How to size a market for a new device
A step-by-step method for hardware and device startups who need a market size they can defend, built from sourced inputs rather than a single large number.
Define the buyer and the use case
To size a market for a new device, start by deciding who buys it and what they use it for. These can be different people. A hospital department may use a device, a purchasing committee may approve it, and an insurer may decide whether it gets paid for. Each of them affects how many units sell.
Write the use case in one sentence, for example "a clinic uses this device to do X for patients with Y." Then name the person who holds the budget. If you cannot name them, the sizing will drift, because you will not know whose count to use.
It helps to run a short list of questions at this stage.
- What problem does the device address, and for whom?
- Who pays, and from what budget?
- Is the purchase a one-time capital item, a recurring subscription, or a consumable?
- What would the buyer do today instead?
Choose the unit of count
The unit of count is the thing you multiply. It might be facilities, clinicians, patients, procedures, households, or installed devices. Pick the unit that matches how the buyer decides.
A rule that keeps things honest: the unit you count should be the same unit the price attaches to. If you sell per device to a clinic, count clinics and devices per clinic. If revenue comes from a per-procedure fee, count procedures. Mixing a patient count with a per-clinic price produces numbers that look neat and mean little.
If revenue has several streams, such as hardware plus consumables or software, size each with its own unit and add them at the end.
Build bottom-up from sourced inputs
Now build the estimate in four layers: how many buyers exist, how many of them fit your segment, how many units each would use, and the net price you would actually receive.
For each input, find a source you can name. Public sources often include government statistics, regulatory databases, industry association counts, published clinical literature, and company filings. Record the source, the date, and the exact definition it uses. When no source exists, write the input down as an assumption and say so.
Use net price, not list price. Discounts, distributor margins, and rebates reduce what reaches you. If you only know list price, state that and treat the result as an upper bound.
Check each layer for double counting. If a hospital and its departments both appear in your buyer count, the same purchase is counted twice. Also check that the buyer count reflects current facilities and not a historical figure that has since changed.
Keep the arithmetic in a spreadsheet where each input has its own cell, so anyone can change one and watch the output move. The guide to TAM, SAM, and SOM for medical devices explains how these layers map onto the three market numbers.
Cross-check top-down
Once the bottom-up number exists, test it against outside figures. Look for published estimates of total spending in the category, installed base counts, or the revenue of established competitors. If a competitor with a large share reports revenue smaller than your serviceable market, ask whether your estimate is too large or the competitor is only one of many.
The goal is a sanity check, not agreement to the dollar. A bottom-up number and a top-down number that differ by a modest amount are normal, because they use different definitions. A large gap means one set of inputs is off, and finding which one is useful work.
Be careful with third-party market reports. Their category definitions often differ from your product, and the methods behind their numbers are not always visible. Note which ones you relied on and what they include.
State assumptions and test sensitivity
List every assumption in plain language next to its value. Then change the ones you are least sure of and see what happens.
Here is a hypothetical illustration, with round numbers chosen only for the arithmetic. Suppose a model assumes 10,000 target buyers, a 30% fit with the segment, 2 units each, and a $1,500 net price. The serviceable market is 10,000 x 30% x 2 x $1,500, which is $9 million. If the fit is 20% instead of 30%, it falls to $6 million. If the price is $1,000, it also falls to $6 million. Seeing that two modest changes cut the number by a third tells you where to focus further research.
Do this for each major input and record the result in a small table. The inputs that move the answer most are the ones to research further, and the ones that barely matter can be left as stated assumptions.
Build a low, base, and high case, and write down what has to be true for each. Investors tend to respond better to a range with reasons than to a single confident figure.
Turning the estimate into a plan
A market size is a starting point. The obtainable share depends on how you will reach buyers, how long adoption takes, and who else is selling. For fundraising, pair the sizing with a competitive view, which our guide to competitor analysis for a pitch deck covers.
If you would rather have this built with sources and documented assumptions, our market sizing for startups page describes the approach, and the packages and add-ons are on the services page. Our work is research and analysis, and it does not constitute investment advice. When you are ready, book a discovery call.
Common questions
How long should a market sizing model take to build?
What if no public data exists for my buyer count?
Should I use list price or net price?
Is a third-party market report enough on its own?
Related
Need a market size you can defend?
Book a discovery call to talk through your device and which sizing scope fits. Pricing starts at $999 for fixed-price research.