Over the last year, my company, Snaps, powered chatbots and voice skills for leading brands like Nike, VICE, Jameson, Marriott Rewards, Simon, Gatorade and more. We also just launched a new “chat story” in partnership with the Ad Council. Inspired by the insight that teens increasingly seek out interactive stories and “chatbots” online, the “Ultimate Party Foul” chat experience leverages this emerging format to seamlessly integrate a message about the consequences of underage drinking and driving.
With chatbots, brands are able to achieve more than 80% week-over-week open rates of messages they send, an order-of-magnitude higher than their traditional email marketing campaigns. Some of our partners have also seen four times higher conversions and 12.5x higher click-through rates using chatbots.
If implemented correctly, chatbots can deliver significant value for marketers and consumers alike. Here are three actionable learnings to consider when launching a best-in-class chatbot that contributes to bottom line – or effectively shares a social good message.
Go Beyond Mobile Devices
The storyline in the Ad Council and Snaps chat story highlights the range of potential consequences of drinking and driving underage, and plays on teens’ fear of losing the freedom a driver’s license represents. So far the chat story has had over 73,000 interactions and users are averaging nearly two and half minutes on the experience.
In summary: Use your bot to drive people to do something – visit a store, do a workout, or in this case…don’t drink and drive!
Personalization Drives Outcomes – What Do You Know, and What Can You Learn?
The chatbots that consistently delivered at least 80% weekly engagement utilized three different data points on the user to hone in on a personal message. For example, Bud Light launched the “NFL MyTeamCan” chatbot experience with the goal of driving demand and purchase of Bud Light’s team cans on game day throughout the NFL season. The Bud Light chatbot acted as an assistant that reminded fans that it was almost game time, and invited them to order Bud Light before the game. Bud Light saw a week-over-week 83% engagement rate with personalization that layered in elements like the fan’s first name, location, favorite team and when that team was playing. This enabled Bud Light to provide a customized chat experience with millions of unique combinations for conversations – and deliver customized team cans within an hour.
When building a chatbot, brands must think about all the different ways to utilize first party APIs (team schedules, delivery partners by zip code), third party APIs (location services) and user attributes (name, favorite team) to create a truly personalized experience for each user.
In summary: If you know anything about a user, leverage it to be more personal. If you know nothing, have a chat that learns something quickly!
Get to the Point Quickly
Unlike social networks, “time spent” or “average session length” are not always the best metric to determine success, or grow a highly engaged audience. When building a chatbot, learn as much as possible up front, extract intent and deliver results.
For example, a leading athletic brand set out to inspire a sneaker style for women across the globe. The brand launched a customized sneaker builder where the user uploads of a photo of her outfit, and magically (with AI image recognition technology), the chatbot delivers a totally unique pair of pair of shoes that matches the uploaded outfit. This personalized two-click experience drove a CTR 12.5X higher than the global brand average. After A/B testing this chatbot across 250,000 users, when the core focus (click to purchase) was beyond five clicks, the brand witnessed significant dropoff.
Chatbots are an emerging platform with huge potential for driving consumers to take action, whether within the messaging channel itself or in the real world.