Conversational AI key to South American cosmetics giant Belcorp’s sales strategy

Gary Flood Profile picture for user gflood March 4, 2024
Summary:
Belcorp has made a major commitment to conversational AI and is using Kore.ai’s ‘Jessica’ virtual WhatsApp field sales force assistant to help its customer-contact workforce

An image of someone applying make up to someone’s face while they are lying down
(Image by Engin Akyurt from Pixabay)

South American multi-national cosmetics firm Belcorp says that the early use of conversational AI (Artificial Intelligence) has proven to be a good foundation for much more extensive investment in the technology.

Specifically, the firm says the use of natural language intelligent assistants makes it much easier to provide support for the 800,000 independent consultants in 13 of its core South American markets. (The 14th, Portuguese-speaking Brazil, is not currently included.)

This independent (and overwhelmingly women-led) entrepreneur cohort working on a Multi-Level Marketing (MLM) basis forms Belcorp’s front-line sales and customer-contact workforce.

Centered on a Spanish-language chatbot running over WhatsApp called ‘Jessica,’ agents can use the tool to get training on new Belcorp products, as well as the latest information on popular offers for products such as L'Bel, Ésika and Cyzone.

But beyond being able to access useful company information, they can also run their own accounts and sales records.

Last year, Jessica was connected with in over 2 million sessions. It is also currently offering a 75% containment rate. That means queries that otherwise would have meant a call to the company’s contact center, about orders or commissions, don’t need to be made - saving significant Belcorp corporate support time and cost.

So successful has Jessica been that the firm has now expanded its use of AI to what it calls its new AI Innovations Lab Platform. This is moving the scope beyond helping staff in the field to other business processes, from R&D to eliminating unnecessary expenses, and the improvement of team efficiency.

It’s a journey summed up by its internal head of IT, Venkat Gopalan:

We have several different business models, from e-commerce and B2C marketplaces like Amazon and Mercado Libre, but in Latin America direct selling is very successful.

That’s because beauty is both very social and very personal. We want somebody to tell us how to use a product because of things like it being right for our personal skin tones and shades.

Another part of our purpose is also promoting beauty to achieve personal fulfilment and enabling women to be successful - so Jessica has been really central in helping us achieve that.

That’s because if I know my order status, I can operate more efficiently as a salesperson supporting myself and my family.

And as this is an MLM model and you have your own channel of five to ten people that sell under you need to keep supplied, being just able to find out all you need by scanning a QR code is so much better for you as a consultant.

Belcorp says that Jessica is a lot more than the first generation of chatbots - even having enough of a supplied ‘personality’ to become a true intelligent virtual assistant.

For example, as Spanish in Chile is different from Spanish in Peru and Colombia, Jessica can talk in the specific local flavor of the language the consultants use.

Gopalan - who holds three IT leadership roles at the organization, being simultaneously Chief Digital Chief Technology and Chief Data Officer—says he started evaluating the potential of AI back in lockdown.

A major driver for this was the need to keep the very large Belcorp sales force updated, as the company creates over 250 new products per year and runs campaigns every three weeks.

Gopalan says he decided to use technology from a conversational AI specialist called Kore.ai to do that.

Kore.ai markets both enterprise conversational AI and generative AI-based products, with an emphasis on both employee experience and customer experience.

Intelligent chatbots deployed over multiple channels

Conversational AI works by trying to resolve a user’s query or problem via dialog.

To do so, the system must be able to fully understand what the user is saying and first attempt to answer them from a semantic search over an internal set of Frequently Asked Questions/FAQs and a synonym look-up table.

The system is also able to go back and ask the questioner for more data to refine its help - so, if the query is about a large pizza, but no toppings were mentioned, the software knows it needs to clarify what kind of base and if the customer wants pepperoni.

Brands use this form of AI to create chatbots that can be deployed over multiple channels, including Facebook Messenger, WhatsApp, SMS, and more.

In the Belcorp case, that means users from Chile to Bolivia. Gopalan says:

There are specific ingredients that go in each of our beauty products. So, there is a lot of education and training that we need to give to our consultants throughout the journey.

Now, we’re also giving them the best digital tools to help them run their business successfully, from when they place an order to all the post-order sales aspects that they need to look at - but it's also all about building you as a Belcorp consultant, because there is also a loyalty program we want you to be in.

Expanding AI tooling

Next steps for Jessica on WhatsApp are all about extending the functionality to allow consultants to not just find out product data, but also directly place orders.

That would mean a much better buying experience for the agents but will also mean Belcorp can start tracking how much inventory it has based on where field teams are in their sales cycle, and even prompt them to place proactive orders.

This, says Gopalan, will feed back into the overall Belcorp supply chain so he and his team can immediately know where the demand is and where the matching supply should be.

The firm is also about to make onboarding of new consultants possible just through conversational AI over WhatsApp, he adds.

Beyond that, Gopalan has big plans for his new, wider AI initiatives.

These include ambitions to use AI, machine learning and analytics to reduce product development time by 20%, lab productivity by 60% and cut consumer risk evaluation periods by as much as 80%.

Early wins, he claims, include cost reduction in laboratory tests, external software licences, and overall optimization of workflows.

But the need for Jessica to keep helping Belcorp’s MLM workforce remains paramount, he concludes:

We want heavy lifting to be done by their virtual assistant to help her with anything that she needs - the best companion in helping me throughout my business lifecycle with us.

Loading
A grey colored placeholder image