, created by Florian Baur
How Predictive Data Analytics could change the way you shop
20% of products in customer baskets can already be predicted - but how will this affect the retailer?
During the week between visits to the supermarket, we all inevitably run out of something and often do not get around to replacing it until the next main shopping trip. Shopping lists can be stored with online retailers but every item is used up at a different rate so some products are still finished before they are replaced and/or you end up overstocked at home.
In the Internet of Things (IoT) world there is talk of different connected devices which could scan your fridge, identify what is missing and add those products automatically to your list for online or offline shopping – a great idea but who keeps everything in the fridge? With a digital assistant such as Amazon Echo, you can just shout ‘I have run out of beer’ and have more delivered in two hours – again an interesting approach.
Recurring subscription models have been emerging in the last few years but usually focus on one product. For example, the Dollar Shave Club sends just the right amount of razors you need for a month. But what if we could use Data Analytics to predict all the key products you need - for example, a grocery basket for the weekly cooking including fresh products?
Recent emnos data research showed that around 20% of the purchases in shopper baskets can be predicted - this could present an opportunity worth over $80 Billion in the US and €150 Billion in the EU 28 zone. The top ten predictable categories showed in the research as pet food (38%), canned seafood (38%), water (34%), cola drinks (34%), milk (34%), yogurt (32%), juice (31%), wine (30%) and beer (30%).
How could this impact the retailer?
If you join the Dollar Shave Club you would no longer buy razors in the supermarket or be interested in promotions. Take this a step further until more and more shoppers subscribe to a recurring service for several items and there could potentially be a substantial loss in category sales. These products typically sit within the retailers’ core categories and if they are taken out of the equation, shoppers could be de-motivated to visit that store. Recurring purchases also tend to make people more loyal to a brand but, on the other hand, also less sensitive to price, offers and promotions.
e-commerce supermarkets (whether independent or linked to bricks and mortar retailers) are probably the most prepared for predictive data analytics and recurring models. However e-commerce only accounts for a fraction of the overall grocery retail business which means less depth of data and, so far, less powerful prediction capabilities.
On the other hand, traditional retailers have all the data depth to run effective data analytics and predictive models but may not support the approach from a strategic point of view. It is a matter of balancing the prospect of categories disappearing (or shifting to a convenience mission) against the powerful influence on loyalty plus costs savings in offers and promotions.
New business models built on data
Some consumer goods manufacturers have started looking into new approaches – they run their own direct service and fund the shipping with savings from distribution and promotion. This means handling the logistics and working the shopper centric data in an unfamiliar way, but control does make is easy to add in new products for consumers to try.
This scenario could come sooner than you might expect – it is just a question of collecting enough data history to support accurate prediction.