Pete Brown, Manager at Elixirr, examines how analytics will be changing retail in 2019.
The following guest article has been written for InternetRetailing by Pete Brown, Manager at Elixirr. Elixirr works with incumbent organisations and start-ups across retail, financial services and telecommunications, which want support to keep up with and respond to today’s pace of technological change. Elixirr works across the world and has bases in Europe, Africa and the US.
New Year, new approach. Retail has seen its fair share of disruption and if market predictions are anything to go by, this will continue into 2019. However, in the midst of the high-street chaos there was a lot of innovation in retail. Most notably, the overlap of the digital and physical worlds whereby physical retailers were using data and consumer facing technologies to enhance their customers’ shopping experience. Every card swipe, share or online search generates data – the challenge is turning this into insight.
Dealing with data has always been an issue for retailers. There are mountains of unused consumer information that clog up back office systems and never reach the surface. However certain companies, such as Sephora, are leveraging this information to create personalised offerings unique to the individual. This approach is the only option for targeted marketing if you want to make successful sales. Instagram accounted an estimated $8+ billion in advertising last year – mainly due to the success of targeted campaigns. Retailers must treat their customers in the same way, removing themselves from blanket campaigns and tailoring to captivate their customers’ interests.
The in-store use of data has changed dramatically. At the front end there have been developments in everything from static traffic counting, which monitor people flows, to smart electronic shelf labelling, allowing the global deployment of pricing and discounting at the click of a button. Before retailers dive into this they must have a clear goal in mind.
The past two years have seen a rise in “tech for tech’s sake”, whether that is iPads in-store or something for kids to play with while their parents shop. We are starting to see real use cases emerge in areas like smart fitting rooms and mobile payment, two areas in which I expect to see enormous change in the next 1-2 years. AR/VR body scanning is also a big opportunity and companies such as Zugara are working with retailers to develop this technology. When I think of data however, one of the significant areas of opportunity is how we use it to drive more sales.
Concise and accurate reporting is the key to sales. A lot of retailers still take the ‘report everything, assign nothing’ approach to information, particularly sales data. However, more recently the amount of information available to you is infinite. This leads to confusion where sales teams on the ground do not have clear success metrics for measuring their sales performance. If you are incentivising individuals or reporting on more than 4-5 targets, you can be sure that more than half of your retail staff will drop the ball.
One of the benefits of being a new player is that you lack the historical backlog of data and can build your systems in a way that only reports information relevant to your particular teams. This enables you to run more efficiently and to test what metrics really lead to better sales. A good example of this is in the user-rated sharing economy.
Take Uber, it must decide which components lead to higher sales and higher customer retention. If it started to measure drivers on absolutely everything about their journey and weighed it all equally, the business would not be able to drive the right behaviours. It is about balance. Too little data and you don’t understand your customers, too much and you will struggle to create a successful, refined strategy. The trick is to start small, grow slowly and carefully, and constantly test that what you are tracking and measuring is driving the best results.
The growth of clicks-to-bricks has been a central trend in the past five years. 70% of consumer research still begins online before the final purchase occurs in-store. This means it is essential that these two channels work with one another and not independently. Many traditional retail businesses still operate online and physical retailing as separate verticals, with little overlap.
The issue is however, that rather than creating synergies that compliment, companies create competitive P&L issues internally, where they are unhealthily chasing the same customers.
I have also seen organisational structures that separate horizontals such as marketing and communication between the online and physical channels, when in fact they are duplicating the same skillset. As a new entrant, Farfetch has successfully built a store proposition that incorporates the best parts of the online world whilst maintaining the needs of physical retailing e.g. operations, communication, formats etc.
Predictive analytics is another area which has seen huge growth, especially focused around the areas of predicting product demand, managing inventory and detecting fraud. Algorithms that analyse sales information and sales variables have vastly helped retailers to forecast demand.
This in turn leads to more effective replenishment and a better understanding for retailers of the cross-dependency between products – leading to more efficient dark stores and store set-up in the retail environment. It is a big challenge for retailers to incorporate historical datasets as they have often not been stored correctly or the information is corrupted. When these models are built effectively however, they can often require few people to operate and produce much more accurate forecasts.
With inventory management there is a huge amount of capital tied into excess stock across the globe, and sometimes this is unavoidable. A study on Fortune 1000 companies demonstrated that a 5% reduction in inventory generated on average $20m in increased profits. If you do not utilise data you are destined to miss-order, have excess stock and be forced into the profit-reducing world of discounting.
Some retailers have adopted a new approach in the form of designated data labs. My experiences of these have been mixed – some businesses manage to integrate the ideas from them successfully, however most fail to launch them at scale. Walmart collects around 2.5 million gigabytes of data every hour on its customers’ behaviour, preferences and purchases. With this quantity of information, the retailer decided to launch an analytics hub where 200 streams of internal and external data can be modelled. Similarly, P&G has placed this as a central part of its strategy with the CIO stating that the business is “2 or 3 years ahead of anyone else”.
Size and money to invest are the differentiating factors for success in this case. If you have neither of these, you run the risk previously stated of creating separate entities that can never be integrated or used at scale – a blunder of many retailers.
I am looking forward to 2019 and retailing. The retailers that will survive on our high-street will be agile, ready for change and looking to other industries in the future. Our shopping visits will become more personal, tailored and engaging – all things that will enhance the customer and retail experience.