Feature article - The hidden cost of running the same experiment twice

Josh Miller, VP of account management at Uncountable, shows how a new data architecture can make institutional knowledge fully accessible

Somewhere in your organisation right now, a chemist is working through a formulation problem that a colleague solved eighteen months ago. They are running trials, adjusting parameters and logging results, completely unaware that the answer already exists. It is not in the system, it is in a spreadsheet on a drive in another facility or in a notebook that never made it off someone's desk.

I remember the moment this clicked for a chemist at a large multi-site polymer compounding company. We had spent several months on implementation of our platform, including data migration, structure, and training the team. It was not until the chemist ran his first real search in the live system that it hit him. 

He was looking for prior experiments by performance specification and raw material and suddenly years of work across the organisation appeared in front of his eyes. He went quiet for a moment before saying: “I had no idea any of this existed”. His world had genuinely turned upside down - not because the data was new, but because for the first time, he could actually find it.

This is not an isolated occurrence. In speciality chemicals, advanced materials and other R&D-focused organisations I work with every day, it is a pattern. And it is not caused by a lack of effort or talent but by how most organisations store their R&D data. 

When experiment records live in disconnected spreadsheets, local electronic lab notebooks (ELNs) and unstructured notes, they are not searchable across teams or geographies. The data exists, but it can't be found. Each lab operates as if it's the first to encounter the problem at hand.

Cost of duplication

The real cost of this pattern is bigger than any single wasted trial. Think about what duplicate experimentation consumes: scientist time, raw materials, instrument capacity and weeks on the development timeline. Multiply that across a global organisation running hundreds of projects in parallel and you are not looking at a minor inefficiency. You are looking at a structural drag on innovation speed that compounds quietly across every product cycle.

The scale of the problem becomes clearest when a customer starts tracing how their default process works. At one organisation, the standard response to any new customer request was to go spend time in the lab: a minimum of six weeks, regardless of what had been tried before. When they achieved full visibility into their historical data, they found that 60% of those projects could have been resolved with a search. Six decades of institutional knowledge had been sitting in the organisation the whole time, but it was completely inaccessible.

The fix is not more documentation discipline. It is a data architecture that makes experiments searchable by default. Every trial, result, and failed attempt is captured in a structured, queryable format that any authorised team member can access, regardless of where they sit.

What can change

When this system is in place, the operational changes are significant. The system proactively flags similar experiments when new ones are being set up, so that a redundant experiment never takes place. A chemist in one facility can research the experiments that colleagues across the network have already run for a given raw material or performance target. 

Failed experiments stop being dead ends and start being signals. Historical data becomes a living asset rather than a passive archive. For example, Repsol, a global energy company, consolidated over 10,000 data points from two decades of formulation work into a single searchable platform and cut experimental workload by 30–40% per project as a result.

One of the first things customers do when they gain cross-site visibility is start mapping raw materials across regions. For instance, a team in Europe might be working with one supplier’s plasticiser, while a team in the US has been running trials with a functionally equivalent material for a different product line. Neither team knew the other’s work existed. 

Once that data surfaces in a shared system, the European team does not start from scratch but from a set of results that already narrows the experimental space significantly. The lab time they recover from their very first search clarifies the value of the system they have invested in.

Here is a question worth asking of your own organisation: If your most experienced chemist left tomorrow, would their discoveries be accessible to the team that follows them? If the answer is uncertain, the cost of that uncertainty is already accruing, one repeated experiment at a time.

Contact:

Uncountable

www.uncountable.com/contact-us