Opinion, advice, and analysis by the TNW community

Why your ‘data-driven’ strategy isn’t working

Data can't do all the work

Fergus Weldon
Story by
Fergus Weldon

Director of Data Science, TrainlineFergus is Director of Data Science at Trainline, where he uses Big Data to make sense of the UK train network, creating time and money-saving solutions for customers every day. Previous roles were wit… (show all) Fergus is Director of Data Science at Trainline, where he uses Big Data to make sense of the UK train network, creating time and money-saving solutions for customers every day. Previous roles were with companies such as Dunnhumby, Tesco and Nielsen. It was whilst studying robotics at Dublin City University that he also discovered his love of coding and hacking. Outside the world of data, Fergus enjoys CrossFit and is an accomplished Ultimate Frisbee player, winning some of the world’s biggest tournaments.

“We must be data-driven!” If we’re not led by what the data says, we’ll get left behind!” is a rallying call I’ve heard numerous times throughout my career. “Data-driven” has become a buzz-term in its own right and it’s not uncommon for businesses to get caught up in the hype 

A data-driven strategy is only half the solution 

Don’t get me wrong, being a data-driven organization in today’s world is critical to ensure businesses remain competitive but I’d challenge the common understanding of what being data-driven actually means.

Does having a data lake or warehouse and a business intelligence team qualify you as a data-driven organization? How do you attribute being data-driven to business performance? Can you really say data is driving your business success if you’re still using Excel for your analysis and reporting? 

The best decisions are led by insight 

I challenge the notion of being data-driven as, alone, it does not necessarily lead to positive business outcomes. “Data-driven,” as it is commonly used, makes too many assumptions about how the organization is using the data and even the value of the data.

One question, which is often overlooked in my experience, is whether the organization has the capability to use data in a manner that will yield beneficial results. It’s good decision-making which leads to better business outcomes; and how do you make better decisions? Through being insight-led.

It’s stating the obvious but it’s important to keep in mind that the best decisions are based on insight and observation — which can only in part come from data.

How to spot the warning signs 

Even with all the right intentions, it’s easy to get led astray by poor process, so how can you tell where you’re going wrong? There are a handful of tell-tale signs that you’re missing a vital step in your data-led approach. 

Firstly, alarm bells should start ringing when you notice that business decisions are being made in the absence of any insight or evidence. 

Following this, less obvious signs are likely to include very few people within the business having access to company data or even to the data experts. Along with this, it’s often common the organization lacks a strategy to up or cross-skill the team, which tends to hinder the development of a versatile team able to tackle a broad range of challenges. 

Staying on the right track 

While I can only speak based on my own experience and observation, it seems to me that many “data-driven” organizations are misguided on what it really means. To combat this, at my company we explicitly call out in our data strategy that our ambition is to be insight-driven, with data playing the key role — data-alone doesn’t cut the mustard for us so we’ve put several measures in place to combat this directly. 

Opening up access to data from within your business will pay dividends in the business insights it returns. Training is a vital part of ensuring your team can reap the most from your data and yet many businesses will limit training to a small portion of those handling data tasks. We facilitate training programs to ensure our team have the necessary skills to work with the data they are given. To compliment this, we put in place adequate governance to allow for safe access to business data.

To support those working outside of data-handling teams, our data scientists host weekly office hours to support the wider organization in their quest for true data insights. 

Final thought 

There are endless ways for companies to manage and govern their data but without the requisite skills, the necessary training and tooling to mine reliable evidence from the data, then even the best governance and strategy cannot lead you to better decision making.

Published November 15, 2019 — 11:00 UTC