ktsdesign - Fotolia

CIO interview: Mike Young, Centrica

We speak to the CIO whose CEO recently met with the head of Microsoft to talk about strategy

The fact that CEOs from the top software companies are visiting Centrica is indicative of a company that is evolving way beyond its original remit as parent company of British Gas.

Microsoft CEO Satya Nadella is keen to involve Centrica in a strategic initiative for the utilities sector, while for SAP and CEO Bill McDermott, Centrica runs one of the largest SAP instances, comparable to the systems run at Apple and Microsoft, and it is set to migrate to SAP’s in-memory Hana enterprise resource planning (ERP) system.

CIO Mike Young heads up a team at the very forefront of data science, where algorithms are not only helping Centrica manage more than 23 million customers, but are also providing a new revenue stream in data discovery through a spin-off called Io-Tahoe.

Computer Weekly recently spoke to Young about how the data strategy was spun out and the strategic partnerships with Microsoft and SAP. “The journey for Centrica started at the back end of 2014,” he says. “By 2015, Centrica had put in place a data lake with MapReduce, Spark and all the usual things you would expect in a data lake. All our structured and unstructured data was going into the lake.”

During the process of establishing a data lake, the company had developed quite a lot of intellectual property, says Young, adding: “As we got grips with the data lake, we developed our own algorithms.”

The company set up a £100m venture capital arm and acquired a complementary data tools company, Rokitt-Astra, which became part of Io-Tahoe. “We brought the two together so we could sell data services to banks,” he says. “With GDPR [General Data Protection Regulation], people want to discover what is in their data.”

Young says Centrica’s data science team is now 300 strong, with specialists in analytics and visualisation. “I have MIT [Massachusetts Institute of Technology] and Cambridge-qualified people as part of the team,” he says. “But two people in the team came from field engineer and customer operations, who bring their experiences to the job. This mix gives us the capabilities.”

Young adds: “We look at real-life problems in Centrica, do a bit of R&D, and create minimum viable products.”

“We look at real-life problems in Centrica, do a bit of R&D, and create minimum viable products”

Mike Young, Centrica

For instance, he says the team has built a live tracking system using machine learning, which tracks customers’ activity online and can provide a level of detail of the customer’s journey on the website. This can be used to pinpoint how far the customer got before the journey failed, resulting in them needing to call the contact centre.

“A year ago, operations didn’t have this information to hand,” says Young.

The team set out to identity something that could speed up call centre operations and have the customer at its heart, he says, and the project took just six weeks to develop an initial release.

After further refinement, the project was ready for deployment in the contact centre, says Young. “We are able to glance at why customers’ journeys have failed and can work on remedies so that the problem does not occur again. Now we rarely see a repeat failure.”

Meeting Satya

In a recent blog post, Young outlined how Centrica had formed a strategic partnership with Microsoft following a meeting with Nadella. He wrote: “What’s really exciting about our partnership with Microsoft is that it means we can move at speed, taking part in innovative pilots and learn how the leading edge of cloud computing can help us accelerate change and move into new markets.

“We heard today about the future of artificial intelligence (AI) from Microsoft, and so we are delighted to have our data science and digital teams working with them, developing AI and machine learning capability that will transform the ways we interact with our customers.”

Young says of the partnership: “Our CEO, Ian Conn, sat down with Satya Nadella two week ago and talked about what we are both doing in Azure, and where we could benefit from doing things jointly.”

He believes Centrica will benefit greatly from a close working relationship with Microsoft. “Satya has something very unique and chose three key partners,” he says. “Centrica is the only one in the energy sector.”

Centrica’s CEO was asked by Nadella to take part in an evaluation of some new big data tools Microsoft has been developing. Young says: “This is a strategic partnership. My chief data scientist sits on the advisory team at Microsoft and has a real say in its roadmap. We have the ability to lean on the 300 data scientists at Redmond. I expect we will end up with something we can use.”

An audience with McDermott

It is a similar story with SAP, another of Centrica’s strategic software providers. Young says: “We have a strategic partnership with Bill McDermott. It will take two years to take on board our legacy systems and move them over to Hana.”

Explaining the challenge Centrica faces with its back-end systems, Young says: “We have 23 million customer across the group. I have eight significant ERP systems in the UK. We run two of the biggest SAP platforms. It is the same system used by Apple and Microsoft. It is a big batch-fed system and is 24 hours behind real time.

“How do we take this batch system and get real-time data so that when a call comes into the call centre, the operations team has the most up-to-date information about the customer?”

After a meeting with McDermott, Centrica has agreed a plan to migrate to an end-to-end digital back end based on SAP Hana to overcome the limits of batch processing. Young says: “We have moved our central financial ERP from from legacy SAP to SAP Hana, which means data is updated every second.”

Read more about Centrica’s IT strategy

The procurement team has been moved onto another SAP product, Ariba, while WorkDay provides HR. “The tricky piece is in the back-end billing and ERP system,” says Young. “We have put in an API layer and use machine learning to trawl through the batch systems, and repair data in the back-end systems.”

According to Young, these measures have enabled Centrica to speed up processing time from the previous 24-hour limit in the batch system to just an hour.

From an enterprise software perspective, field service is another area of the business that needs addressing, says Young. Centrica’s 50,000 field engineers use a patchwork quilt of 11 different systems, but over the next two years, “we will be using a new cloud-based field management system”, he says.

Digital challenges

Young has rolled out digital transformations across a number of corporations and, in his experience, it is never easy. “There are three things to do,” he says. “Use the right tech for the right data outputs, wrap this round a proper business process in the most efficient way, and third is the people. If you haven’t got the people trained, then you have a problem. That is always the hardest to address.”

Young adds: “A third of your employees get it, adapt quickly and become your digital agents, a third need coaching and time to adapt, but a third never really get it.”

To help everyone along at Centrica, Microsoft Office 365 is being used as a collaboration stack, says Young. “This is not just email and Skype for Business, it is also Microsoft Project and PowerBI.”

Digital transformation is often described as a journey. Young says that at Centrica, there are three levels of maturity.

At the most basic level of digital, Centrica runs lunchtime sessions, which are used to help people understand what “digital” means to Centrica, he says. The next stage in digital sophistication is helping people do their jobs through digital initiatives. To help, Centrica has about 600 digital champions across the business, says Young.

At the highest level, Centrica has recruited a digital leadership, which can be parachuted in across the business to investigate opportunities in digital transformation, he adds.

Read more on CW500 and IT leadership skills

Data Center
Data Management