Practical Guide to Delivering Value

A common problem in the technology world is that we focus too much on the technology and too little on the people, and especially on the way the brain works. Central to this is the ability to frame problems correctly, and Richard suggested this is the root of value in any analytics work, and that the value that flows from this is at the heart of digital transformation more widely.

Whether slowing Chinese demand, Brexit or the prospect of tariffs, there are a number of challenges facing the company, and they have responded with a range of programs operating under the Charge & Accelerate umbrella.

The corporate analytics look after everything but the vehicle software behind things like autonomous vehicles. They have a team of 40 people and began their journey 2.5 years ago. The team aim to drive transformation across the business so that analytics becomes part of the DNA of the business. They operate a hub-and-spoke approach, with the core team acting as a centre of excellence across the business.

In the last financial year, the team were able to attribute £146 million of value to their work, and this wasn’t just their own bravado as it was signed off by an internal audit. This represents a 2,000% ROI for the team as a whole.

This value was delivered via a number of projects. For instance, they developed a bayesian model to help the company forecast more effectively. They also developed a graph database to help the company manage a hugely complex manufacturing process. Whilst these were pretty cool projects however, none really contributed to the bottom line figure. What really moved the needle was problem framing and structuring. It’s the human brain that delivers the real value, not the technology.

A Practical Guide to Delivering Value

There are three main aspects to help you derive value from analytics: process, training and diversity of thought. A key process to get right is to ensure that both the business problem and the analytics problem are defined at the same time. Only then does that work filter down and the solution fleshed out and implemented within the business.

Whilst the business and analytics problems are determined concurrently, it’s important that the structure of the problem

is distinct from the technical design. They’re two completely different processes with two different teams working on them. The problem has to be clearly defined before technical solutions are proposed.

We live in a world awash with content to help us learn all
manner of technical disciplines. There are books, websites, MOOCs, conferences and workshops. Even amidst this deluge of opportunities however, there remains precious little available on analytics problem framing. This shortage of material resulted in JLR writing their own, complete with case studies. It represented one of the best investments of time in their entire analytics journey.

Diversity and quality of thought allows you to narrow the search space with brain power rather than computing power. This is an important distinction as in technology projects it can be tempting to just throw computing power at the problem in the hope of cracking it, without ever really defining the problem to begin with. We don’t want to turn unique people into robots when our talent has so much potential for high quality thought. This is not easy however, not least when you’re asking people to think differently to the norm in their team and their organisation.

Pitfalls to Avoid

1. Quantify your value – It’s tempting to think that value can’t be quantified of your analytics work, but this is a mistake, even if the value is not financial in nature.
2. Learn to challenge respectfully – It’s really difficult for people to challenge institutional norms or those in senior positions, so this is a big cultural issue to tackle.
3. Don’t let your processes stifle change – If your process stifles innovation and fresh thinking then analytics will always struggle.