How do we unlock data’s potential and ensure it works effectively, efficiently and fairly within our organisations? The Data Management Specialist Group summarises a recent virtual event focused on helping data practitioners and leaders develop their skills and their teams.
Making Data Good for Society was a recent virtual event hosted by the BCS Data Management Specialist Group and DAMA UK. The event saw members and data practitioners enjoy ten webinar sessions, each focused on a difference aspect of managing data.
Taken as a whole, the event explored practical aspects of gathering, owning, managing and leveraging value from data. Talks also explored how correctly curated data can help organisations innovate, develop new products and solve valuable questions.
The event’s final panel debate provided a route in to summarising the event.
Watch the video
Panellists included:
- Karen Alford - UK Environment Agency
- Nick Sorros - Datakind
- Stuart Kitney – Head of Department - Data Science, National Physical laboratory (NPL)
- Neill Crump - Digital Strategy Director, Dudley Group NHS Foundation Trust
- Julian Schwarzenbach - Chair DMSG
- Mark Humphries - Chair of DAMA UK
1. How can organisations get benefit from data?
‘Organisations used to limit their thinking about data to just dashboarding and reporting,’ says Karen Alford from the Environment Agency. ‘Organisations are, however, becoming much more mature in their approaches and thinking. This maturing is evidenced through data being thought of much more holistically by businesses.’
‘Data is king,’ Alford explains. ‘Today we can get so much more value from data and also from technology... we’re seeing so many more opportunities.’
Nick Sorros from Datakind echoed these sentiments: ‘Organisations are becoming much more mature in their understanding of data, their expectations and their wrap-around processes. This maturation is, however, very much an ongoing process which needs to be driven and nurtured within organisations.’
‘When it comes to creating a “data mature” organisation, there are two key ingredients,’ Sorros believes. First is leadership. Leaders need to “buy-in” and they also need to be data-literate. They must be technically informed - they need to know what is possible and what’s realistic.
Next, leaders need to focus on creating a data friendly culture. This can involve inspiring, training and enabling people to use data. ‘Organisations will only get the value from data if there is the right culture within the business,’ he says.
Neil Crump explains that, within his organisation, they are working to realise a data friendly culture through creating job roles specifically designed to build awareness of data. With these people embedded in the business, conversations about problems, opportunities and solutions can be facilitated.
Crump also says that providing people with the correct tooling is imperative. Leave your team working with old, inefficient or out-of-date tools and they spend their time using the tools and not making use of the data. ‘Correct tooling [frees] a lot more time to have the correct conversation,’ he says. ‘And these conversations lead to the outcomes we’re looking for.’
‘There also needs to be confidence in data.’ explains Stuart Kitney. ‘[We] need to understand the data’s provenance: where it has come from, how it has been captured, what method was used to capture that data and whether it is fit for purpose.’ If we trust data, we can use it to its full value.
2. Can an organisation succeed if it doesn’t have a data literate leader?
‘These roles are absolutely key,’ says Kitney. ‘Ownership and responsibility for the data held within an organisation - there has to be accountability - not just for the data but for its structure, the tools that are applied to it and ownership of the data. Every organisation should be implementing these roles.’
Karen Alford echoes these sentiments: ‘It’s essential. I have this type of role and I ask all [of] the awkward questions. Sometimes we think IT or a piece of technology can do everything for you but if you don’t have good data - that is fit for purpose - the technology isn’t going to bring the benefits that you’re expecting.’
3. How to raise levels of data awareness and literacy
Unlocking data’s power and potential at an organisational level, is all about educating and empowering people. This journey starts with leadership and permeates an organisation through cultural change.
Nick Sorros believes that an organisation’s journey towards data literacy starts with data owners and leaders making a strong case for its use. ‘This process is akin to role modelling: sharing, discussing and learning from case studies of data adoption within organisations. It is about exploring what data makes possible and how these possibilities can be unlocked.’
Stuart Kitney explains that, in the world of science, things are certainly changing. ‘We’re refocusing how scientists think about data,’ he says. ‘Historically, careers were built on peer review publications and your peer network. What’s becoming more important is reproducibility and providing your base data sets alongside publications... Using internationally recognised data repositories and making that data openly accessible...'
