How companies are using chatbots for marketing: Part 2
What does it take to be successful with chatbots? Contributor Daniel Faggella continues his series showcasing ways companies are already employing this technology.
In Part 1 of this series, I profiled how two companies — Domino’s Pizza and KLM Royal Dutch Airlines — are using chatbots to boost their brands and smoothe interactions with their customers. This time, we’ll continue by looking at cosmetics retailer Sephora and sharing some of the lessons marketers can learn from these leaders’ experiments.
Company description: The French cosmetics company offers makeup, fragrance, skin care and hair care goods for men and women featuring more than 300 brands and its own label. It has approximately 2,300 stores in 33 countries. Its parent company, LVMH Moët Hennessy Louis Vuitton, reports revenues of 4 billion euros in the first nine months of 2017 in the Perfumes & Cosmetics business group, which Sephora is a part of.
How it’s being used: The Sephora Innovation Lab was set up in 2015 in a converted warehouse near its US headquarters in San Francisco. It seeks to integrate cosmetics expertise and digital innovation to improve in-store experiences, offering applications such as makeup tutorials on mobile devices. As a part of its innovation projects, Sephora worked with chat app company Kik in 2016 to build the beauty brand’s chatbot. The virtual personality offers makeup tutorials, skincare tips, video clips and product reviews and ratings.
You can watch how the chatbot interacts in the video below:
It also launched the Sephora Reservation Assistant (a chatbot that uses natural language processing to interact with clients in identifying store location) and the Sephora Color Match (a chatbot that uses artificial intelligence [AI] to pick a color from an image and find a match from its beauty products) on Facebook Messenger. Here’s a demo of the Color Match chatbot:
It also launched a marketing campaign using the chatbot where teens going to the prom submitted 1,500 questions that were answered by the company’s makeup artist. After the live streaming Q&A, viewers were sent back to the chatbot to learn about other prom-related content.
Value proposition: Sephora reportedly claims an 11 percent increase in booking rates through the Sephora Reservation Assistant. In addition, it reported an increase in in-store sales, claiming an average spend of $50 by clients who have reserved an in-store service via the chatbot. The prom-related project reportedly helped drive ongoing engagement on the chatbot. Kik claims in its case study that it spurred 600,000 interactions.
Key takeaways: Selling your products through chatbots will be more effective with value-added features such as tutorials and tips. According to a Digiday interview with Bridget Dolan, vice president and head of the Sephora Innovation Lab, “Digital is a critical element in retail — however, it is not just for the sake of adding new, cool technology. Our intention is to help our clients.”
While most chatbots are likely not generating more revenue (or savings) than they cost to implement, we suspect that large brands (like those highlighted in this short series) will be the trailblazers in fleshing out chatbot use cases that will become popular with users. We covered retail virtual agents in a previous MarTech Today article, and indeed, we suspect that sector to be exciting in the next 12 to 18 months.
Companies most likely to pull ahead in the chatbot race will probably have the following qualities in common:
- Massive existing budget to develop chatbot applications and hire the requisite talent to build and integrate this technology.
- A very large existing customer base already used to engaging with the company via mobile (Starbucks comes to mind).
- Relatively high-volume, low-ticket offerings. Expensive purchases that require customization and in-depth conversation (buying a car, shopping for a new apartment) may have a harder time gaining chatbot traction because of the lower volume of transactions and the inherent challenges with building chatbots with extremely robust Q&A ability (the simpler the better).