Data has never been more valuable in decision making. Here’s how to use it effectively

This has been a year defined by unknowns: from the nature of the virus and how it affects different people, to the impact that the measures put in place to control it are going to have.

Dealing with unknowns and unforeseen factors is one of the biggest challenges that anyone can face in business.

For technology businesses disrupted by lockdowns and trading restrictions, perhaps the most critical concern has been how to keep cash flowing in order to cover overheads.

On top of this, every business wants to improve sales and be able to make predictions into the next year. 

These crucial, high risk, high impact decisions are of the sort that keep even the calmest among us awake at night.

For me, this unprecedented degree of uncertainty has emphasised the importance of how we can use data in our decision-making processes. 

Start with the facts

This year, I’ve found myself returning to a study conducted by the Swedish statistician Hans Rosling that formed the basis of his 2018 book Factfulness.

The basic premise of the book is that most people are wrong about the state of world. 

When you consider the best available data on income levels, health, safety and so on, you find that things are not as bad as you may think. 

The majority of people, however, have their perspective skewed by existing biases and other influential factors – such as the stories they read in the news. As a result, they believe the world is becoming a worse place to live.

Whether you agree with Rosling’s conclusions or not, the central argument that when making important decisions we should a) consult the best data available, and b) recognise and moderate for our biases, is one I find hard to dispute.

This is all the more important in our current situation, as making the right or wrong decision today could be the difference between your business thriving into the new year or collapsing entirely.

This is especially true global businesses, especially those involve in the complex inter-dependent worlds of consumer electronics manufacturing and supply chain.

So, how do you make sure you’re doing it right?

Select the best data

Your information has to come from reliable sources. Before you consider what the data is telling you, ask whether you trust where it’s come from. 

As is the case with many things, you’ll only get out as good as what you put in.

Similarly, just because you have an abundance of data available, that doesn’t mean you should use it all.

Ultimately, you should be selective: it’s better to source information from a smaller number of trustworthy sources, than to have lots and lots of possibly shaky data.

Parse it, consolidate, then act

The information you’ve gathered should be crunched by a select few people, and run through various validation checks.

I always involve myself directly in this process, so that I can measure my own point of view against that of my team and those to whom I report in.

Once the data has been seen and analysed by all of these smart people, we present it ready for feedback. We then take this feedback, calibrate it against other findings and responses that we’ve already gathered, before identifying trends and outliers.

At this point, it can be tempted to discard anomalies and information that doesn’t fit your assumptions or preliminary conclusions. But I would warn against doing this. 

Instead, question your reasoning. 

Weigh it with the perspective of your smart colleagues: am I discarding this because to do so would confirm my bias? What could the existence of this outlier suggest? How significant is it?

Then, with the answers to those questions and the feedback received: recalibrate.

Data itself is not free from bias, of course – University College London’s Dr Hannah Fry has written extensively on this –  but it’s still possible to account for these biases in your decision making: build a diverse team; be selective but representative with your information sources to ensure the highest quality; be prepared to ask the hard questions, and so on.

As you can see, this is a very iterative process – in which feedback loops counter biases should distil the information and outcomes to provide something you can act on.

What does this look like in reality?

One area that this process is particularly applicable to is acquisitions – which is itself a significant element of the high tech and electronics space.

If you take effective integration as a measure of the success rate of acquisitions, then most fail.

So what’s more valuable, the data from those integrations that did work out, or those that didn’t? Learn from the successes, I say.

Because the number of integrations that go well is relatively small compared to the number that fail, it’s possible to focus in on this group as a select, valuable dataset – without having your learning muddied by the multiple different ways that the many failures occur.

Of course it’s worth being aware of pitfalls and common mistakes, but in terms of data-driven decision making it’s better to use the success stories as your core set in this instance.

Losing focus on the desired outcome of an acquisition is one of the top reasons integrations fail. 

By concentrating on the successes – and how they were achieved – you ensure a clear focus on the ultimate goal of your own successful integration.

Ask yourself: what is it you want to want to achieve? Then follow that with the next best question: can the data tell you about how to get there?

Let what you know define your future. That’s as good a motto as any for data-driven decision making. 

I'm interested in your perspectives in decision making with quality data. Where do you think it falls down? And what tips do you have for others? Please let me know in the comments below.