Lithium CEO Pete Hess on helping brands keep user data safe and why it doesn't build bots

Last month, San Francisco-based analytics firm Lithium Technologies opened its new office in Bengaluru. The company, named as the next big thing in the social media customer relationship management segment by The Wall Street Journal, also said it would double its workforce in the country to 200 in the next year.

President and CEO Pete Hess, who was in Bengaluru to inaugurate its campus, spoke to TechCircle about how analytics is helping in automated conversations by integrating with third-party bots and how Lithium's products can help brands protect their customer data in the age of strict privacy laws. Hess also talked about how Lithium and social media marketing firm Spredfast, which it acquired last month, will create a large customer engagement platform. Edited excerpts:

Could you talk about your product suite with the recent acquisition of Spredfast?

We have a product where brands can communicate with their customers directly. Then there is our community platform, where customers can pose questions and other customers will answer those questions, which is cost efficient for the brands. There is the community play and if the end customer is going to social media for help, we help them do that through these channels. 

We are building an asynchronous messaging platform for the brands on social media platforms. Instant messaging is going to replace SMS, emails and calls eventually.

When customers tweet, they might be doing that with the brand directly or they may be complaining about the brand and the software listens and we get the feeds from Facebook, Twitter etc. The software creates a case for an agent using 
our software to respond to help that customer or to know what the customer has said. It is a listening software at one level but also provides for support. So, it is like digital customer care. 

Customers are increasingly using social media to get help from brands. Spredfast, which does organic marketing through social media, will help us in providing a comprehensive suite of digital customer engagement.

What is the role of the Indian office?

We have several customers in India, some of whom are multinationals and hence the contract could also be global. We do engineering and product support also from India. 

We have teams that do design work, core engineering, back office, professional services, quality assurance, some entire products are done by our Bengaluru centre. 

The objective is to mirror all of the functions here as well and ultimately sales too, which will help us understand how we market our products here. It helps our local teams to have local customers, who they can engage with, which will help them to better design the product, so that they don't need to work the strange hours of our US office timings to understand customers. 

If we get more customers here, they will be less dependent on the US teams (for customer feedback on products).

How do analytics and machine learning help you in automating some of these digital customer services?

The product helps in automating digital customer care through the response channel or what we call as brand direct, which is a more efficient way of talking to customers than over a customer call centre or emails. The first level of conversation with customers is often taken care of by bots before directing the issue to a human agent. 

We don't build the bots, but we integrate our tool with the bots. Brands often want a personalised approach and the idea is not to replace humans with bots as of now. 

Many brands are integrating our software for Facebook Messenger or iMessage or other channels to communicate with the digital agent. This is a private asynchronous messaging tool, which often knows the customer well. 

With Spredfast, brands can take that information and derive business insights for customers or groups of customers and effectively market your products.

But for larger brands, the level of engagement could be high for humans to take care of all that conversations, right?

A lot of them do. A subset of responses is automated. Sometimes a human agent will be employing to bot to make sure the interaction with end customer is right. 

For instance, T-Mobile in the US does that. Because sometimes the issues are out of the purview of the bots. Sometimes a spike could be because of a new product launch, most of which can be automated. The community centre, which we help the brands run, also help their customers to get most of their questions answered.

You mentioned that you don't build bots. Why do you stay away from creating your own bots?

Eventually, we will want to develop as a standardised configurable bot. Right now, it is not a scalable business and is a proprietary product and we would not be able to our customers like Google or Microsoft. 
Many brands have their own bots already and many bots are domain/industry-specific. Companies already have a lot of domain knowledge and we need to let the brands leverage our platform to work on top their expertise.

Are you overly reliant on social media outlets like Facebook and Twitter?

That is where the traffic originates from. We want to help the brands integrate with all their social channels. But a lot of customers are using brands' own app to communicate. 

There are community forums, too, for dedicated fans and our product will help brands to keep the conversation in their own territory. It helps them monitor and ability to moderate the conversations and address the concerns. It is a great way to build loyalty among the consumer base.

The asynchronous messaging platform that we built can work inside brands' own apps. Especially when the strict privacy laws are being implemented worldwide like the European Union's GDPR, brands can control the customer data better if it happens on their own channels like community forums or own apps.