Data governance, process culture and predictive modelling in a post-COVID-19 world were all hot topics as The Bridge Analytics hosted a virtual roundtable about data strategy adoption experiences across multiple industries.
Hosted by professional Chief Data Officer (CDO) Barry Smith and The Bridge Analytics' Dan Holden and Chris Fernyhough, the successful online event was the first of its kind for the team and bought together and bought together senior stakeholders in data management to discuss and share strategy integration tips, best practice and challenges.
What is a Data Strategy?
A data strategy is a plan that enables a business to identify, acquire, store, manage and exploit data to achieve the business objectives, and as such needs to be prioritised alongside other business transformation. A data strategy needs to be a balance between being on the ‘offensive’ from a competitor advantage point of view as well as ‘defensive’ from a compliance and regulatory perspective.
Integrating a data strategy can prove problematic in different ways across different businesses and the panel discussed some of the sticking points they’ve encountered. Host Barry, who has over 20 years of experience in business change, IT, business intelligence and data governance with leading brands like Npower and Virgin Media, led the conversation on integration and process ownership.
Integration and Process Ownership
“In my experience some heads of data said their strategy was being adopted early and flourishing, driving the organisations to become very data centric. Others were seeing the opposite and while part of the data strategy was being adopted and pushed and part of it couldn’t get off the ground.”
Asked whether therefore a data strategy needs to be a separate entity rather than something that’s embedded within other business transformation, Interim CDO James Morgan offered his insights:
“What I’ve found that’s worked for me over the years has been a blend of both. I’ve fallen short in the past by trying to run the data transformation and data programme just via funding from other programmes. The problem is when some of those fall by the wayside you’re left exposed and the cost ramps up the others. What I’d therefore move towards is having a large element of any implementation of a strategy - the ‘plumbing and drains’ element - existing in it’s own right, but then having it couple into as many of the other activities as possible.”
Director at Oakland Andy Crossley agreed on the all-encompassing scope of a data strategy as opposed to a singular focus:
“There’s no point having a data strategy that isn’t embedded in all facets of the business because otherwise why are you doing it?”
Having clear and understood data owners, as well as stakeholder buy-in, was identified as a critical part of any data strategy, as CDO’s must rely on a concise and consistent data pipeline from across many areas of an organisation, as Director of Analytics and BI Richard Longhurst put it:
“You have to think about who are going to be the users of that project. In terms of reporting output you also have to think of GDPR principles - who are going to be the consumers within the business, are they going to self-serve or are they getting reports sent to them? What data is going to be required to feed that project? I think the head of data needs to be providing all the use case templates across all of those projects so there’s a standardisation in terms of how those questions and those items are answered.
In terms of the data that’s going to be serving KPIs, you’re going to get the same data points that are going to be coming across all of those projects. So rather than having a change team going back to the head of data saying “we need to do this report or that report”, actually all those questions will be coming together to the head of data to figure out that they need to capture one data point to answer many questions.”
Having consistent and reliable data owners across a business for this can often be a challenge, and one strategy to tackle this is to make data stewardship and responsibility role specific and part of the job description and associated KPI’s. This encourages employees to take part and solves the issue of data management models failing should individual employees leave. As Data Practice Lead Catherine Wilks put it:
“It’s not just as simple as agreeing the business data owners because they’ve got to understand what their role is, its importance and be committed to performing it. I think that getting them to agree to do it is sometimes the easy bit - getting engagement is where I’m seeing clients can sometimes struggle.”
Showing ‘quick wins’ when implementing a data strategy can often help highlight the importance of a good strategy even in highly sales-driven businesses where data management practice is seen as an unnecessary obstacle, as James Morgan explained:
“The purely data stewardship and governance angle won’t work - it has to be understanding what they’re driven by, their sales targets, what their objectives are and how you can help them achieve them.
The one that really makes some senior sales people take notice is just showing them how some of the shocking quality of individual line level records of their B2B prospects and current accounts is damaging their revenue. Often this is things like the same company on the system with five or six different names, different account manager names and misspelled classifications of the business type. It’s effective if you can name and shame who entered that onto the system and then demonstrate the cost as a result. This can be the amount of comms they do to the wrong people or the inability to actually contact them and sell to them because they haven’t got their details right.”
Interim CDO Ed Wynn gave an example of how a simple piece of streamlining work at one business opened to door for buy-in for a more enhanced programme of data management:
“I was at a building society for an extended period of time and it was obvious there was so much information that didn’t seem to be being used. We quickly catalogued it out and there were over 5,500 different reports in circulation across the organisation. A quick win there was to rationalise it down to about 2,000 over the course of a couple of weeks. That’s is probably still too many but it was amazing the traction we got from doing a simple thing like that. It’s really just housekeeping but it got badged up in a different way.”
Data in a post-COVID-19 world
The way data is being managed and modelled for prediction is changing rapidly as behaviours change due to the Covid-19 pandemic, and staying agile to that change is key. For example, many supermarkets are facing challenges when using their last 10 years worth of sales data to form future action models for stocking stores as consumer behaviour has changed substantially in the last three months. Businesses face a challenge to become quicker in analysing fresh data, as Group Head of Knowledge Management at Pepper Financial Services Group Jon Haigh summarised:
“Aside from the horror of everything going on with Covid-19 at the moment this is going to be fascinating time with a lot of interesting things coming out. The whole remit of predictions based on what you’ve done in the past and what happened two years ago is now somewhat irrelevant. That hopefully gives you a scope to speak quite differently about the way of thinking.”
The roundtable ended with a discussion on how there might be a potential change of attitude surrounding how the new data being sourced for the NHS in order to manage the pandemic would be managed - and how long that highly personal data would be able to be stored for. Jon Haigh reflected:
“If you’d have said three months ago that Google and Facebook have been releasing data as they have been thats basically tracking people there would be uproar, but that has sort of changed now. Who knows how long that will be allowed for?”
Interim Head of Data Craig Stirk said of the session:
“It was an excellent conversation. A key thing for me was how it showcased that it’s easier to sell your data strategy if you can demonstrate how it releases benefits for other change programmes and the business. They will become advocates for you and it’s easier to fund your data strategy if you can get it classed as a foundation activity.”
With experts already worried about the reliance on a centralised database when it comes to storing data for the new app, there’s sure to be plenty of discussion as the system is rolled out to the general public.
The comprehensive findings of the roundtable will form the basis of an upcoming whitepaper from The Bridge Analytics.
To keep up to date with the latest tech news, follow us on Twitter @BridgeITRec