Until we discovered that chatbots had the potential to completely revolutionise how consumers communicate, search, browse and buy with a brand, the emerging technology was lending itself to numerous gimmicky headlines. You might recall seeing stories such as: ‘Turn to a chatbot – your customer service companion’; ‘How people are turning to chatbots to tackle love, depression’; even ‘The chatbot that helped beat USD 4 million in bogus parking tickets.”
It is easy to get carried away with the fancy headlines though. After all, you can sell products and services, make payments, personalise your marketing, check the weather, or even chat with your favourite movie character on the chatbot platform.
But while we are certain that the trend is here to stay, there is a plenty to be done to ensure that the investment in this new-age communication tool is worth your while – and worth every penny. Just this month, Facebook announced changes to its use of Artificial Intelligence (AI) after its bots hit a failure rate of 70 per cent. Many other brands, too, have begun dropping their bots. Frightening isn’t it?
The real challenge is therefore for brands to develop chatbots that are genuinely useful, and are a game changer for their business. What does this look like?
It needs to work
At the most fundamental level, a chatbot needs to work well in order to overcome the stereotype that it is in fact, well, powered by technology. Will it fail to understand your request, or produce useful results – as in the result of Apple’s Siri?
Perhaps because chatbots are still in a very nascent stage, and are being developed by ambitious developers who want to master AI, chatbots tend to be needlessly complicated to engage with. By keeping the interactions simple, providing guidance to first-time users, and having buttons instead of an over-reliance on natural language chat, chatbots will provide real utility.
It needs to solve a real problem in a more effective way than the alternative
The sudden increase in the number of rideshare apps and options in Singapore means that all of us are constantly opening and checking Grab, Uber and the taxi company apps, trying to find the lowest possible fare to our destination. Enter TaxiBot, which compiles publicly available data into an easy to use chatbot, activated through a few taps on Facebook Messenger – saving consumers time and money.
As these apps continue to jostle for brand loyalty in a price-driven market, the breakthrough moment comes when a chatbot tackles the challenge of unnecessary inconvenience. Which is why having one single app that gives us the lowest quotation becomes genuinely transformative.
It needs to delight the user
The important thing to recognise when designing bots is that the user interaction needs to be directed, and the quality must stay consistent.
Singapore’s public transport category, for instance, is one of the most data-inundated categories possible. As a consumer, you can walk over to the bus stop and check an electronic board of arrival timings, or download one of the hundreds of apps available, attempt to figure out your bus stop number, and finally, find when your bus will arrive.
With data being so dispersed and wide-ranging, the Bus Uncle chatbot on Facebook stands out because it allows consumers to use natural language on the front end of the interaction, and get an answer in a few taps. The real utility lies in its ability to help consumers sort through the clutter, while entertaining and delighting the consumer with remarks interspersed with bits of Singlish.
Another great example is Sony Pictures Malaysia, who leveraged the interactiveness of chatbots to promote their movie Magnificent Seven, leading users through an interactive battle against the titular villains from the movie. While most media and entertainment brands were posting video clips, trailers and static brand posts to promote the movie, this experience allowed moviegoers to feel like they were part of the movie – even if just for a few moments.
It needs to be personalised to your needs
The real potential of using a chatbot stems from the ability to receive information only when you need it, and not be inundated with information. A study from Blueshift backs this up, concluding that sending a push notification to a mobile user based on events particular to that user, can lead to a 2,770 per cent increase in engagement.
GOTObot.ai’s first bot, the Currency Converter & Exchange bot, illustrates this perfectly. It allows users to set up alerts for when a currency pair hits a predefined exchange rate, or receive a daily notification at a time of their choosing, while also providing on-demand currency conversion. Just like a personal assistant would, the real potential we see here is the personalised communication it offers.
It needs to trigger repeat usage
Poor retention rates plague chatbots today, driven in part through a way of thinking of them as one off campaigns, and not making an effort to create habit amongst consumers to drive repeat visits.
The North Face did this extremely well when they implemented a chat-based shopping experience through their website that mimics the experience of being in a store – creating a faster, more intuitive conversational approach that improves daily with the rapid development of AI itself.
Apart from these considerations, as with any marketing platform, conversational commerce solutions have to integrate well with existing physical and digital channels, be it at point of sale, on a website, or via email or SMS.
Ultimately, the opportunities for personalisation are endless, and natural language interactions will become the norm in terms of how people expect to interact with technology. If done well, chatbots will become the cornerstone of any brand’s engagement and sales strategy.