How to Size a Market for a New Product

To size a market, build the number twice: top-down from published market data (TAM, SAM, SOM) and bottom-up from buyers x purchase rate x price. If the two estimates land within roughly the same order of magnitude, you have a number you can defend. If they do not, one of your assumptions is wrong - and finding out now costs nothing, while finding out after tooling costs six figures.

Top-down is fast and it frames the opportunity, but it is only as good as the report underneath it. Read the methodology note in any market report before you use its number: many are extrapolations of a single survey, and category definitions vary wildly between publishers.

The bottom-up number is the one investors interrogate, because every input is a claim you can be asked to defend. It is also the number that drives tooling decisions - cavity count, line rate and minimum order quantity all follow from annual units, not from the size of the category.

Purchase intent is the one input you cannot borrow. Run a real concept test with a price attached before you convert intent into units - stated intent without a price is not a forecast.

Size the closest substitute market, then adjust. Count the buyers who currently solve the problem another way, estimate what they spend on that workaround, and treat that spend as the pool you compete for. Validate the adjustment with a priced concept test rather than intuition.

TAM is total demand for the whole category. SAM is the portion your product, price tier and channel can serve. SOM is the share of the SAM you can realistically win in a defined period, given your budget and channel access.

Bottom-up is more defensible because every assumption is visible and testable. Top-down is useful as a boundary check. Serious plans include both and explain the gap between them.

It needs to be right about the order of magnitude and honest about the assumptions. A number accurate to the nearest thousand units is false precision; a number that tells you whether to buy a single-cavity or a four-cavity mold is doing its job.

Abstract sizing advice is easy to nod along to and hard to apply. Here is the same bottom-up method applied to a hypothetical $149 kitchen appliance sold direct and through one national retailer, with every assumption exposed so that any of them can be attacked.

The number that decides this business is not the 48 million; it is the 3 percent. Discovery money is best spent attacking that figure, because a factor-of-two error there changes the entire plan, while a factor-of-two error in the household count changes nothing operationally.

In B2B the discipline is different: the counts are small enough to verify by name. If you cannot list two hundred of the twenty-one hundred qualifying facilities, the estimate is not yet grounded.

A market can be real and still not worth entering. Run the volume forecast straight into a margin model before celebrating the number, because the two together — not the market size alone — are what determine whether the product can fund its own development.

A single number invites disbelief; a range with named drivers invites discussion. Vary the two or three inputs that carry the most uncertainty and present low, base and high cases, stating what would have to be true for each.

The common convention is to count only the top box of a purchase-intent scale, then apply a further discount of roughly three to five times depending on price and category friction. Higher price, more approval steps and longer install effort all push the discount up.

Then bottom-up is the only honest route: count buying units from registries, licences, establishment counts or distributor lists, and validate the counts through interviews. A genuinely absent market report is often a sign the category is emerging, which is worth stating explicitly.

It sets the tooling strategy. A market that supports thousands of units a year justifies hard tooling and a lower unit cost. A market of a few hundred favours bridge tooling, machining or urethane casting with a higher unit cost and far less capital at risk.

Yes — the first two quarters of real sell-through data replace the weakest assumptions in the model. Rebuild the forecast with actual conversion, channel and repeat rates rather than defending the pre-launch number.

Knowing how to size a market is half the work; presenting it so an investor or executive believes it is the other half. The most common credibility failure is a single confident number with no visible assumptions behind it, which invites the audience to test it rather than discuss the decision.

Lead with the range and the single most sensitive assumption. Audiences trust a presenter who names the weakest part of their own model far more than one who defends a point estimate.

Work with LA NPDT: if you are moving from here to execution, start with our product development consulting or talk to us about end-to-end product development.