The Value of Predictions

By James Lea

June 14, 2024

This was a Podcast I gave to Jonathan Norman, available here: https://www.podbean.com/ew/pb-tttuy-16404d8

Here’s what I said!

Podcast

Hi, I’m James Lea.

Thank you Jonathan for the chance to be on the podcast!

I’ve been asked to talk about the Value of Prediction. This is a subject close to my heart, as it’s been central to my work over the last 16 years. I will tell you why in a minute.

Expertise

My background is physics, engineering and data science. In my early career I was a software engineer, and I even founded an AI startup long before it became popular.

Since then I’ve been involved in project and programme delivery for over 20 years - aerospace, transport, consumer goods, defence and energy – and I’ve used all the methodologies – waterfall, iterative, hybrid, agile, R&D.

Along the way I’ve become a fellow of the APM, and of the BCS.

Predictions are a common theme in all of my work, and central to my message in this podcast. Without predictions, we cannot deliver. Their value is immense.

Prediction Story

Let me tell you a story why – dating back to 2007.

Picture the scene – I’m managing a large team delivering a complex software project – an air traffic control system. We’re running late, and the programme manager wants to know – when will we deliver? There’s a large team downstream, waiting for the product to test, and the pressures is immense.

What happens?

We fail. We miss the deadline. Finances and reputation take a hit.

But out of that failure, was borne an enormous piece of learning. We discover that it is possible to systematically predict the future. The client supports us, and together we learn.

Learning

We learnt that the past is a guide to the future. It allows us to make predictions.

We collected data on every aspect of our work. We measured the sizes of everything – the effort, the amount of code we wrote, the number of components in the system, the number of requirements.

We studied correlations, explored which indicators seemed to work – many didn’t, but one did.

We found a way to reliably predict how much work a work package would take.

We had uncovered the physics of the project – and we exploited it.

Transformation

Over a year, we transformed our capability. We went from not knowing how long a work package would take, to being able to predict in advance within 5% the total duration and effort.

This proved sublime. We had a team ‘in the groove’ where the work flowed.

They excelled under sustained pressure, in the sweet spot in performance. A fully data-driven optimised team. We handled change with aplomb, team rotation, knowledge transfer, holidays, training, risks.

We spent the next two years delivering, and winning a commendation from the client. That is the value of prediction.

Scaling Up and Lessons

Since then I’ve worked on many projects, and every single one I’ve sought to uncover the physics behind the work – the data, the analytics and the science that allow us to make those vital predictions.

I’ve learnt a lot. Systematic delivery pays off. Working out requirements up-front. Having disciplined ways of working, that get the job done and with creative freedom.

The value of prediction is enormous because:

A prediction gives you a map, a chart - you can steer the ship
It forces you to think and plan – forgetting something is expensive!

Since then I’ve gone on to found a business Project Science Ltd, where we help clients build data-designed and enabled projects. We help them to predict – and thereby shape – the future – through our services and products.

We’ve recently released an app that uses machine learning to predict how long tasks will take, using past actual performance data, and this is proving really exciting – a great application of predictive analytics.

Call to Action

I encourage you to take any opportunity you can take to make a prediction, and then compare the outcome with the forecast you made – you will learn.

Despite what they say, the past is a guide to the future.

Our challenge is how to relate the past to the present, and then the future.

This is a big challenge because projects are complex systems. That needn’t discourage us.

Physicists have made sense of exotic phenomena through science. We can do the same with our work.

Value of Prediction

Ultimately – the value of a prediction is that it puts us in control of where we’re going, and that builds trust with our stakeholders as we deliver our commitments – and it gives us the potential to exceed them, and find new ways of working.

If you want to find out more – there’s lots of material on my website www.projectscience.co.uk, come along to one of my talks with the APM, or find me on LinkedIn – get in touch!

Thank you.