Deep tech investors often face a difficult imbalance: the founder understands the science better than the person deciding whether to fund it. The answer is to turn the investment into a set of questions that qualified people can test, with clear consequences for the capital plan.
A theme can tell you where to look. It cannot tell you whether one company's result is reproducible, affordable at scale or valuable to a customer.
Choose a theme you can investigate
Start with a bottleneck in a specific industry: an expensive production step, a material limitation or a capability customers need but cannot obtain economically. Identify the technical approaches that could change it and the evidence needed to compare them.
Then assess your ability to investigate. Can you recruit independent specialists, speak with potential buyers and understand the funding required between demonstrations? A theme is more useful when you can name those people and milestones. If every important judgement depends on accepting one founder's account, broaden the evidence base before narrowing the investment focus.
Separate the claims in the pitch
Write the central investment claims as statements that could be wrong. “Our process is superior” is too broad. “The process achieves the required purity at the stated throughput and energy cost” creates a basis for investigation.
Then separate the evidence:
| Claim | Evidence to investigate | Who can help assess it |
|---|---|---|
| The underlying effect exists | Methods, measurements and reproducibility | Relevant independent technical specialist |
| The product works in its intended setting | Tests under realistic conditions | Engineering or application expert |
| Production can meet the cost target | Yield, inputs, equipment and scaling assumptions | Manufacturing or process specialist |
| Customers will adopt it | Qualification requirements, trials and purchasing decisions | Prospective users and buyers |
| The company controls what it needs | IP ownership, licences and dependencies | Appropriate legal and technical advisers |
One positive answer does not resolve the others. A scientific result may be sound while the proposed manufacturing process remains uneconomic.
Use readiness levels carefully
NASA's technology-readiness framework distinguishes stages such as proof of concept, validation in relevant environments and an operationally demonstrated system. It is useful vocabulary for describing technical maturity.
It does not establish customer demand, production margin or an attractive investment price. Ask which version of the framework the company uses, what evidence supports its stated level and what remains before deployment in the actual customer's setting.
For software embedded in hardware, examine the complete system. A component's laboratory performance may not survive integration, environmental conditions or production variation.
Commission a question, not a general endorsement
An expert review should have a written scope. Identify the claim being tested, the material available, conflicts of interest and the limits of the reviewer's expertise. Compensate the work transparently and distinguish technical findings from a recommendation to invest.
Ask the reviewer to state the strongest alternative explanation for the reported result and which additional test would distinguish it. Have them identify missing data and assumptions they could not verify.
Where a claim is decisive, a second independent view or a reproducibility test may be more valuable than a longer report from one person. The objective is to reduce a specific uncertainty, not collect respected names for the investment memo.
Connect each milestone to a capital decision
Suppose an illustrative company needs £2m to validate a pilot process. Its next financing would fund £8m of production equipment. Before investing in the pilot, identify what its results must show to justify that much larger commitment.
The milestone might require a minimum yield, an independently witnessed performance test and evidence that a buyer will qualify the output. Set the thresholds before results arrive, alongside a reasonable response if they are narrowly missed.
Distinguish a technical failure from a timing delay. Additional money can buy more experiments, but it cannot make an invalid physical assumption true. A delay caused by a replaceable supplier poses a different decision from a result that contradicts the core mechanism.
Put scaling costs in the commercial model
Prototype unit cost is not the same as production cost. Examine equipment utilisation, scrap, energy, maintenance, quality assurance and the staff needed to keep the process running.
Customer qualification can add time even after the product works. Ask what the buyer must approve, what integration costs it faces and whether a paid trial leads to a defined procurement decision. A letter expressing interest may demonstrate relevance without creating an obligation to purchase.
Use a cash schedule that includes unsuccessful tests and a slower qualification process. Then establish who could finance the company through the next stage and what evidence those investors would need.
Make the remaining uncertainty explicit
The investment memo should name the technical claims verified, the commercial assumptions still open and the tests funded by the proposed cheque. Record who assessed each important point and the material available to them.
You may never understand every technical detail as well as the founder. You can still understand which claim the valuation depends on, how it will be tested and what happens to the investment if the evidence changes.
FAQ
Do you need a technical background to invest in deep tech?
You need access to expertise appropriate to the claims you are underwriting, plus enough understanding to frame the questions. Independent specialists can assess the science and engineering. Their findings should state what was tested, what remains uncertain and how those uncertainties affect the commercial plan.
Does a high technology-readiness level make a company investable?
It addresses technical maturity under the framework being used. It does not establish customer demand, production economics or an attractive valuation. Ask for the evidence behind the claimed level, then assess manufacturing, qualification and sales separately.
How should investors budget for deep tech follow-ons?
Map the capital needed between meaningful milestones, including failed tests and slower customer qualification. Distinguish money already committed from funding that must be raised later. Your initial cheque should have a clear purpose even if you cannot finance every subsequent stage yourself.
What should a deep tech syndicate establish before investing?
Agree how technical diligence will be commissioned, how confidential findings can be shared and who decides on follow-ons and exits. Budget for ongoing administration and a potentially extended hold. The structure should support the company's development path, rather than assume the first demonstration creates liquidity.
