Those of us keeping our eyes on the mobile landscape are hearing more and more about ‘chatbots’. But what are they? And how might they fit into the way mobile supports mobile business? The answer to the first questions is simple enough - a chatbot is a text-based service, either located within a messaging app (e.g. Facebook Messenger or Whatsapp) or inside a brand’s own app, that allows a user to interact with a business via natural language. The promise of that approach is that it supports transactions (like ordering a taxi or booking a table) within an environment that the modern consumer spends an awful lot of time in and offers a natural user interface - texting. But what about the reality? We’ll look at 2 examples of Chatbots in action, but before we do that, let’s think about a typical customer service call that you’d have over the phone (which we all thoroughly dislike) vs. the quicker experience a consumer could get using a chatbot inside a messaging app.
On the phone, it would go a little something like this:
You search for the company’s website and track down their customer support number. Then, after calling the number from your phone, you would follow the step by step directions of the automated voice prompt system, often choosing the wrong option and being redirected right back to step 1. The next step is waiting on hold until a human becomes available. After they finally arrive, you answer several security questions to prove that you are, in fact, you. From there you may or may not get your answer, mostly depending on whether you were connected with someone knowledgeable.
Here’s how it would work with the chatbot experience:
1) Open Facebook Messenger and search for the business name
2) Start the conversation by making a request
3) Receive rich media feedback (text + images + hyperlinks + voice) that answers your question.
Of course the chat experience is better, because there are no phone numbers to find, no phone tree to navigate, you don’t wait on hold, you are pre-authenticated, there’s a computer replying, which is infinitely scalable and has a consistent knowledge base. In short, the customer service you’ll receive from a chatbot could very well be superior to that of a typical customer service call.
“Let’s ponder this example from the homepage of Microsoft’s recently-launched Bot Framework. Here’s how they think we’ll be ordering pizzas in the future:”
According to this analysis, it will take over 73 taps of typing to place an order with the pizza-bot - feeling hangry yet?
The same task, carried out on a pizza delivery app, takes a mere 16 taps. Today we’re used to getting what we want in a ‘native mobile’ fashion - with the tap of a button. And we as consumers don’t want to type on mobile if there’s an easier alternative.
Even assuming the chatbot knows that your favorite pizza is a medium pepperoni, the actual experience of typing word or phrase responses takes much more effort than clicking through a well-designed app. When you’re logged in, your preferred payment method is known, favorites are tagged, and the normal conversational conventions are dispensed with (e.g. you can click a picture of a pepperoni pizza vs typing ‘pepperoni, please!”) - things are a lot simpler.
Chatbots are certainly an interesting development, and will likely deliver some useful interactions. However, as with the pizza example, they cannot easily replace much of the ‘shorthand’ that we’ve developed through standard UI conventions over the years with good app design.
I would predict that chatbots augment the mobile app experience but are not going to replace the mobile app experience. There are places where a chatbot can provide great service, but that doesn’t mean they are the next great user interface for everything. The native mobile app is a well tuned platform for rich interactions with consumers, where good design can trump a language interface for intuitiveness and efficiency for most tasks.