Analysis: How Marks & Spencer’s virtual assistant drives sales

From a “game changer” of a strategic alliance with Microsoft to a deal with tech education specialist Decoded to create “the world’s first retail data academy”, the retailer is making bold moves to shake up how it operates.

In the context of these partnerships, the quiet and methodical development of the chain’s virtual assistant is relatively small – but it demonstrates how M&S is using technology on the front lines to change how it communicates with customers.

The chatbot was designed to reduce the volume of calls to its customer service centre, which digital operations manager Sophie Bishop says is to “ensure the advisers on the phone are dealing with someone who really needs a human”.

Marks & Spencer’s virtual assistant therefore only presents itself when there are indications that a shopper is struggling to navigate the retailer’s website in some way.

It was initially launched to handle two customer queries – those who wanted to track their order and those who needed more information about returns and refunds.

Bishop says: “So far it is doing a fantastic job in the places we’ve installed it and we are learning every day from the conversations that are going through it.

“We are adding content weekly depending on where we see the need. It’s an agile approach so we might put a journey live for a week, take it down and see what we can learn from it.”

Additional functionality

Today, the virtual assistant is used across 11 different customer queries, with two of its latest additions – helping with access to rewards and giving order details – being amongst the most popular amongst online shoppers.

Bishop says that the number of shoppers that have used these functions is “absolutely massive”.

However, Bishop is keen to stress that the virtual assistant is not viewed as a replacement for its customer contact centre, and that not every type of customer query is appropriate for a chatbot to deal with.

“Anything that’s a bit emotional, like order queries especially around peak time, isn’t really right for our chatbot – we’ve all experienced the frustration when a delivery doesn’t turn up,” Bishop explains.

“We know it works for certain instances – the trick is making sure the virtual assistant handles the [queries] it can handle and gets the customers it can’t help speaking to a person as soon as possible, so they’re not frustrated.”

The chatbot has measures in place to avoid this situation. The use of certain trigger words or two consecutive requests to speak to a person and the chatbot will automatically put a customer through the retailer’s customer contact centre.

Nevertheless, the queries Marks & Spencer’s chatbot can answer have made a tangible impact.

Since its launch, approximately 70% of customers who have interacted with the retailer’s virtual assistant have had their issue resolved without needing to speak to a customer service adviser.

M&S also estimates that, to date, its chatbot has driven around £2m in revenue by shoppers that have interacted with it and then completed a purchase.

While Bishop would not be drawn on how Marks & Spencer plans to roll out its chatbot across its other channels such as its mobile app, she had no doubt it would play a bigger role in the retailer’s ecommerce platform going forward.

“It is quite an agile rollout but we’ll see what we can do because we know we are increasing revenue with [the chatbot],” she says.

“It’s all part of that digital first retail and this is a big step for us. We know this is working so how we make its presence bigger is something we are always evolving. It’s all a part of our bigger mission around digital transformation and how we become a digital-first retailer - this is definitely a part of that.”