Nobody could imagine that millions of people in the UK would be discussing algorithms. But the A-level fiasco has changed all that. In all walks of life we rely on algorithms. Health, justice, marketing, entertainment etc all heavily rely on algorithms. I wonder whether the A-level shambles may have changed the perception of data models.

As a data-driven marketer I have worked with algorithms for many years. I’m not a data analyst, I’m a data planner and so it is my job to brief, work with, oversee and present the work of the data analyst.

I have always taken the view that a data planner is a vital component in any data analysis work. There are always vital questions to be asked and decisions to be made in the key stages of building any data model and the data planner should be a key part of those questions and decisions as they will bring a customer and business perspective to the party. As they are not as technically blessed as the data analyst they can ask the stupid questions that often turn out to be far from stupid. They are not as concerned with the process and methodology but are focused on the outcome.

I’m not sure if there were data planners involved in the building of the A-level algorithm but if I was involved I would have wanted to look at the deviation between the predicted grades and modelled grades at a student level. I would have wanted to look beyond the averages and see what it meant for the students. If the model is outputting grades three or four levels below the predicted then something is clearly wrong.

I’m a champion of data modelling but I’ve always believed they are only as good as the people who worked on them. As we move ever faster towards AI and machine learning the need for good old fashioned data planners to interpret, challenge and ask stupid questions has never been greater.

Further Reading

Data is the lifeblood for CRM Marketing, so we have written a lot of articles about what it is and how to use it.  In this related article we explore why data storytelling can help you turn complexity into business outcomes.