Dialogic Blog

How Do You Market to a Chatbot?

by Thomas Schroer

Jun 26, 2017 4:40:27 PM

Tesco Intelligent Shopping

At last week’s TechXLR8 in London, I was listening to Paul Wilkinson with Tesco talk about the various initiatives his company is exploring to improve the customer's buying experience through technology and innovation.

His group at Tesco has been experimenting with some novel ways for customers to order and buy the things they need, and then make that experience easier and more automated through the use of Artificial Intelligence (AI). They’ve looked at devices including ordering buttons and bar code scanners to help consumers order the things they need, or reorder when the things they use regularly run out.

They’ve also been leveraging the conversational interface capabilities of Alexa and Google Home as a front end to an ordering and fulfillment app to provide intelligence and context to the buyer’s journey. By adding AI, the application knows what the customer buys and when they need to buy it again. It can also anticipate based on different events and what it has learned in the past by basically applying “if this, then that” type of rules. For example:

  • The customer tends to order milk on Thursday, so if it’s Thursday, then automatically order milk, or
  • A customer's birthday is coming up later this week, so on that day, they should treat themselves to champagne.

But if an intelligent bot is now doing the shopping and making the decisions on what to buy, how do you market to that new kind of decision maker? Patrick did admit that there will be impacts to brand marketing in an environment where a bot pushes the consumer again and again towards purchasing the same products.

I spoke with Seri Mesquida Head of Innovation and New Ventures at Hotelbeds Group. He was at the conference talking about “AI as a Game Changer for the Travel Industry.” The use of AI and bots in his company's application includes interactive suggestions to consumers on travel packages based on a customer’s preferences.  I asked him how do you provide consumers variety and choices of products and services in a situation where those very things are being curated by an intelligent bot that is basing its decision on a customer’s historical buying patterns.

In his case, what they would do would add some selections that are not necessarily perfect matches to the consumer profile, but may be of interest to the consumer nonetheless. Regardless, the intersection of AI and online shopping seems to work. Tim Rogers, founder of +rehabstudio, was sharing several case studies of his company’s work that reinforced that point.

One of the projects he discussed was the Cocktail Coach, a chatbot that helps cocktail enthusiasts create high quality mixed drinks from the comfort of their own home.

Cocktail Coach Intelligent Bot

It was rolled out prior to the Christmas holiday season in 2016 and included the ability to purchase the necessary ingredients for making the drinks. Tim indicated that the conversion rate for this app was a whopping 30%. Most online apps globally only see about a 1.3% - 3.3% conversion rate at most, and those numbers are highly device dependent (games see conversion rates up to 10%).  When I asked him why he thought that this particular app had such a high success rate, he indicated he did not have quantitative analysis for the reasons but attributed the success to the topic it covered and the value of the information provided to the user. I guess there’s a little bit of mixologist in most people, and the help provided on concocting drinks coupled with a high “time in app rate”, and an interactive smart interface makes for a good combination for conversion success.

Tim’s message on the future of company branding provided some insights to Paul Wilkinson’s comment earlier on the impact intelligent shopping bots will have with brands in general. Tim pointed out that a company’s brand is the experience its products and services provide and vice versa. When powered by AI, it becomes “Brand as a Service”. The new definition of brand that will drive valuation for a company becomes a combination of

  • Propriety data or intelligence
  • Branded services built on those capabilities, and a
  • Subscription versus a traditional sales commercial model

The advent of intelligent bots and analytics will have an impact on branding and marketing, and companies that can incorporate "Brand as a Service" concepts may be able to better position their products and services in this new marketplace. What do you think? Email me at Thomas.Schroer@Dialogic.com or send us a tweet at @Dialogic.

Topics: Internet of Things, Communications Application Development, Analytics, Artificial Intelligence