Getting a business to become data-driven can be tough, especially if there is already a strong culture in place which makes decisions on instinct. So how do you get analytics adopted? David Reed looks at two examples of what works to deliver transformation.
“Data people are the most influential in our company.” Would you like to be able to make that statement? At the moment, there is at best a 50/50 chance of it being true, since around half of companies have adopted data and analytics and made the shift towards data-driven decision making.
In order for Pete Williams, head of enterprise analytics at Marks and Spencer, to be able to make the comment above required considerable effort and a cultural change in the organisation. “Retail is an incredibly intense and fast-moving business that makes lots of last-minute decisions based on lots of data. The goal is getting better at doing that and it was also something we had to do because of changes to our business model,” he told a Business, Analytics and Data Science meet-up organised by Pivigo in October.
As a multi-channel retailer serving 36 million customers, M&S is awash with data, but also had an embedded culture about how it decides to respond to their demands. “We will never be eBay or Google because we have too much physical presence which generates no digital trail,” he admitted. “So we had to pick and choose what data we had and make progress with that.”
Unlike many data practitioners, Williams recognised that data silos are an indicator that data is considered important within that function. In building an enterprise analytics function, however, that local adoption can also become an obstacle because managers do not want to lose their power over a source of critical information.
But Williams argues that, “the biggest danger is not thinking through the process.” Given this data incumbency and the retail culture within the business, how did he manage to drive through the change? “I used Spotfire as a catalyst for disruption - it has been amazing and has driven adoption in the business,” he said.
Before the project started, he describes the situation as one of “unconscious incompetence” in which the organisation didn’t know it was asking the wrong questions. “I started to show that data was available, it would help to understand customers and see results. That’s when people started to ask different questions,” said Williams.
Since running that proof of concept with the reporting tool back in 2013, his team has now grown to 200 who are collaborating and influencing decision-makers right across the company. While new technology proved to be the key to unlocking this shift, it was simply a tool to demonstrate how data and analytics could be deployed.
A much bigger change was needed in order to embed this new way of thinking. Williams explained: “There was a cultural challenge around vertical collaboration, getting to the leaders of the business. So we have connected analysts and data scientists with executive directors so they can drill into the data and explore what it means to the business.”
It is commonplace in the data industry to say that achieving a paradigm shift like the one experienced at M&S is always about people and processes, rather than technology. As this example shows, however, the impact of a new solution should not be under-estimated or ignored.
In fact, in the Royal Mail Data Services-backed research carried out by DataIQ, the database or CRM platform emerged as the single biggest obstacle to marketing becoming data-driven, identified by 42.9 per cent of respondents. This put it above the culture (named by 39.6 per cent) or budget (31.3 per cent). A further 30.8 per cent explicitly said that technology was holding them back.
Remarkable as it may seem in an era when everything from data, databases, analytics and marketing automation can be tapped into as a service, making technology access simpler and quicker than ever, there are still hard-wired IT barriers. Adopting data and analytics often involves deploying a new infrastructure which can have implications for what is already in place.
To be willing to undertake that, the business needs to be convinced it is worth it. Showing senior executives a new set of reports or data visualisations may be the best way of doing that. Or it could be down to the personal qualities and drive of the project’s champion. That was certainly true of Robbie Burgess, director of data and technology at RBI. As she recalled during the DataIQ Link conference, “in 2008, we were a niche publisher of magazines and Reed Elsevier put us up for sale because we didn’t look like them.”
Being put on the block is a tough place from which to start a technology transformation. Ironically, Burgess noted that it was the financial crisis which saved RBI because no buyer could be found. “So they decided to transform the business into being an information company,” she said.
That set out the vision for a raft of programmes which included creating a new customer database to provide insight right across the business. “It was the first time IT, marketing and data people had got together. It was like a trifecta bet - if you don’t get all three right, it will fail,” she said. “Our system and operating model is complicated - you can’t just pick out one piece and add a new one. It all has to join up.”
The early stages of that process were not pain-free because the technologists found it hard to understand why marketing wanted information about customers after they had already signed a deal. “They didn’t understand the need to build a relationship beyond the sale,” said Burgess.
Similarly, marketing was making mistakes in its approach to capturing information from customers through progressive profiling. She said: “They thought it just meant chopping up a long form into smaller sections, rather than building an information exchange and also not just asking for an email address.”
Getting this triple-headed team to work together involved understanding the strengths of each partner. According to Burgess, “the technologist is ‘Mr Clever’ - very bright and able to understand concepts quickly as well as being innovative. He is an excellent planner.” In fact, her first encounter with the agile working method was within the web development space.
Marketers, by contrast, are also smart and excellent communicators who can also be challenging. They adapt rapidly and well to new information during the course of projects. “Data people are smart problem-solvers who are truly ‘multi-lingual’, speaking both technology and marketing,” said Burgess. She has actually recruited linguists into her team for those skills, rather than specific domain expertise.
The idea that adoption of data and analytics might result from improved communications could seem fanciful if the transformation is only considered within a framework of hard factors, such as new systems, altered KPIs, cost-savings or enhanced revenue. But there are soft factors involved in moving the business culture towards data, ranging from working practices and cross-functional teams through to terminology and customer descriptions.
Any one of those could prove to be the trigger for change, just as much as the appeal of a new software application or visualisation tool. You never know what might happen until you look into its eyes...