How to manage data analytics programmes and teams
Data analytics technology won’t deliver business value by itself – it needs to be deployed, and its users organised to deliver value
Although data analytics technology has been around in one shape or form for years, it seems that organisations are still having problems in obtaining the full value from their often rather pricey initiatives.
The age of analytics: competing in a data-driven world report, released by the McKinsey Global Institute in December 2016, revealed that while a few – mainly digital native – organisations were using their data and analytics technology effectively, most were a long way from doing so.
The most successful initiatives were found among retailers and firms offering location-based services. But McKinsey indicated that companies in manufacturing, the public sector and healthcare were obtaining less than 30% of their projects’ potential value.
Beyond greenfield sites that have been built with data in mind from the ground up, Jason Foster, founder and director of data and analytics consultancy Cynozure, believes there are three main types of organisations that are struggling to getting it right.
These include companies in which senior leaders have said they are keen to benefit from data analytics but where the business is struggling to deliver.
There are also firms that have been investing and working in this area for a long time but, because they are still not gaining value, are becoming disheartened. Finally, there are the organisations that are a mixed bag in terms of maturity.
So although their e-commerce and digital marketing departments may gain a lot of value from their data and analytics activities, more “traditional” areas such as logistics do not.
The biggest obstacles to success are not so much technical as organisational, according to a study entitled Plotting the data journey in the boardroom: the state of data analytics 2017, conducted by software supplier MHR Analytics.
It found that while just more than three-quarters of the 300 UK-based C-level executives questioned planned to undertake a data analytics or big data project over the next 12 months, the most significant barrier to doing so was finding and training appropriate staff (42%).
Another key issue was developing a coherent business intelligence (BI) strategy for the entire company (29%), followed by how best to manage BI initiatives that are being undertaken by individual department heads and using BI to generate actionable business insights (28% respectively).
Overcoming business challenges
While Cynozure’s Foster agrees that finding people with good quality skills and experience is not easy and generally requires investment in training, he also points out that “at the component level, the skills are out there, but the issue is that people try to find a few individuals to solve all their problems”.
“Data analytics is a team sport and you’re unlikely to find all the necessary skills in one person,” says Foster. “You need to have a good mix of people with overlapping skills that complement each other.”
Such a team will likely be composed both of internal staff who have been given relevant training and people who have been hired in from outside, perhaps through graduate programmes. Key roles, on the other hand, will include software engineers who can build data pipeline frameworks to correlate data from a range of sources.
Also vital are data engineers able to model data in such a way as to make it accessible to business users, and data analysts and scientists. They analyse the data, uncover insights, present them to business users in a pertinent way and then collaborate to turn those insights into action.
Last but not least is the chief data or digital officer (CDO), whose job it is to understand the value of the organisation’s data and what opportunities it offers, while also orchestrating how it is handled and ensuring that governance is sound.
As for challenges around developing a coherent enterprise-wide BI strategy that does not end up fragmenting into individual departmental initiatives, Matt Jones, lead analytics strategist at data analytics consultancy Tessella, advises ensuring that it is jointly owned by IT in the shape of the CDO and the business.
“It’s important to have an overall, joined-up strategy and to understand where you want analytics to take you,” he says.
“If you don’t, the danger is that you end up with a huge technical programme that takes years to complete, when really it’s about having a vision and delivering wins quickly and iteratively.”
The issue is that a lot of organisations simply jump in and buy technology without thinking about the business problems they want to solve or the new opportunities they would like to exploit.
“Identify your problem, work out what data you need and then think about the skills and technology you require,” says Jones. “All too often, people gather lots of data in an analytics platform and then look for a problem to solve, but that’ll only threaten your return on investment.”
Cynozure’s Foster agrees. “While building a central data warehouse should be a goal, you need to pick a suitable use case and start with that,” he says. “You don’t need every department and every data set from the outset – start small and grow.”
Aligning around the customer
Another challenge people often struggle with, Foster points out, is developing the right “cultural mindset”. This is important, he believes, as “this makes things stick – it’s about how you turn your vision into an executable plan that will resonate and get buy-in.”
Original source: Computer Weekly