are conditional scenarios, not probability bounds.

The smaller audience and higher price are separate changes. The old conversion rate may not survive the price difference or the new product's purchase friction. Show the evidence supporting each adjustment and identify which assumption has the largest effect on the funding decision.

Document deliberate exclusions as well as inclusions. A previous launch backed by an unusually large reseller commitment may be unsuitable for estimating ordinary direct demand, even when its product was similar. Keep it available as a separate distribution scenario rather than quietly deleting evidence that does not fit the central case. This makes it easier to challenge the comparison set without rebuilding the entire forecast.

Several analogues can reveal how sensitive the result is to distribution, pricing or customer familiarity. Keep the reasons for including each one. A simple average of incomparable launches does not necessarily provide a better forecast than a carefully bounded single comparison.

If the team assigns weights, explain them. Similar audience and sales process may justify more influence than a superficially similar product with a different route to market. Separate the judgment about similarity from the eventual sales result so the weighting rule is not chosen only after seeing which forecast management prefers.

Include a reference-class view of failed launches when the data supports it. Looking only at products still actively marketed can create survivorship bias. If unsuccessful launches are missing from the available records, state the limitation and widen the scenario discussion rather than inventing their outcomes.

A demand estimate is not yet a cash forecast. Model the time between customer interest, order, delivery, invoicing and collection. A product that requires implementation can have a capacity limit even when market demand is strong. Deposits, refunds and cancellation rights can also change the cash path.

In the illustrative 240-sale case, suppose only 150 customers can be onboarded in the first quarter. The remaining demand must be assigned to a later period, a waiting list or lost sales according to explicit assumptions. Counting all 240 as completed sales while budgeting capacity for 150 creates an inconsistent launch plan.

Ask accounting to determine the revenue-recognition treatment under the actual contract and applicable framework. Keep that decision separate from the demand forecast. A customer payment at launch may fund delivery without all of it becoming revenue immediately.

Separate changes in the accessible audience from changes in conversion. If only 3,000 of the planned 6,000 customers were reached, lower sales do not yet establish that the offer converted poorly. Report delivered outreach, qualified exposure and completed purchases so the team can tell a distribution shortfall from weak customer interest.

Set review dates around observable events such as the first customer invitations, completed demonstrations, accepted orders and delivered units. Define the measurement consistently. A free signup, a refundable reservation and a paid completed purchase provide different evidence about demand.

Compare actual conversion with the relevant eligible population, including people who did not respond. Early adopters may be more interested than the later audience, so avoid applying their conversion rate to everyone without considering selection. Track whether outreach and service capacity match the conditions assumed in the comparison.

Retain the original forecast and explain updates as new evidence, changed assumptions or execution differences. The resulting history helps the next launch team understand which analogues were useful. Finish each review with the commitment that the evidence now supports, such as another production lot or a later hiring date, rather than treating the first forecast as a permanent promise.