An interview with Anthony Morris, ex Director of Data, Analytics and Insight at Dixons Carphone
Introduction to Anthony
Anthony Morris spent 10 years as a Management Consultant across a number of different businesses, supporting with different types of challenges from growth strategies and cost reduction to turnaround. This led him to Dixons where he spent 12 years in a strategy role, which soon developed into a specialist position unlocking the value within the business through increasing the use of analytical data. There he built a 100-strong team focusing on business insights and intelligence, supporting all areas of the company to make better use of their data.
In this interview, Anthony helps us to understand how retailers can better use data to their advantage, and why a pragmatic and intelligent approach is crucial.
What are your biggest frustrations when it comes to how retailers use their data?
Firstly, it’s the ability to bring together efficiently multiple data sources to make faster, more effective decisions. Retailers typically have rich sources of data, but often it’s not readily available in a structure that can be easily combined to deliver actionable insights. This lack of prescriptive insight is one reason why historically retailers have had to rely a lot on gut instinct to make decisions.
Retailers need to be responsive to their customers and previously it wasn’t easy to be able to use data fast enough to adapt. Whether it’s prices, promotions, product range – retailers need to be able to efficiently respond to and anticipate customer needs.
Now data has become more prevalent and more sophisticated tools, like AI, are available to mine the data, retailers need to develop their advanced analytics capabilities to ensure they can make smarter, customer-led decisions. It can be challenging to build a deep understanding of customers and behaviour – having the ability to look forward rather than backwards with predictive reports. Data is more often used to hone the existing offer and make incremental refinements than to drive more radical changes to the overall customer proposition. The key to enabling more effective use of data is ensuring that it reaches decision makers in a timely manner and can be easily used by the people making day to day decisions. Data maturity is a perpetual process and needs to become part of the DNA in any business – minute by minute, hour by hour retailers are amassing more data from which they can optimise all aspects of their customer and product lifecycles.
Something many are missing is effective marketing measurement – when we can understand customer behaviour across channels we can more effectively shape propositions. It’s easy to look at the wrong metrics – often, companies focus on the metrics available to them and try and optimise them. You can’t easily measure the underlying drivers of customer satisfaction, so you put that aside and do it less often – but then you won’t get under the skin of that data. Instead you focus more on sales, stock, margins. It’s easy to get into that routine with the blinkers on, only focused on what you can see meaning you effectively put aside the things you should be looking at.
Another issue is the blind use of tech when interacting with customers. All too often personalisation is clunky – whether it’s emails missing names or excessive retargeting. There are some AI techniques offering great potential, but we need to ensure we get good value from them. A poorly designed chat bot is worse than an interaction with a human. It’s vitally important that businesses can keep customer experience in mind at all times rather than blindly chasing tech for the sake of it because they are conscious of the importance of personalisation – even though it’s great they’re trying to utilise innovative solutions.
Last but not least, it’s the inability to be able to measure what really works. There are often many tests going on within retailers at any time, but without a robust testing methodology, experimentation strategy and platform they typically default to that confirmation bias or look at tests and pick those you like the answer to rather than selecting them based on validity and accuracy. In the vacuum of having reliable robust data it’s too easy to make the wrong decision.
Do you think that confirmation bias (and other biases) are a problem for retailers when they analyse their data?
Confirmation bias is an interesting concern, because everyone has a natural tendency to seek information that confirms their existing beliefs. Unfortunately, this type of bias can prevent us from looking at situations objectively, impairing the decisions we make and ultimately leading to poor or faulty choices.
The key here is going back to that measurement aspect – being able to use the data intelligently and honestly to support decision making. Being able to run good experiments and having timely access to data is really critical.
In terms of other biases, something that is closely related to confirmation bias is the belief that what has worked in the past will work again – and conversely what hasn’t worked won’t work in the future. Preferences change and evolve – the way people interact with your business changes. The last decade or so has seen a dramatic change in the channels people are shopping through, how different segments behave – so the precedent is no longer there. Old approaches to forecasting are no longer so valid which creates challenges, but also opportunities provided you can efficiently manage and tailor your approach.
What would be your top three recommendations to retailers to help them improve how they use their data?
The first involves creating a single source of the truth, a platform that is dynamic and enables new data to be quickly and easily ingested. Not just consolidation for the sake of it – but a live platform that enables new assets to be brought in and effectively utilised.
Second would be a shift in how organisations use their people – bringing in that cultural change and awareness of what tools are available. In some ways advancements in technology make it easier to use and exploit AI, but there’s much more businesses can do to be open minded and find ways to explore that potential. This cuts across a number of things whether it’s training and reskilling for competence or looking at the use of new tools – breaking down barriers to the adoption of new ways of working.
Thirdly, I recommend focusing on building or acquiring effective and efficient platforms for decision making. In a world where cloud enables us to consolidate data and we have a lot of fantastic tools at our disposal, we need to have the right types of decision forums so that we can unlock the benefits quickly. As we’ve moved online dynamics have changed – and we’re still working on how best to use some of those different decision-making platforms. This affects and informs all kinds of decisions – from training and upskilling and ideating for a new project down to run of the mill budgeting. Retailers need to consider a model that really works to take advantage of differences of opinion, but at the same time find conformity in decisions.
How do you predict data will be used in the future to strengthen a retailer’s market position?
I think critically we’ll see data analysis used to achieve that deeper understanding of the customer – better anticipation of consumer demand, what they want, how they want to shop. That information shapes all aspects of a retailers’ propositions, whether it’s price, products, placement.
Now more than ever, it’s key that customers feel they are getting goods in a convenient (and safe) manner. AI will undoubtedly help as we move away from ‘one size fits all’ mentality. Instead of all stores of a same size having the same core products, we’ll increasingly see stores which better serve their local catchments. We’ll see online propositions that feel more immersive than the catalogue formats of old. Ultimately this is good from a consumer perspective, but will require ongoing innovation from retailers.
One positive for businesses is that as technology develops and evolves, it naturally becomes more accessible financially. This means that technology costs are no longer a barrier preventing businesses from becoming data driven. Now there are innovative solutions available to a wider range of retailers which don’t require significant investment upfront.