Lockdown Lifts Usher in a Big Data Experiment for Retailers
An article by Natalie Cramp, CEO of Profusion, first featured on Customer Magazine Experience: https://cxm.co.uk/lockdown-lifts-usher-in-a-big-data-experiment-for-retailers/
The phrase ‘unprecedented times’ has been used so much in reference to the Coronavirus that it can seem like an empty cliché.
We have after all been presented with new extraordinary circumstances on a daily basis for months on end. Now, with lockdowns easing and the retail industry opening back up, we can be forgiven for thinking the unprecedented times are now at an end. However, this could not be further from the truth.
All of us, and particularly retailers, are entering uncharted waters. Many businesses are moving from the relative clarity of being fully shut or online only to the more complex position of trading under restrictive and unique circumstances. Navigating this new normal is going to be very difficult but it can be made easier by collecting and using data to make more informed decisions.
The biggest question facing retailers is how will consumers react to the easing of lockdown. We saw in the first week of non-essential shops opening up queues around the corner for some outlets.
Undoubtedly, as the weeks go on we will get plenty of news reports indicating whether there has been a complete bounce back, a steady increase, an initial glut followed by a sharp drop off or a worryingly flat response. The reality is that these industry wide stats will take months to give us even a general picture of what is happening. For individual retailers, this will be too late to react and adapt.
Retailers need to take control of the situation by having a proactive strategy underpinned by data. The exact approach will vary markedly between businesses, however, for SMEs right up to large chains, there are overarching principles that I believe will help frame their response:
Identify the data you have
Website analytics, sales data, marketing and social media engagement stats etc. are available to nearly every business and will provide the bedrock for understanding how conditions have changed and (most importantly) could change in the future. If you don’t currently collect this data, it is absolutely critical to set up processes that will store it in a manner in which it can be easily analysed.
Identify the data you don’t have
Often the biggest knowledge gap retailers have is in store. Beyond simple transaction data, retailers are blind to footfall, customer journeys around the store, connecting in store customers with online identities, how and why sales are abandoned and the general sentiment of customers. Add to this incomplete information on online customers and many businesses can be surprised by how little they actually know about their customer base.
This data dearth can manifest itself in low marketing engagement stats, inaccurate predictive modelling for sales and, most visibly, an inability to answer simple questions – for example, ‘which of my customers are vulnerable?’. Recognising what you don’t know is an essential step towards getting the data you need to know.
Define the questions you want to answer
There are dozens of metrics that can be collected and fed into data science algorithms, however, more data isn’t necessarily better. Not all data points are created equal, nor is it practical or wise to attempt to collect everything you can – it is a pointless and time-consuming exercise.
The best approach is to first think about the questions you want and need answers to now. For example, how does in store footfall compare to pre-lockdown? How many customers are not making purchases due to queues because of restrictions? Are the in store customers new or online customers that prefer to now shop in physical locations? From your list of questions, you’ll be able to work with your data specialists to determine the exact data points you need and therefore what new information you need to collect.
Create a plan to collect the data
This is where businesses can often get bogged down in complexity. Dreaming up cunning marketing or engagement strategies to get customers to part from their data may look good on paper, but when time is of the essence and money is tight they are far from practical.
Companies often approach data collection with a near-fatal misconception that customers need to be duped into providing their data. The reality is that many people are happy to provide information directly if they approached in a transparent manner and can see the clear benefit of doing so.
Asking customers questions via in store surveys or online polls is often the most effective strategy. With conditions as they are, most will understand why you would need to know their thoughts and feelings.
For information that can’t be easily obtained by direct questioning or observations from your in store staff (e.g. on footfall), technology can provide assistance. One example is using fintech software that captures and collates card data from purchases, helping you to link in store and online identities.
Analyse the data – quickly
The depth to which a business analyses data is of course dependent on their resources. There will also be large variations in the approach that will work best – updating existing models, creating new algorithms, looking at how data is integrated into the existing enterprise software and Business Intelligence platforms etc..
However, with the retail environment likely to be highly volatile, the key will be to ensure that the insights gained are updated as regularly as possible – ideally in real time. It is also worth noting that getting assistance from data scientists isn’t as expensive or daunting as it sounds.
Often they can work with a business to quickly set up the models that are needed and, with the right data architecture in place, they can, with a little training, be run by any member of staff to continually provide insights.
Ensure your teams understand it
This is worryingly so often missed. There is little point in having a plethora of data or indeed any insight if it cannot be acted upon. This should not be the purview of a data science team but something that every team member across the business is confident to do. They need to understand the insight relevant to their role, its limitations and be able to take quick action – be that switching off a customer marketing campaign, or influencing decisions on ordering from your supply chain.
Test, learn and adapt
Exactly how to ‘act’ will not necessarily be clear – we are after all in uncharted water – however a scientific approach can help to mitigate risk and light the way forward. This is a perfect time to test strategies and retailers need to be bold in doing so, and adapting quickly from the data they get. If, for example, you’re presented with data that in store sales conversions are very low, you can test different solutions to identify which is the most effective. A good data scientist will be able to assist you by stripping out other variables to tell you exactly which factors are moving the numbers.
I know for a lot of businesses data collection and analysis will be far from top of their list of problems to tackle in a post-lockdown world. Nevertheless, I can’t stress enough how important it is to make business decisions based on facts rather than emotion or ‘gut feeling’. Any data that helps to illuminate these unprecedented times may be enough to give your business the edge that helps it not only to survive, but also to thrive.