Skip to main content

Data decentralisation - the next big thing?

Steve Farr 6 minute read Data quality

In the mid 90’s Volvo produced an advert that spawned a cliché - “I’m a control freak!”. It quickly moved from being a badge of honour however, to an insult. None of us want to work for one, though we may secretly wish that we were one… grasping hold of that elusive control.

End users of business systems want to take control of their lives to get the job done; make the boss happy; fulfil the KPI - whatever the business goal. This has manifested itself in a number of ways over the past 15 years and I’m sure you’ll all be familiar with at least one of these scenarios…

  • In product evaluations, worthy, functional software can often lose out to the product with a superior user interface and user experience because the end user finds it more intuitive.
  • Centralised systems controlled too tightly by IT will often get bypassed by end users with cheap desktop tools.
  • Whole areas critical to the business have shifted over to departmental control – just look at how website design and management commonly now sits in Marketing rather than IT, or how Tableau and Qliktech have gained enormous ground against the old BI giants

The next big thing?

So, what’s next? At Experian we talk to business users day in and day out and the vast majority have something in common – a desire to take control of their ‘mission-critical’ databases, predominately customer databases. These business users need data to do their jobs – be that running marketing campaigns, applying analytics or reporting to the business. All too often we’re told that when they can’t directly access this data (often because access is controlled by IT) then meeting business objectives becomes more difficult.

As businesses and not for profits increasingly adopt digital transformation strategies, a clear trend is emerging and the old-world order where IT controls mission-critical databases is beginning to crumble. We’ve noticed gathering momentum in the trend for businesses to give control of data back to those who need it. Today the trend is small, but we see rapid growth, starting with some particular types of businesses and databases. And it’s not just anecdotal - let’s look at the numbers.

We conducted a research campaign across 853 mid to large sized enterprises across the UK, the US, Australasia and SE Asia during the Spring and Summer of 2018. They told us that while only 5% had divested control of this data from IT to the departments (Marketing, predominantly), a staggering 81% wanted to. And we saw no significant variations by either vertical industry or geography.

Fuelling a growing trend

We asked what was behind this control-freakery, and there were two major conclusions:

1) Speed

The relationship between IT and the business is threatened by an increase in the pace of business. Fewer than 1 in 5 businesses are agile enough to turn around a data request in less than a day, and half must wait a week or more.

2) Data quality

Secondly, businesses feel that there is a lack of control over the quality of data. There are numerous underlying reasons for this including human error or a lack of IT time to fix problems.

In combination, this leads 86% of respondents to believe that their lack of control over data negatively affects the customer experience. Moreover, the same percentage believe that the solution is to move responsibility for data quality firmly away from IT.

I want doesn’t mean I get (or does it?)

Clearly departments want to take more control. But how practical is this, given that only 5% of companies have made this move to date? “I want” doesn’t always get, as every toddler is told.

So, we dug a bit deeper, and this perhaps was the most surprising finding. A massive 42% of companies declared that they were likely to move to a decentralized model within the next two years. This is a staggering fallout from digital transformation and must be placed in the context of the GDPR and recent publicity about data breaches, which, one would imagine, would place a brake on decentralization. This is a nothing short of a sea change.

What if you want to become a control freak?

If you want to start doing the groundwork, here’s a few considerations to get you started:

1) Tools

For businesses to truly take control of their data, we need to give them the tools to do it. The revolution in data visualisation would not have happened without the efforts of Tableau, QlikTech and others. Without massive improvements in processing power, and cheap deployment the AI revolution would have been a non-starter. If businesses truly want to manage, cleanse and enrich their own data, the tools need to be there. Look for tools that can be used at a departmental level and allows those that understand the data to be the ones that dictate how and when it is manipulated to business effect.

2) People

Data literacy is an essential skill in today’s digital workplace, just as much as Word or Excel used to be. Experian has just become a founding member of the Data Literacy Project to help champion this important issue both in businesses and at an educational level. Check it out for ways to get support and get involved.

3) The Business Case

Quantifying the lag that centralization can cause, may look too much upon the negative side of the equation. Building a case for giving business people more control must look at the positive effects to be had upon processes, and assigning a value to them. Faster and better management of customer data might mean more cross and upsell; better customer satisfaction ratings; and even fewer product returns. Better control of inventory will mean faster stock turns, less wastage and fewer complaints. These types of statistics will be readily available in most organisations and the business is ultimately responsible for them: they need the means to control them.

Our own data management platform, Aperture Data Studio, is built with the business user in mind. We’ve created a short video to bring to life just how straightforward it can be to get hands-on with data.