Find out if the idea works on paper before it costs real money.
How big is the market, what does one unit cost to make and sell, and can the thing physically be built? We answer all three with the working shown, then write down whether to build, wait, or walk away.
Three situations where this pays for itself
You're about to spend on a build
An engineer, a factory, or an agency has given you a quote. Before you sign, you want an independent read on whether the market is big enough and the margins hold. A two-week check is cheap next to a twenty-week build.
An investor asked for numbers you don't have
Someone asked about market size or cost per unit and you gave a guess. You'd like the next answer to come with sources, a spreadsheet, and assumptions you can defend.
It's hardware, and physics might say no
The concept involves a sensor, a battery, a radio, or a moving part. You need someone to check the bill of materials, the power budget, and whether the thing can detect what you say it detects, before tooling is ordered.
Numbers with the working shown, and a decision at the end
Every figure in the report has a source or a stated assumption. If a number is a guess, it's labelled as a guess. Change one input and you can see what happens to the answer.
TAM, SAM, and SOM: the total addressable market, the part you could realistically serve, and the share you could obtain in the first few years. Each layer footnoted to a public source or a stated assumption. Our article on market sizing before you build walks through the method.
What it costs to make, deliver, and support one unit or one customer, set against what you could charge. Includes the costs people forget: shipping, returns, support, and the survey or certification runs that hardware often needs.
For hardware: bill of materials, power budget, range, and whether the sensing approach can actually do the job. For software: which parts are solved problems and which parts nobody has made work yet.
The handful of assumptions the whole plan rests on, ranked by how much damage each would do if wrong and how cheaply you could test it.
A written call with reasons. Build means proceed as scoped. Wait means a specific thing needs to change first, and we say what. No means the numbers don't work, and we show which ones.
Two weeks from first call to a build, wait, or no
A feasibility sprint typically runs two weeks. It's desk research, modelling, and a few expert calls rather than customer interviews, though we'll pair it with interviews if the demand question is still open.
Frame the question
A working session to pin down what we're testing: the product as currently scoped, the price you have in mind, and the number that would make you stop. Everything downstream is measured against that.
Size the market
We build the TAM, SAM, and SOM from public data, industry reports, and your own sales evidence where it exists. Each step from the big number to the small one is written down, so an investor can argue with an assumption instead of the conclusion.
Model the economics
A unit-cost model with the real inputs: materials, labour, delivery, support, and any regulatory or field-testing costs. We test it at three volumes so you can see where the margin appears, if it does.
Check the physics
For hardware we review the bill of materials, power and range, and whether the sensor can measure what the pitch says it measures. Where we need a specialist, we bring one in for a call and note what they said.
Decide and write it down
A readout call and a written recommendation. If the answer is wait, you get the specific change that would flip it to build. If the answer is no, you get the numbers that made it a no.
PavIQ: a $40B market and a $580 bill of materials
PavIQ is a hardware concept. Feasibility work put the addressable market at $40B, footnoted, with a CAD $580 bill of materials and a 20-week build plan. The number that shaped the call was different: the field survey runs needed to validate the product cost between $26K and $105K each. On GridSentinel, a hardware concept at about CAD $95 per node, the same check re-scoped the product from theft detection to outage detection.
Read the PavIQ case study →Straight answers before you book
What do TAM, SAM, and SOM actually mean?
TAM is the total addressable market: everyone who could conceivably buy. SAM is the serviceable part of it: the segment your product and channels can reach. SOM is the obtainable share: what you could win in the first few years against the people already selling there. The first number is for the pitch. The third is the one you plan around.
Isn't market sizing just a made-up number?
It is when the assumptions are hidden. Ours are footnoted, and if you disagree with a step you can change it and the model recalculates. The $40B figure on PavIQ is defensible because the case study shows how it was reached.
Do you do hardware as well as software?
Yes. PavIQ and GridSentinel are both hardware concepts. For hardware, the feasibility work includes bill of materials, power budget, and a physics check on whether the sensing approach works. Where a question needs a specialist engineer, we bring one in for a call.
What if the answer is no?
Then you've spent two weeks instead of twenty. A no comes with the numbers that caused it, so you can see whether a different price, market, or scope would change the answer. GridSentinel didn't die at this stage; it got re-scoped to a problem the hardware could actually solve.
Can you combine this with customer interviews?
Yes, and for a new product we usually recommend it. Feasibility answers whether the business works. Interviews answer whether anyone wants it. A combined research sprint still fits in about two weeks. See product research for the interview side.
Bring the idea, the price you have in mind, and any quotes you've received. In thirty minutes we'll tell you which question is most likely to be the problem and what a two-week feasibility sprint would cover. Pricing is quoted after the call.