In 2017, the biggest gap in the chatbot market will continue to be perfecting consumer experience. Growth of the overall chatbot sector is entirely dependent upon a consumer's desire to engage with these services. A good chatbot adds value to the consumer's life while only showing up when the consumer wants it to.
In 2016, major brands powered millions of chatbot sessions. Looking back, it's clear the real opportunity is delivering the right message to the right person at the right time, otherwise called "intelligent messaging."
Intelligent messaging is context-aware messages provided to users based on a smart combination of valuable information, including:
1. Information brands have. This is either information the brand has about its customer, or data about its products and services. Customer data includes CRM customer files email addresses, purchase history, product preferences and phone numbers. Product and service data includes product SKUs and descriptions, store hours and pricing.
2. Information platforms have. This is any information that the platform on which a bot lives knows about a user. For Facebook Messenger, this is a particularly robust data set (name, locale, likes, etc.).
3. Information users provide. During the customer's onboarding experience, messaging apps and services receive and store a number of properties on a user -- a perk that's unique to mobile messaging.
Intelligent messaging is a simple concept that stands to be the future of messaging for both brands and consumers.
Below are a few examples of intelligent messaging at work. (Note: Snaps worked with the Nike team to help create the Jordan Brand Breakfast Club Workout bot.)
Jordan Brand Breakfast Club Workout
When rolling out its Breakfast Club platform, Jordan Brand had a goal of reaching elite high school athletes with a rolling admission workout, leveraging content it created. The brand decided to launch this experience in Facebook Messenger so it could deliver the workout to users without having to create a new app and force a download, and because it believed a conversational interface would be the right user experience for its users.
Because this workout IS a rolling admission workout, on any given day one user might be on day one, and another on day 30. This custom data model leverages users' local time and creates single-user cohorts independent of the calendar date (simply a user's progress through the 30-day workout).
Users are sent a primer message at 6:23 p.m. local user time every night, and then a wakeup message at 6:23 every morning there's a workout scheduled. Throughout the workout, users are prompted to provide feedback on how the workout is going, and if they're ready for the next batch of exercises.
Platform data: Locale / time zone
Brand data: NikeID Login for name matching
In-bot data: Registration day, workout feedback
American Express is an early adopter in chatbots, currently running a pilot for its customers in Facebook Messenger. The AmEx chatbot sends messages about each card purchase and also sends relevant messages about card benefits. Users are asked to login to their American Express page with their user ID and password.
American Express knows the user has booked an international flight leaving LGA (NYC) on December 24th. A day before the trip, the chatbot sends a personalized, informative note to inform the user: a) when using an AmEx card abroad, there are actually no international transaction fees; b) that there is actually a private AmEx lounge at the user's departing airport (with an accompanying airport map option).
Platform data: Locale / time zone
Brand data: Amex ID login for name matching, member status, purchase history, travel destinations and travel dates
In-bot / custom data: None
Numbers back the claim: brands using intelligent messaging have a 70% response rate. Compared to other communication forms (email, social posts, etc.) intelligent messages provide a massive opportunity to drive engagement and conversions.
What's more, bot users can be acquired for significantly less money on Messenger than through native app installs, without losing any value.
Bots continued success paints a not-so-distant future in which consumers can have fewer apps but recieve even more value from their mobile services, all while using platforms they're already on. Bots can offer new product releases based on previous preferences, a shopping assistant timed to walking into a store, or a follow-up for feedback after checking out of a hotel.
