News Stay informed about the latest enterprise technology news and product updates.

Automat powers conversational marketing bot for L'Oreal

The new L'Oreal chatbot helps customers find the perfect gift -- and requires an enterprise-grade marketing stack to do so.

NEW YORK -- Chatbots may be a natural fit in a customer service department, but can they upend traditional -- some...

would say stagnant -- digital marketing efforts? Andy Mauro, CEO and co-founder of the Montreal-based Automat Technologies Inc., is betting on it.

Automat is the core technology behind L'Oreal's conversational marketing bot, called Beauty Gifter, which launched in May. The bot helps L'Oreal customers ascertain and figure out which products make the perfect gift, a beauty-products buying scenario that Mauro called "a legitimate customer problem."

To make recommendations to customers, Automat swaps out cookies and clicks for a highly personalized interaction. The bot uses artificial intelligence to essentially ask: What do you like?

Digital marketing is ripe for disruption

According to Mauro, digital marketing techniques have stalled. The most common practices include cookies and tracking pixels to follow customers around the web, internet search to understand behavior and email marketing campaigns that whisk customers away to digital properties -- often with little success.

"This is a crazy way to talk to your customers," Mauro said during his talk at the Strata Data Conference. "It builds no brand loyalty. We all hate it."

L'Oreal's Beauty Gifter aims to modernize digital marketing for a clientele increasingly accustomed to talking to their phones via digital assistants, such as Siri and Alexa. The technology relies on natural language understanding to converse with customers and help match the right beauty product gifts to the right person. But customers don't interact with the bot on L'Oreal's site; instead, they do so through Facebook Messenger, a channel that now boasts 1.2 billion users.

The conversation starts with the bot asking the gift-giver a few questions, including the gift-giver's budget. The gift-giver then shares the bot via Messenger with the gift recipient, who is prompted to answer a few quiz questions ala BuzzFeed -- pick one: bubble bath or breakfast in bed?

The bot collects the responses, analyzes them and then provides a few price-appropriate gift recommendations to gift-givers, who can purchase and ship the gift right in Messenger.

It's still early days, but Mauro said he's been pleased with the results he's seen so far, including an 82% net promoter score. Even better, the technology starts up a conversation that, if all goes well, continues to pay dividends. "The re-engagement loop is really where the magic happens," Mauro said.

It's a data, data, data, data world

Mauro admitted conversational marketing isn't for every industry. Insurance, for example, may not be the best fit. But opportunities exist for companies in retail, consumer packaged goods, fashion and luxury, he said.

And competitors and venture capitalists are also taking notice. In the last couple of weeks, HubSpot, a marketing and software company, acquired, a small chatbot platform company. And Drift, a conversational marketing startup, announced a $32 million round of funding that will go toward building a presence on the West Coast.

There are types of data you can learn in conversations with someone that you cannot get through a search, through a like or through the traditional digital monitoring means.
Andy MauroCEO and co-founder, Automat Technologies Inc.

CIOs interested in helping their marketing departments pursue a conversational marketing strategy will need to start with the customer data they already have. Mauro said one of the mistakes companies make with chatbots -- especially in the early days -- is to go narrow and shallow. "I think the point is you want to go narrow, but relatively deep on a specific problem and do full integrations," he said.

Companies will need to build out an enterprise-grade marketing stack with purpose-built artificial intelligence technology related to marketing -- and not something companies can procure from a chatbot builder alone. "The [customer relationship management] integration, card-to-payment integration, inventory integration planning -- we had to start that early," Mauro said.

Plus, if the conversational marketing bot is going to be successful, customers have to be able to find it. Marketing practices, such as paid advertising campaigns and other optimization practices, as well as getting featured on something like a Facebook Discover -- Facebook's chatbot marketplace -- and, yes, even stodgy email marketing campaigns still apply to conversational marketing.

The reward for companies that successfully digitize and automate conversational marketing? That would be more -- and richer -- customer data. "There are types of data you can learn in conversations with someone that you cannot get through a search, through a like or through the traditional digital monitoring means," Mauro said.

Next Steps

Facebook's chatbot platform will change mobile computing as you know it

Data is key to unlocking the potential of chatbots

A look at IT's biggest chatbot development hurdles

Whether chatbots or humans have an edge on social selling

Dig Deeper on Enterprise artificial intelligence (AI)