The data conundrum: democratisation, commoditisation and the executive mindset

Reading Time: 4 minutes

The data conundrum: democratisation, commoditisation and the executive mindset

As companies grapple with making data pay, the role of technology leaders is front and centre. But it is never simple. Inherited legacy stock, conflicting viewpoints on off-the-shelf solutions and a lack of understanding in the broader business, mean data is complicated. Our recent Digital & Technology Leadership roundtable provided plenty of food for thought as CIOs, CTOs and CDOs thrashed out the challenges and opportunities in today’s data landscape.

 

Data – is a valuable asset but who owns it?

Democratisation of data was a recurrent theme. Attendees agreed there are challenges with users understanding data and that can undermine the effort put into data collection and analysis. One participant gave an example of a global healthcare business drawing a line between data that can be disseminated more generally and higher-level predictive analytics best kept in the hands of data scientists. Not true democratisation perhaps but a middle way.

“It’s a people problem,” commented a delegate from the private equity sphere; “Data culture, data literacy, and data politics, from the tools all the way to leadership and culture..” One leading CDO made the point that teams of today look very different from those of just five years ago; “Deep coding skills are being replaced by people with broader business skills.” Another participant questioned how to embed business analysts into existing business divisions. Data is everybody’s business but how it looks in practice is still a matter of debate.

Does a company need a CDO? This sparked some of the liveliest discussions. One delegate commented that the appointment of a CDO could imply that the CIO doesn’t understand data. Alternatively, could it be a device to raise the profile of data without giving the role any real authority? Most cynical of all, is the creation of a CDO role simply a ploy to allow the rest of the exec team to wash their hands of data?

They appoint a CDO thinking they will sort our data out… but the real work is to convince people in the organisation to use data better and that’s not solved by just creating a CDO role,” noted one attendee.

Other participants countered that the need for a CDO depended on the stage in the journey – some companies may appoint a CDO to unlock data they haven’t got the expertise internally to tap into yet.

 

Commoditisation and the search for utopia

What can you commoditise and how do you ensure it fits the purpose? A leading CDO with a private-equity background noted that in the past, private equity businesses would keep IP in-house as a value creation exercise. He commented that some parts of the industry have since become commoditised with solutions you, “Plug in and apply to some of your most difficult data problems… do things that would take a team of humans a very long time.”

The discussion moved onto “new frontiers of data analytics” with the father of the World Wide Web Tim Berners Lee and his ‘semantic web.’ This goes beyond purely numerical data to build graph databases of interrelated events, generating valuable market insight. But it does take time. “Building a graph database is like building a wine cellar – you must be patient,” noted one delegate wryly.

Other attendees cautioned that companies should start with a specific problem then work back to see what they need data for; “there must be a purpose.” And some delegates made the case that whatever black-box product you layer on top, you are only as good as the input data. There will still be a team of data engineers fixing data quality issues.

All agreed that a change in technology leadership can be frustrating. “You produce a utopia… a new team comes in and they create a new utopia,” noted one attendee. Another participant likened the choices made around data as being a “marriage” – you need to be considering whether this is still the right choice six years ahead, not now.

There was a common acknowledgement of the legacy challenge; “the need to transform outdated legacy data stack, show results quickly and deliver value for a CEO.” One participant took a pragmatic stance admitting that whilst multiple legacy systems are a non-ideal solution, they remain better than not having data at all; “it’s a balancing act.”

 

Alignment with strategic priorities is vital

How to deliver in the face of these challenges? One theme was echoed by many participants – the importance of ensuring data strategy dovetails with corporate priorities. “Pick out a few big bets and focus on just one,” advised one attendee. Getting an early win goes a long way towards championing the data cause and securing further buy-in and investment.

The relationship with the CEO and indeed all functions with the business is crucial. So too is the “executive mindset.” One participant noted that; “A CEO who already thinks of data as an asset is different from a CEO who you are trying to convert.”

The power to influence was seen as a crucial weapon in the data leader’s toolkit; “You are not just running a tech team; you need to partner with other executive members.” One attendee made the point succinctly; “You’ve got to make everyone around the table care about it (data).

What if the other executive functions don’t understand the data challenge? One attendee spoke of CEOs who, “Look for the silver bullet.” They hear about a new product, think it will solve all their data problems and then push it to the technology lead. But these CEOs may fail to appreciate the full extent of legacy challenges.

Using data as a platform for enhanced revenues remains the holy grail. But as one attendee put it; “That’s stage two, it has to wait until we have got our house in order.” Does the board really understand what it takes to get there?

The insights shared at our roundtable suggest data strategy and ownership must be a collective effort driven by all members of the executive, not the preserve of the data lead. With a cohesive approach, challenges can be more easily overcome and a clear path through the data landscape can be navigated.