Artificial intelligence (AI) is changing the world. We’ve now got cars that can drive themselves, fridges that can order milk, and chatbots that are capable of holding human-like conversations. And it’s these bots in particular that are having the biggest impact on the way brands communicate and engage with consumers. So what can chatbots really do? What are their limitations? And how can your brand take advantage?
Modern retail is experiencing a role reconfiguration. Traditionally, it was the product that brought customers to a retailer. Not anymore. Now, customer engagement drives the industry, and in an increasingly digital marketplace the essence of customer engagement lies in personalization. Customers want to connect with their favourite brands and services, but brands that offer great service experiences are in the minority. In 2015, a Customer Care Measurement & Consulting study found that 54% of retail customers are dissatisfied. So what innovations should be made use of to engage customers and encourage brand loyalty whilst keeping cost of innovation low? Chatbots can be part of the solution.
Chatbots are artificial intelligence platforms that can respond to text or voice commands and deliver a response. And they’ve been around for ages. From the earliest days of computers, simple bots have allowed users to return pre-set answers to specific commands or questions.
But we’re light years ahead of that now, as chatbots have grown in (artificial) intelligence to a point where they can now lead conversations, suggest related topics, and display certain emotions or feelings. These advances have been key in establishing chatbots as a useful marketing tool for brands looking for increasingly efficient – and effective – ways to engage consumers on social, on mobile and through their own customer service channels.
What bots can do for your brand
Regardless of the size of your business, if you have a customer services department then there’s a bot for you. Chatbots are there to make things easy – for brands and consumers – by automating simple processes and returning answers to questions in real time. In fact, after the avalanche of apps being released over the last few years, messaging is hailed as the new anti-interface of the next phase of the internet.
And bots can also be used throughout your brand’s customer life cycle to generate interest and awareness, enable purchases and nurture brand loyalty. If you are a fashion brand, for instance, you could use a bot to share personalised styling advice, accessory recommendations and reviews with your customers. You could also link directly to your e-commerce site to make it really easy to view and buy the products you reference.
You can even use bots to enhance your loyalty programme by helping your customers select the most attractive redemptions or offers for them (and therefore strengthening their relationship with your brand too). What’s more, if the backend of a chatbot has payment and ordering information stored for a user, ordering is only a few taps away—a great improvement from having to dig out your card and typing in all your card and address details. For example, regular Amazon customers know that one of the reasons why ordering from Amazon is the preferred choice for products available from its online store is because the ordering process is simple, quick and easy. Additionally, with the wide range of goods available on Amazon, the customer can track their order for a new winter coat in the same environment as rating the purchase of a bathroom mat. The same principle applies to chatbots.
And that’s not all. As chatbots are completely automated, they can be a great cost-saver too. Unlike human customer service employees, chatbots can interact with hundreds of consumers at the same time, answering more queries in a shorter amount of time (which leaves your human staff free to work on other things).
When not to use a bot
As with any technology, you should think carefully about when and where it would be right for your brand to use a chatbot. Used at the right times bots can provide valuable support for your customer service teams; but you should be wary about relying too heavily on them.
If your brand provides a particularly complicated product or service, for instance (such as financial or medical services), you might enjoy limited success with bots. Because if your bots fail to understand a highly specific question or query, then your customers could get very frustrated, very quickly, which could harm your brand. We all know how annoying automated phone menu systems are, so you need to make sure your bots provide a better experience than that.
In any event, it’s important to create an experience that requires minimal input from your customers. So think long and hard about the most intuitive and contextually relevant ways you can engage with your customers. These interactions should be simple and quick, and they should help your customers reach their objectives as quickly as possible. And where your bot does not have an answer, make sure you direct your users to the best person who does.
What to think about when setting up your bot
The most successful chatbots provide relevant – and increasingly personalised – responses to consumer queries. And they use data to do this. So to make your bot’s experience feel as tailored as possible, it makes sense to integrate it into your brand’s existing CRMs and backend systems. Your bot can then use this data to filter out irrelevant responses and tailor its answers based on your users’ history of engagements with your brand.
If you are an e-retailer, for instance, you can programme your bot to avoid talking about products you know your customer has already purchased, or even make suggestions for accessories that complement the look that they have purchased in the past.
Of course, as AI becomes increasingly sophisticated, so too will the algorithms powering chatbots. We’ve already reached the point where bots can adopt natural speech patterns, and they are becoming better and better at understanding colloquialisms and slang. But there’s still a long way to go. After all, there are infinite permutations and scenarios a bot may have to respond to, which makes coding these algorithms very complex. And this is where chatbots can still come unstuck. So, to get ahead of the game, you should try to pre-empt what kind of queries and questions your users will ask, and then spend considerable time and effort programming your bot to answer them. It is also worth considering that if a retailer’s level of customer service is a core part of their brand strategy, one way to introduce bots safely and successfully is having bots work alongside live customer service agents. A conversation could start with a bot but it could transition to a human customer service agent if basic information has been gathered and the query becomes more involved. That way, both the customer service agent and the customer can start the conversation at a relevant point, rather than spending time gathering basic information.
H&M and chatbots – The ideal customer experience
Ten years ago, if a customer wanted to buy a dress from retailer H&M, they would browse their local stores or order online. With the introduction of smartphones and apps, the online experience transferred from the personal computer to the smartphone. Still, to find the desired style of dress, the customer would first have to select a gender, category of clothing, and apply various other filtration criteria to be presented with a page of search results that they could finally make a selection from. This year however, H&M moved to the forefront of the future of shopping with the release of its chatbot via the messenger app Kik. Kik had launched its bot shop, based on the principles of the apple store, in 2014 and currently has 275 million monthly active users, 40% of which are in the 13 to 24-year-old demographic—a crucial age group when it comes to creating long term brand loyalty. So while the customer is using the Kik app, they can now simply select the H&M bot and communicate with it either via messaging, for example requesting a ‘blue cotton shirt dress in size 10’. The interaction is much smoother, no installation is required and the customer does not get frustrated on their journey to finding the garment they are looking for. This also means no matter what your customer is up to, when they are ready to make a purchase, they are able to do so in the blink of an eye without having to leave the app (and more importantly, other conversations they might be engaging in).
The future of chatbots
Traditionally, brands and marketers have marketed and advertised to their audience through one-way conversations. Bots encourage consumers to initiate personal two-way conversations, at a level far more efficient and cost-effective for the retailer or service provider than interaction via social media or call centres can be. The enthusiastic adoption of chatbots by major players in the industry, such as Facebook and Google has shown that users are keen to adopt this technology to manage their busy lives without having to leave their messaging app (and potentially interrupting another conversations that they are having simultaneously with another human being). The general trend points towards the fact that users will engage more frequently, deeply and efficiently with third-party services if they are made available within a conversational environment instead of separate apps, thereby turning users into customers.
Tom King, Commercial Director
A modified version of this article first appeared in ADMAP magazine, March 2017