What can scientific labs and an innovation company learn from each other?
Before we start, I would like to introduce myself. I’m a chemistry student used to immersing myself in data, creating graphs and drawing curly arrows.
I joined the Fluxx team as an intern a month ago and found that they too, “do experiments”. Albeit in a very different way, there are still some similarities to the chemistry experiments familiar to myself.
Let’s start by defining an experiment. Simply put, it’s ‘a procedure undertaken to make a discovery, test a hypothesis, or demonstrate a known fact.’
In a scientific setting, this may be used to discover a property of a compound or prove a theory. In an innovation company, it’s used to decide whether an idea is worth investing in. In both cases, there is a process by which one must follow to prove or disprove the hypothesis.
“No amount of experimentation can prove me right; a single experiment can prove me wrong” -Albert Einstein.
You will always know that an idea may work to a certain extent. But you need to know where major flaws lie, and this is only achieved through experimentation.
And when you look at their basic structure, experiments in science aren’t actually very different from those at Fluxx.
Comparisons:
Both methods of experimentation begin with a hypothesis. When working with a client this may be an assumption (e.g. about the customers). In a scientific experiment, this may be assumptions that physical laws are always obeyed and there are no exceptions.
Both also require a method of gathering data. In science, you might find something like an infrared spectrometer. In business, you might find fake doors and wizard-of-oz tests. Following that, you’d analyse the data and relate it back to the hypothesis. Does the data agree/disagree?
There are still a few key differences. Notably more BANG! POP! FIZZ! And bright colours with scientific experiments. (However, it can’t be ruled out for an innovation company). Scientific experiments also require a set-up of instruments, vessels and compounds which need timing and adjusting by hand. Science definitely involves a certain empirical exactness. Repeats are performed with maybe only a few data points gathered however the same thing is done each time.
Innovation experiments normally involve interactions with people (like the aforementioned fake doors and wizard-of-oz tests). This therefore requires a lot more data due to the probabilistic nature of this experiment type, showing whether the hypothesis is more or less likely to identify trends and quantify benefits.
Tests in chemistry should normally demonstrate the hypothesis, as most things follow laws of science (although as ever, there are exceptions). However, when you are gathering data from a single person you can be unsure whether the hypothesis is correct, or you just got lucky.
Likewise, within the methods there are very different materials required: chemical equipment, compounds, facilities vs notepad and pen, voice recorder, video camera. However, specialist roles are still given depending on the task.
Data Analytics
Data analysis is a key component of any experiment. This is how one can decide the success of an experiment. Both the way data is gathered, and the process of analysing it can be quite different from each type of experiment and even from any individual experiment depending on the method.
Within chemistry the data may be gathered using spectroscopic methods to analyse a compound you have produced and comparing this to a literature proven graph. This is a process method whereby you are making certain checks along the way to make sure you will end up with the correct compound.
On the other hand, it could be a more physical experiment, using a computer programme to harness a number of results where a graph may be plotted. This then may be used to find a trend line and a gradient value to plug into an equation and from there refer back to the hypothesis.
At Fluxx, we base our hypotheses on customer research. We begin our projects by getting close to customers, observing and conversing with them in their everyday environments in order to get an understanding of their needs, behaviours, and motivations. We then form hypotheses about ways in which we could add value to their lives.
In some cases, we could use the NPS (net promoter score) to measure success. Your hypothesis may be that Agent X will have an effect on the NPS and it will change by Y points. You would then test to see how much this is true i.e. whether it is more, less or the same as Y points.
Both the physical chemistry experiments and the NPS experiments will provide you with a number of data points. This can then be measured against a control group, thus proving that you achieved the results from the thing you changed.
Other aspects and conclusions
Fluxx’s motto is ‘start by starting’. This is a very good approach when it comes to business, as it drives the task forwards. Within a lab with dangerous chemicals however, there are things we must do before starting to mitigate the risk of death and serious injuries. COSHH safety forms being one of them, however much of a pain they are.
However, the purpose of the experiments within Fluxx is to avoid any major long-term risks within the business. With innovation you are using the innovation process itself to drive the hypothesis and quickly prove or disprove the hypothesis. Going in all guns blazing with an idea, without first testing the hypothesis, can cost a lot of time and money. This can show that people agree there is a need for a given product, but it can also show a disagreement with the same initial assumption.
If there’s something that innovation companies could learn from scientific experiments, it’s that if you follow such a rigorous method, including safety checks, there really are the most minimal risks to the business, and clients you work with should understand this. If you follow proper procedures, there will be little to no consequence.
If there’s something that scientists can learn from Fluxx, it’s that you may need to adapt your plan at will. Innovation in the technique you are using, new instrumentation, different methods to reach the end point or a new catalyst may be the key to improving time to market or quality of the outcome.
