Technology research, compared.
Also called: Technology scouting · Solution survey · Tech option appraisal · Approach selection
Working out which technologies could deliver the product, then comparing them on cost, maturity, risk and fit before you commit to one.
List the candidate ways to build it. Score each on cost, maturity, risk and fit. Pick the boring, proven option that meets the spec, not the clever one. There is a worked options table below, and a checklist you can run.
What technology research is
Every product can be built more than one way. Technology research is the short, deliberate phase where you find the candidate approaches, lay them side by side, and judge each on the same four things: cost, maturity, risk and fit. Then you pick. It sits before you commit, not after, because changing the core technology once design starts is expensive.
Maturity matters more than people expect. A proven, dull approach that thousands of products already use carries known costs, known suppliers and known failure modes. A novel one might be cheaper on paper and a nightmare in practice. The job here is not to find the cleverest technology. It is to find the one that meets the spec with the least risk you can get away with.
The four lenses
- Cost. Bill-of-materials cost at your volume, plus tooling and any licensing. A penny on a part you ship in thousands is real money.
- Maturity. How proven is it? Off-the-shelf and widely used, or bleeding-edge and barely documented? Maturity buys you suppliers, data sheets and people who have already made the mistakes.
- Risk. What can go wrong, and how badly? Safety, supply, obsolescence, the skills you would need to hire. Rank the risks; do not just list them.
- Fit. Does it actually serve the product, or are you bending the product to suit a technology you find interesting? Fit is the lens that catches engineer-led mistakes.
Here is what the comparison looked like for the proofing box, where the core question was how to heat the box and hold 26°C ±0.5°C overnight on under 30W, so you can see the shape of a real appraisal rather than a generic template.
Notice the pick is the second-cheapest option, not the cheapest. The thermostat lost on fit; the PID won because it met the ±0.5°C spec while staying mature, sub-30W and repairable. Cost alone never decides it.
The two approaches below are the ones I see go wrong most often. The red column is the trap; the green is the better move.
- Picking the newest, cleverest technology because it is interesting.
- Scoring on cost alone and ignoring maturity and risk.
- One unproven supplier, no data sheet, no second source.
- Bending the product to suit the technology.
- Picking the dullest approach that still meets the spec.
- Scoring every option on cost, maturity, risk and fit together.
- Choosing parts with data sheets, suppliers and a second source.
- Letting the spec decide the technology, not the other way round.
How it fits the bigger picture
Technology research is activity 03.04 in the framework, inside Stage 03 Innovate. It builds on the ideas and concepts generated earlier in Innovate, and it feeds straight into assess sustainability (03.05), which weighs the chosen approaches against material, energy and end-of-life cost before Stage 04 Evaluate tests the commercials.
What it can do
It narrows a wide field of possible technologies down to a defensible shortlist, scored on the same four lenses, so the team commits with eyes open. It catches the expensive mistake (the wrong core technology) while changing it is still cheap.
What it can’t do
It can’t prove the chosen approach works in your product; that is what prototyping and testing in later stages do. The scores here are informed judgements, not test results. A promising option can still fail on the bench, and you should expect one or two to.
See the full 10-stage process →
Try it yourself
Take the hardest technical promise in your spec. List every realistic way to deliver it, three or four approaches, no more. Score each one on cost, maturity, risk and fit, in plain words, on a single page. Then pick, and write down why the others lost. If you cannot say why an option lost, you have not researched it yet.
Want a guided first pass? Start the Free Sprint → and the GPT will help you frame the technical question worth researching.
Your tech-research checklist
Project notes: choosing how to hold the heat
▸ From the notebook · optional reading
An afternoon with a Manchester PCB partner, three ways to heat the box on the table, and why we picked the second-cheapest.
3 min read · click to open
The promise was fixed: hold 26°C, give or take half a degree, all night, on under 30W, with no app and a price of £149. The open question was how to heat it. I sat down with a small PCB shop in Manchester who had built low-voltage controllers before, and we put three approaches on the table.
What we weighed
Option one, the cheap thermostat. A resistive element and a bimetallic thermostat, well under £10 of parts, repairable with a screwdriver, the kind of thing that has heated greenhouses for decades. I liked it. Then the Manchester engineer pointed out a simple thermostat swings around ±2°C. That breaks the ±0.5°C promise on the box. Out on fit.
Option two, the PID controller. The same element, but driven by a small microcontroller running a textbook PID loop, a cheap thermistor reading the air next to the dough rather than the element. A few pounds more in parts. Mature, well documented, holds the temperature comfortably under 30W. The only new risk was firmware, and PID control is about as well-trodden as software gets.
Option three, a clever off-the-shelf module. A newer all-in-one heater-control board. Fewer parts to assemble, but one supplier, a thin data sheet, and no second source. I asked what happened if that supplier discontinued it mid-production, and nobody had a good answer. Too much risk for a product we wanted to make for years.
The decision
We picked option two. Not the cheapest, but the one that met the spec with the least risk we could get away with. We worked through the BOM that afternoon and landed it between £38 and £55 all in, ceramic shell included, which kept the £149 price honest. The choice also kept us inside reach of UKCA and BS EN 61010, because every part had a data sheet behind it.
The boring option won, as it usually does. I have rarely regretted choosing the proven approach; I have regretted the clever one more than once.
— Innovate stage, project notes, 2026
— Next in Innovate → Assess sustainability