Why retailers should be deploying conversational commerce

Convenience. Personalisation. Customer service. These are three areas of retail which digital technology has drastically changed its meaning.

It used to be convenient to hail a black cab off the street, until Uber came along. Personalisation was something you did to a t-shirt or a keyring, not the homepage of a website. And customer service used to be ‘can I help you ma’am?’ with a smile, now it’s offering multiple delivery solutions and ensuring mobile holds the answer to every possible customer query.

And answering customer questions via email, call centre or face-to-face is just not good enough anymore. Customers now want to speak to retailers in a similar way to communicating with family and friends – via SMS and, increasingly, messaging apps like Whatsapp and Facebook.

Facebook M and chat bots

At the end of March, Facebook Messenger unveiled integration capabilities for businesses. One of its first partners was the airline KLM, which now offers customers the ability to receive confirmation messages, access boarding passes, as well as check-in reminders, flight status updates and most interesting, customer service questions answered from within the Messenger application.

In a big step ahead of its 'buy' buttons, the social network now wants businesses to use the platform – Facebook Messenger – to actively engage with customers.

Behind the scenes, Facebook is using AI (artificial intelligence) chat bots which plug into Messenger to answer consumer questions, meanwhile Facebook M is an intelligent assistant – like Apple’s Siri or Amazon’s Alexa – which intermediates.

Facebook said its bots will be able to answer questions about the weather and traffic automatically, while also providing more useful automated notifications, such as receipts or shipping details, which is where the platform could become powerful for retailers. Meanwhile Facebook benefits because it keeps its users within the platform for longer as they no longer need to hop between different apps.

Facebook M is powered by both humans and artificial intelligence, so that when the AI (or bot) isn’t able to answer a consumer’s question, a person will step in and take over the conversation. Additionally, because artificial intelligence uses machine learning, the bots can learn from the human takeovers in order to handle the question solo next time.

Ramzi Yakob, a consultant at digital transformation agency, TH_NK, said all businesses which interact with customers should be paying attention to Facebook M, mainly because messaging apps have the largest number of users – 900 million on Facebook Messenger alone. 

Indeed, Forrester’s eBusiness analyst, Julie Ask, points out that consumers spend 84% of their time in just five apps each month, and there is a high change that one or two of those are social media or messaging apps.

The design paradigm of the messaging interface is incredibly familiar,” explains Yakob. “All the way back to the first time we started using SMS, so they’re very easy to use for the widest possible number of people.”

He says there is a growing opinion that intermediaries like Facebook M will become the dominant channel for digital transactions in the future. “Not being plugged into intelligent assistants and conversational UIs by 2020 will be akin to not having a website in 2015.”

Intelligent assistants

Intelligent assistants are not new. Apple launched Siri in 2011 as part of iOS 5, but has failed to capitalise on the software in recent years. Meanwhile, Amazon in the US has been throwing advertising dollars at its Alexa personal assistant and Echo speaker hardware, with Alec Baldwin demonstrating how the ecosystem can order an Uber or anything within the Amazon ecosystem using voice commands.

But Yakob says Facebook's bots are suffering from teething problems. “Their limitations are painfully obvious when contrasted with the demonstrations of their more powerful big brothers, the intelligent assistants [eg Amazon Echo],” he says. “The retailers that will benefit the most from these new technologies will be those that have been slow to adopt and invest in digital so far. The old fashioned retailers could actually benefit the most by leapfrogging into this future as the step change for them would be incredibly pronounced vs. modern retailers that are already investing heavily in multichannel strategy and infrastructure.”

Ask agrees that the technology isn’t sophisticated enough as yet, but with machine learning they will only continue to improve. More worryingly is Forrester found that only 25% of enterprises use location data to make their mobile services more relevant, never mind utilising context and insights from that data to potentially engage in intelligent conversations.

Paddy Earnshaw, chief customer officer for Doddle, said conversational commerce could be used to keep customers up to date with information, such as the tracking of parcels.

