Follow the science is a phrase repeated endlessly by politicians, institutions and advocates of various causes of late. This to the point where the phrase itself has become more of a motto than anything else.

But what exactly does it mean to follow the science?

This blog outlines in simple terms how science and academia gathers knowledge and how you can use this process as a best practice to make evidence based decisions in your own business.

The evidence pyramid.

Although there are many different formulations of the evidence pyramid, often tailored to the specifics of the field they are addressing, they all have common traits.

At the bottom of the evidence hierarchy you typically find anecdotal evidence in various forms, like expert opinion, single case studies etc.

This data is great for generating hypotheses’, but does not provide strong evidence, and is therefore situated at the bottom of the hierarchy.

Edging towards the top you’ll find randomised test and control trials.

These are experiments in a controlled setting where participants are randomly assigned to either a test group which is exposed to some sort of intervention, or a control group which is not.

The randomised component makes sure that any difference in performance between the groups can only be explained by the intervention.

Finally, at the top you find systematic reviews and meta analysis.

These are trying to find patterns of results across the studies conducted and represent the gold standard of scientific efforts because this is where you typically find actionable insights.

Gathering evidence in your business.

The evidence pyramid gives a rough guide. 

Below follows two simple tips as to how to incorporate these best practices from the academic community in your business, and ensure that your business is ‘following the science’.

First, embed experimentation into the way you govern 
your business. 

Whilst having domain knowledge experts in your business is extremely valuable, unless projections and projects are rigorously tested through experimentation and numbers, they should only be counted as anecdotal evidence.

Experimentation is a great way to ensure your business is data driven, can challenge cognitive biases and help decision makers with valuable insights.

Second, set up an analysis database. 

As mentioned earlier, the holy grail of scientific discovery comes from the patterns observed across different studies.

The only way to perform systematic reviews or meta analysis is by having access to the analysis done previously on the topic.

Therefore, make sure that you document the results and other parameters of any analysis you do in your business and gather them in a single database.

Future posts will outline in more detail how to structure your company’s data to be research ready, how to design a database for historical analysis, and how to perform simple meta-analysis of your experiments.