How data and AI can support the eCommerce sector during a pandemic

The COVID-19 pandemic has had a dramatic impact on our lives, with the retail industry being dealt a particularly hard blow. As many outlets have no choice but to move their services online, Forbes reveals how in Summer 2020 (a mere six months into the pandemic), eCommerce had grown at a rate that it ordinarily would have taken four to six years to achieve.[1]

In this blog, our Director of Data Science Finn Wheatley looks at how eCommerce retailers can use data in their decision making, to benefit all aspects of the buying process.

Reducing churn rate

Customer churn can be a major issue for retailers, and data science can help retailers understand how and where they are losing customers within the purchasing process. One important metric is the shopping cart abandonment rate – a high abandonment rate could signal a poor user experience. This can be turned around through implementing a series of marketing actions such as automated emails and customer experience videos – read on to find out how.

For subscription-based digital products, machine learning models can be used to predict whether a customer may churn, or whether a customer is appropriate to target. Such models are usually discriminative classifiers, using deep neural networks, tree-based methods, or logistic regression. Generative models or recurrent neural networks can also be used.[2] Both kinds of models can provide a probabilistic assessment of whether a customer is likely to take an action.

Sentiment analysis

A customer experience strategy that does not integrate sentiment analysis as a core functionality will not capture the overall customer journey. Far too often, users can be put off purchasing from a specific retailer, or a specific product, due to representation on social media platforms and negative reviews.

Text mining techniques can help eCommerce businesses identify and fix product or service related issues, to enhance the overall user experience by analysing the language used on external platforms. Natural language processing techniques can pick out negative or positive sentiment towards the brand, which can help retailers improve their products and services in line with consumer needs.

Inventory management

Good inventory management is key for retailers looking to optimise their customer experience, as they need to be able to store the right goods in the right places and quantities to ensure they are able to meet customer demand efficiently.

Machine learning can play an important role here, as it is able to analyse the data between supply and demand to detect patterns and correlations between purchases. This data in turn is able to optimise the order picking process, through improving picking routes and managing inventory stock efficiently. Furthermore, these algorithms can detect problems within the data itself which can show whether items are genuinely in low demand, or whether there is a problem with the data recorded.

Predictive forecasting

Previous sales history, economic indicators, customer searches and demographic data all provide data sources which can inform predictive forecasting – the ability to predict customer spend and preferences based on past experience. Predictive intelligence technology allows retailers to offer customers the right products even before they realise they need them.

Predictive models are able to analyse historical data to classify customers and certain characteristics related to their purchasing habits. This data modelling allows retailers to customise product suggestions for new customers based on the combination of price and product characteristics that are likely to lead to a purchase, according to the model. Furthermore, retailers can also use these models to create metrics such as customer lifetime value (CLV), or incorporate a marketing mix model to understand how exactly customers should be targeted.

Loyalty cards

Loyalty cards are always popular with repeat customers and rewarding them with deals and discounts not only helps build your relationship, but also allows you to collect their data on a large scale.

Similar to predictive forecasting, loyalty cards provide retailers with the opportunity to track regular customers’ purchases, both in-store and online. This data can then inform targeted advertising and presenting products in a way that will encourage future sales.

Pricing optimisation

Machine learning algorithms can help retailers develop the optimum price point for products at different locations. Analysing a number of parameters such as flexibility of prices, customer location, buying attitudes and competitor pricing leads to a resulting price point that is optimised to benefit all parties.

Upselling and cross-selling

Retailers can track user product purchasing habits by combining customer data and product performance analytics to optimise personalised marketing based on previous experience. For example, if a customer frequently buys a sandwich and a bottle of water separately, it may be advantageous for the retailer to market these products together as a bundle.

Shopping cart abandonment

When all the necessary steps have been taken to ensure a smooth sales funnel runs from start to finish of the buying process, it can be frustrating when shoppers abandon their carts at the final step. Today, the average cart abandonment rate in online retail is 69.57%, which is $18bn lost every year.[3]

There are a number of reasons why customers may abandon their carts – they may be saving the purchase for a later date, they may have found the products they intended to buy elsewhere, or they may have even just forgotten about their cart. However, many of these customers may be persuaded to return to their carts, so retailers shouldn’t have to miss out on a sale:

  1.  “We think you’ll like…” For users that may not have reached the point of entering a mobile number or email address, ad retargeting can be a powerful tactic. Retailers place an ad pixel on their checkout page which can be used to remarket to those users on other platforms such as social media and Google.
  2.  “Did you forget something?” If the user was logged into the website, or entered their email address or phone number at any point during the checkout process before leaving, then there is the opportunity to follow up with an abandonment message. This usually takes the form of an offer or discount code to entice the user to return to the site and complete their order.

Conclusion

Retailers must leverage data to make more informed and powerful data-driven business decisions if they want to stay ahead of their competitors in this increasingly technology-driven landscape. Data-driven decision-making can help retailers to improve and personalise user experience, predict purchases and optimise inventory management, all of which will drive profits.

To learn about how we can help you take full advantage of your data, visit our dedicated retail page.


About the author

Finn Wheatley, Director of Data Science

Finn has over a decade of experience working in lead data science and quantitative roles in both the public and private sectors. Following his undergraduate degree from King’s College London, Finn worked for several years in the hedge fund industry in risk management and portfolio management roles. Subsequent to an MSc in Computer Science from University College London, he joined the civil service and helped to establish the data science team at the Department for Work and Pensions (DWP), delivering innovative analytical projects for senior departmental leaders. Since joining Whitehat Analytics, he has been involved in establishing the data science team at EDF Energy.


References

[1] COVID-19 Accelerated E-Commerce Growth ‘4 to 6 years’, Forbes

[2] A. Y. Ng , M. I. Jordan, ‘On Discriminative vs. Generative classifiers: A comparison of logistic regression and naive Bayes’, NIPS 2001

[3] Tackling shopping cart abandonment with data analytics – Millimetric