IN BRIEF

Technical value, an initial customer and repeatable demand are distinct commercial achievements.

  • Early wins reveal what customers will fund and what delivery costs.
  • Market, value, evidence, delivery and demand must develop together.
  • Software, operational systems and physical products face different adoption requirements.

A customer can pay for a successful project without providing much evidence that anyone else will buy. The result may depend on an unusually committed technical champion, privileged access to data or months of work by the founding team. These are often necessary investments in learning. They also make the first sale a poor guide to the economics of the tenth.

For AI-native chemicals businesses, demonstrating technical value, winning an initial customer and generating repeatable demand are distinct achievements. Progress depends on understanding what made the early engagement possible, which advantages transfer and which concessions cannot become standard practice. A strong result leaves those questions open.

A purchasing decision has different tests

Technical validation establishes performance within defined conditions. Commercial validation asks whether that performance justifies a purchase against competing uses of the customer's money and attention.

A formulation team may welcome better experimental recommendations. Its budget holder still needs to understand whether fewer experiments release usable capacity, accelerate a valuable launch or simply move the bottleneck into testing and qualification.

The alternative may be an established tool, external expertise or continuing with an imperfect process. Switching introduces costs and responsibilities that a benchmark cannot capture. A technical champion can demonstrate usefulness while lacking authority over spending, data access or implementation. An early project becomes more commercially informative when it reveals how those interests were reconciled, and whether another buyer would reach the same decision.

Five conditions that must hold together

Market. A useful segment shares more than an industry classification. Its customers face a similar problem, with comparable urgency, operating constraints and purchasing responsibilities. Two speciality chemicals producers may need very different offers if one must answer customer specifications quickly while the other faces a laboratory capacity shortage. Narrowing the segment improves relevance and comparability, but can expose a market too small to support the intended business.

Value. The economic advantage must survive a realistic comparison with existing practice, including implementation, validation and ongoing support. Faster discovery is valuable when it changes an outcome the customer can capture. It may be less compelling where qualification dominates the development timetable. Pricing must leave sufficient benefit with the buyer while supporting the supplier's delivery costs. A project can create substantial customer value yet offer unattractive supplier margins at the price customers will accept.

Evidence. Proof should resolve the uncertainties that prevent purchase. Scientists may need confidence in predictions; procurement may worry about continuity of supply; operations may need clear accountability when conditions change. Intellegens' FUCHS case reports two tested formulations meeting the project target. That supports the experimental approach, while leaving other buyers to assess its relevance to their formulations and workflows. Evidence becomes more transferable when it explains the conditions behind success, including customer effort.

Delivery. Repetition requires knowing which work is reusable and which remains specific to each customer. By the fifth deployment, data preparation, integration or specialist support should be sufficiently understood to price and resource reliably. Bespoke engineering can remain commercially attractive where contracts justify it. The difficulty arises when a standard product price conceals a bespoke service obligation. Counting customers without examining delivery effort can disguise deteriorating economics.

Demand. Further customers must be identifiable, reachable and able to progress through a recognisable buying process. Founder relationships can secure early access without establishing how a wider commercial team will find qualified opportunities. The next test is whether the same problem, value argument and evidence help another customer commit budget. These conditions develop unevenly: entering a new segment may expand the prospect pool while weakening evidence and increasing implementation work. More apparent demand can therefore make repeatability harder.

The business model changes the burden

Software selling better scientific decisions can earn its place before a finished material reaches market. Intellegens' formulation work illustrates this: adoption concerns include data readiness, scientist confidence and integration into experimental practice. Repeatability depends partly on how readily another team can use the approach.

Operational systems face different obligations. With Imubit's process optimisation systems, recommendations enter plant operations. Implementation must accommodate existing controls, oversight and changing conditions; sustained performance matters beyond initial commissioning.

Physical chemicals, materials and ingredients require qualification, consistent manufacture and dependable supply. Orbital Industries' cooling technology illustrates that burden. Its separately announced CurieOS licensing agreement with BASF ECMS concerns a discovery platform; it cannot establish qualification of the cooling product. Shared AI capabilities can support different businesses, each with its own adoption costs and route to repeatable revenue.

Locate the next constraint

Leaders need to distinguish stalled opportunities from expensive wins. If suitable prospects recognise the problem but cannot justify spending, the value proposition or buying context needs examination. If they accept the economics but repeatedly request further trials, the unresolved risk needs identifying. If customers buy but each deployment absorbs scarce experts, delivery may be the constraint.

Some answers require better models, additional testing or engineering development. Others require sharper customer selection, clearer differentiation or a better understanding of how budgets become available. Reviewing where comparable opportunities stop, and what successful ones cost to deliver, provides a firmer basis for investment than the number of pilots announced.

What carries forward

Early commercialisation teaches a business what customers will fund and what it costs to fulfil the promise. Growth depends on how much of that learning carries into subsequent engagements. The revealing question is how much must be rediscovered for each sale. As technical capabilities mature, understanding the market, articulating a defensible advantage and building evidence that travels between customers become continuing management responsibilities. Those capabilities need time to develop alongside the technology they help commercialise.

Sources and further reading

Company-reported results and agreements are treated as bounded evidence, without assuming scalable demand.

KEEP READING
Thought Leadership in Technical B2B Is About Risk Reduction, Not Content MarketingB2B Marketing Execution Is Really the Management of Strategic Ambiguity

Working through a similar challenge? Start a conversation with Richard ↗

← All articles