2016 has seen talk about bots explode, with Facebook opening Messenger to bots and brands like Sephora making recommendations via messaging app Kik. The reviews have been mixed for current bots. Much of the discussion has revolved around whether bots in their current form are a value add for consumers and the brands launching them. For me, playing with Sephora’s bot on Kik was interesting from an industry perspective, but not something I would regularly use. News app Quartz was also interesting, but not likely something I’ll add to my daily routine. My experience with bots had not majorly impressed me, but I still follow industry news around bots to make sure I am keeping up. One such article mentioned Whole Foods Messenger bot that used emoji to recommend recipes. As an enthusiastic amateur cook and a regular user of Pinterest, this piqued my interest. I don’t regularly shop at Whole Foods, but I am familiar with their reputation and the strides they have made in being a digitally engaged brand. Sounds like the perfect opportunity to give bots another shot.
I do not use Messenger very often, mostly to send my sister goofy videos I see in my News Feed. My engagement with brands on Facebook is also limited, though there are those that I follow. Whole Foods was the first brand I had ever messaged. Opening the Messenger window I am provided some basic information and a prompt to type “Get Started” to begin. The bot sends me a message with more instructions and how to interact with it. The article that sent me to Whole Foods’ Facebook page highlighted the bot’s ability to make recommendations based on emoji, so of course that is where I start. My first request of the bot is recipes with apples. Instantly I am presented with 9 recipes and the option to explore further apple recipes if none of the options presented suit my needs. Selecting “View Recipe” opens the recipes page within the same window, allowing me to take a closer look at the recipe and the required ingredients before quickly and easily returning my chat window. The overall user interface is simple, clean, and easy to use.
So emoji bring back good results, what about text? My next experiment is “quinoa.”Here came a pleasant surprise. Not only was I sent recipes that included quinoa, but the first article that appeared was “How to Cook: Quinoa”. As someone who had to google cooking instructions the first time I made quinoa, I thought that this was a clever value add for users. Cooking comes easily to many, but not every one is Julia Child. Especially with trendy ingredients or those that have very specific instructions, providing this information in line with recipes could be a help to many users.
Suppose you do not have an ingredient in mind to start. Whole Foods still has a solution for you. Typing “menu” into Messenger delivers a prompt to either continue searching for recipes or browse Whole Foods selection. Typing “menu” also acts as a reset for the bot, starting a fresh query. You can browse by type of dish, cuisine, or even special diets. The special diets include common players such as vegetarian and gluten free to things such as “sugar conscious.” This may be helpful for hosting dinner parties with guests whose diets differ from the users. Within each category there are further filters users can choose from to narrow their selection. Depending on your path it may take several selections to receive recipes. Each menu also has a back button that allows users to clear the last selection; helpful if you change your mind about what you want. However, you can only select “Back” from the latest menu, so if you change your mind from Chinese to Mexican, you’re best bet is to start over.
Overall, I was satisfied with my Whole Foods bot encounter. The user experience is smooth and the suggested recipes looked quite delicious. I see limitations in what bots can be used for in the consumer market, but use cases such as this that are on brand and create value for the consumer are an area for growth.
What is your take on bots?