In our new report, Inside the Datavores, we quantify we think for the first time the link between higher levels of online data use in UK businesses, and their economic performance in terms of productivity and profitability.

Everyone is talking about data. Magazine articles, white papers from technology vendors and even dedicated television programmes extol the benefits of ‘big data’. Case studies in the business literature abound – Netflix, King.com, Spotify, Channel 4, Tesco.

Why should you read our contribution? Simply put, because we need more data about data. Businesses need quantitative information on how more effective use of their data can impact their bottom line. And they need quantitative research on what management and organisation practices can maximise their return on investment.

It would be ironic if the business case for using data was based on generalisations from case studies, descriptive analyses and self-reported measures of performance. That’s not the data-driven way of establishing the benefits of data.

The stakes for UK businesses are high. Our previous report, ‘Rise of the Datavores’, reported that only 18% of UK businesses in the internet economy rely on data and analysis when making decisions on how to grow their sales, despite all the vaunted benefits. Understanding how to grow this number should be a top priority for the UK economy.

So, what is the economic reality behind the data hype? What do businesses need to do to go beyond data, how should they transform their processes in order to benefit from the ‘data explosion’?

These are the questions that we seek to answer in ‘Inside the Datavores’. The report presents the findings of an econometric analysis of the link between use of online data (measured using the survey data that we collected for ‘Rise of the Datavores’) and productivity and profitability of a sample of UK businesses. We source data on the latter from company accounts, and in our models control for a host of relevant variables, like which sector a company is in, how old it is and how many IT specialists it employs – all variables which we’d expect to impact on business performance.

What do we find?

1. There is a data dividend. We find a strong and statistically robust link between higher levels of data use (a metric summarising various data collection, analysis and deployment activities) and business productivity, in terms of value added: those companies in the top 16% of data use are, other things equal, 8% more productive than their competitors. Their enhanced productivity means that they are more efficient and competitive than businesses which do not make such big use of their online data.
2. When we consider different types of data activity – collection, analysis and deployment – separately, we find that higher levels of analysis have the strongest connection with productivity. Analytical firms that probe the data they collect with a wider variety of methods (such as controlled experiments on their websites, or data and text mining to discover patterns in their customer data) are 11% more productive. They also see a boost in their profits: their EBITDA (Earnings Before Taxes, Interests, Depreciation and Amortisation – a measure that captures how good is a company at generating net profits per operations) per employee – is £3,180 higher than the average, and their return on equity is 4.3% higher.
3. Collecting online data is of little value on its own. Firms that do not analyse their data or act on its insights do not enjoy productivity benefits or higher profitability.
4. Businesses need to go ‘beyond data’, and reorganise their workplace and adapt their processes in order to experience the full benefits of their data. We find, in particular, that those companies that give their employees more autonomy, and allow them to act upon their data driven insights, see a boost in productivity four times higher than those which micromanage their employees. Contrary to the idea that more data will result in ‘big brother’ workplaces, our results suggest that data creates more value in flatter, more democratic organisations.

What do our results imply?

Our analysis makes it clear that, despite the hype, managers ignore the potential of data at their peril. At the same time, building up their IT infrastructure to collect and process more data on its own will do little. The data needs to be probed and analysed, and their workforce needs to be empowered to act on what is learned.

This raises a host of issues around workforce and management skills – a topic Nesta is exploring in partnership with Creative Skillset and the Royal Statistical Society in an ongoing data skills inquiry as part of the Government’s Data Capability Strategy (see this blog post for a summary).

We look forward to your views and comments. Drop us an e-mail/tweet if you have any questions at @hasanbakhshi or @JMateosGarcia

Image from Datamatics by Ryoji Ikeda, taken by Von Boot