This is critical, Kitney explains, because it brings about a profound change - it changes the scientist’s mindset from focusing on the peer review paper publication and encourages them to see that the data is just as important.
‘The data is more important than the research publication. That encourages a want and a need for data to be curated in a fit and proper fashion,’ he says.
The net result of this change in mindset is that science becomes more efficient. Duplication of scientific work can be costly and time-consuming. Reproduction allows other scientists to start where their colleagues may have stopped and so build out and upon others’ work more effectively.
‘When we get to this stage, peer review publications will be the secondary output. It’ll be all about the data sets - how you produced them and which tools you used. That’d be a massive culture change in science. It’ll build leadership. Our future scientific superstars will build on the data sets they’ve developed and their experimental knowledge [not just] their research and their results,’ he concludes.
A data literate society may also be a healthier one, Neil Crump explains. Traditionally, healthcare has taken a ‘paternalistic’ position on patients’ data. A health provider would effectively own our data and use it to provide us with a diagnosis.
In the future, however, this model may be flipped on its head: citizens could be provided with data about their health - ideally so they can make informed decisions and prevent themselves needing medical intervention.
Crump, however, concedes that this ‘prevention through data model’ of health care has one critical blocker: language. Clinicians, like all professionals, have their own internal bank of words, meanings and usage.
‘What we need to do now is work out how we can share data and insight in a very easy and accessible way,’ Crump says.
'A way in which people can take meaning and use it to improve their own health. We won’t release all the data about people’s health. Rather, we’ll release really important data about people’s health and put it in a context which people can understand.
'Diabetes is a good example. What are the important points around the diabetes pathway? Can we share back to the patient so they can control their own health? Every condition has its own intricacies, so, we’ll talk to citizens and ask them “What’s most important to you?” We’ll co-produce and share more data than in the past.’
4. Why do some projects fail?
Why do some projects fail while others succeed? One answer might be: ‘[Getting] carried away with sexy technology before you get the data bits right,’ says Julian Schwarzenbach. ‘They might spend a lot of money on a technology fix but are forgetting the data. Building up the foundations - getting the data right first is fundamental.’
Karen Alford from the UK Environment Agency echoes Schwarzenbach’s observation as she reflects on how her organisation has clear data standards - but they were written in a document which nobody knew existed.
‘I’ve been making them more visible, more accessible, and making them API and machine readable,’ she says. ‘That lets us humanise the data - we give things the same name in different places. [Previously] we’d call the same physical structure something different depending upon the asset family we were talking about. But, the biggest thing on this journey is making sure that the data is fit for purpose.’
Summing up, Schwarzenbach says: ‘We need to get the fundamentals right, so we avoid the need for data magic.’
5. Key data personnel roles for a successful project
For you
Be part of something bigger, join BCS, The Chartered Institute for IT.
Mark Humphries explains that data scientists have been portrayed as the rock stars who can solve all our problems. ‘They’re the first to admit, however, that they don’t exist in a vacuum,’ he observes. ‘There are other data roles - other data skills.’
The panel sees the following skills as being critical to organisational and project success:
- Defining and understanding. Conceptualising how data fits together such that interoperability can be achieved.
- Networking skills. The ability to work with other organisations and stakeholders.
- Analytics. Data scientists have analytical skills in abundance and at a very high level. What’s equally important is having a diverse pool of widely skilled analytical generalists.
- Coaching skills. To improve organisation-level digital literacy, there’s a need for data experts to take their knowledge out into the wider business and have conversations with everybody.
- Engineering. People who can use standards and bring different data sets together so processes can be automated.
‘We also need to think about the end user,’ explains Stuart Kitney. ‘We need multi-skilled expertise and experience. We need to put more emphasis on skills and education for, maybe, a clinician or pathologist - somebody who is now being asked to apply machine learning and algorithms in quantitative imaging.
‘People need to understand how AI is going to help them make better decisions. You’ve got that challenge in many fields. The rock stars need to be the people who apply the tools, not the data scientists.’