"Conversational commerce will add another dimension to customer experience when shopping with brands. The opportunity to tailor transactions in the 'now' is immense – the key is nailing this sense of immediacy right through the transaction," he explains. "With this in mind, platforms such as Facebook should be looking at how the fulfilment experience is presented to customers. Providing a delivery option that guarantees first time delivery and avoids needless hassle, must be high on the priority list for customers using this technology."

Forrester also warns retailers may have to lose control and data to Facebook’s walled garden platform. Principal analyst,Thomas Husson, says: “Facebook has not yet shared details on how it plans to monetise Messenger. However, if the past is any indicator, marketers will have to spend money with Facebook’s broader ecosystem to get access to the affinity data Facebook collects. Marketers risk losing direct data and insights if their customers use more of Messenger and less of their own branded apps and sites. Several marketers we interviewed are wondering how such new insights will be used by Facebook moving forward and still fear losing some control over their relationship with customers.”

"Uber-fication" of retail

But Facebook Messenger is not the only conversational commerce platform out there.

Adam Levene, founder and CEO of commerce messaging app Hero, says this new wave of communication is all about convenience.

“Just as they order their food and taxis with a couple of taps from within an app, we’re seeing that people expect effortless personal assistance at their fingertips from the retailers and brands they interact with.”

In fact, Made.com is one of the first retailers to partner with Hero in order to communicate with customers in this manner.

“No longer is waiting on hold, sending emails or even tweeting at a brand enough. Consumers are starting to expect real-time responses via their most popular form of communication – messaging,” explains Levene. “Some may call it the ‘Uber-fication of retail’, but we see this as a huge opportunity for savvy retailers. 99% of requests we see at Hero are related to purchase intent, posing a great opportunity for brands to make the lives of customers easy, and in doing so increase sales, loyalty and frequency of spend.

Hero makes its money through its retail partnerships, keep the service free for customers to use. And 100% of its conversations are answered by humans, with automation only used for personalised greetings at the beginning of conversations or when no one is available to respond.

“We’re big advocates of a human-to-human approach, though we employ bot technology where we can to make the experience better for all. We use automation to make things run smoother operationally, instead of using it to try and replace human expertise and support,” he explains.


A similar proposition is BarChick Concierge. The service began five years ago as an online publisher uncovering the best bars from around the world. But the business discovered friction between group discussions on where to go for a drink (messaging apps) and the time-intensive way people looked for new bars to visit (trawling through apps or the web).

Customers can instead text BarChick and receive a suggestion from a human. And using conversational language, users can book a table at a bar in minutes.

The service is also integrated with group messaging platform, Slack, so friends can talk to BarChick when organising a night out, but Facebook Messenger does not currently provide APIs to replicate this. “Work drinks are a big use case for us and with slack we can be part of that," explains BarChick MD, Tatiana Mercer.

"Bots aren’t quite there yet, as anyone who has interacted with the various bots available on Facebook will probably understand – they’re pretty dumb," she says. "It’d be amazing if they opened the platform to businesses, for example, allowing us to hook our concierge platform powered by human operators."

Its partnership business model with bars and spirit brands sees BarChick make its money from sending time-poor Londoners to bars which have a reduced footfall during the week. Additionally, BarChick can provide valuable data to the businesses.

"Our conversational user experience for the customer allows us to get very detailed information about the user’s request, desired experience, and this information is invaluable to bars to provide great service," she adds.

But unlike Facebook bots and similar to Hero, the BarChick service is human-first. "They are assisted by software we’ve built in-house," she says, noting how the business may look to use artificial intelligence tools in the future.

"For us, AI is really going to be about facilitating operational scalability, but much like Facebook M’s strategy, use humans as ‘trainers’ in the future, constantly monitoring conversations, and providing manual assists or intervening only when necessary. More data, more conversations with users, means lower per-user operational cost once we deploy machine learning technology."

For more information, click below:






What’s Hot on Essential Retail?