Virtual assistants for enhanced experiences

Estimated reading time: 5 minutes

Yellow.ai‘s co-founder and Chief Product Officer Rashid Khan proposes to help enterprises automate their ‘conversations’ across two major fronts – consumers and organisations’ employees.

He opined that too much money is being spent, but consumers are still unhappy because the customer experience is still not what is it supposed to be.

Three observations have since emerged.

Another, is that every customer call is manually filled out as a ticket for action to be taken. “More and more CIOs are investing in automation, but that is not being reflected as happiness for consumers.”

The first is that phones and emails used to be the only mode of communications. “This has changed and we are active on other platforms now,” Rashid said.

Another, is that every customer call is manually filled out as a ticket for action to be taken. “More and more CIOs are investing in automation, but that is not being reflected as happiness for consumers.”

The final observation is about how data are located in different siloes.

Customer queries are seldom being solved and Rashid pointed out after a certain time on phone, it is not uncommon for a marketing push to happen ie. an automated voice message, or even an SMS that tries to upsell another product or service.

Rashid said, “So, we are just solving three major problems for enterprises.”

Siloed data

The third scenario stems from data being very siloed in the enterprise. When a customer calls the contact centre with feedback or queries, the data just goes to the customer support team.”I think at a global level, or an organisational level, there is no single view of the customer today in the enterprise.

“So, when marketing teams try to run campaigns they don’t have that view to help them more send relevant marketing messages, or to ascertain whether customers are happy or not.

“Segmentation studies of customers are usually based upon geographies, or languages they speak, or past purchases.”

Rashid shared, “Most enterprises try to do some personalisation/customisation when they are trying to reach out to you. But they are not able to do this because they don’t have a lot of your personalised data in the first place.”

How do we reach this state of silo-less data? That’s where solutions like a customer data platform comes in, and it refines data across multiple functions like CRM, ERP, customer support tickets, and so on.

“That said, most enterprises spend immense amounts of money and resources to try ensure how interaction with each customer can be more personalised.

Removing the siloes

How do we reach this state of silo-less data? That’s where solutions like a customer data platform comes in, and it refines data across multiple functions like CRM, ERP, customer support tickets, and so on.

This starts the important process of unifying all the data.

This is the first part of doing this, and a lot of insights starts to emerge when customers visit a website, or when they interact with a virtual assistant, or when they call a contact centre.

“The whole experience becomes customised for you,” Rashid said, giving the example of contact centre agents constructively giving updates and recommendations without asking for a lot of information, when a customer calls in to purchase something.

A customer data platform (CDP) is something that Yellow.ai has just rolled out in tandem with their virtual assistant solution.

Evolution of virtual assistants

Rashid shared about how virtual assistants (VAs) have evolved, based on three different aspects; channels of interaction, typical use cases, and industry types.VAs have broken barriers in terms of what channels they can be deployed on.

“We ourselves are a platform deployed on more than 50 plus channels, including all the major ones like WhatsApp, Instagram, mobile apps, and even on telephony systems.

“Voice automation will play a big role in this whole automation piece moving forward, and Yellow.ai has built out integrations with existing telephony systems.

Initially used largely for sales and marketing use cases like lead collection, lead verification, and lead qualification, Rashid added that recently there has been an uptick in VAs deployed for other use cases.

“And the experience which we are envisioning is that you call up a contact center, customer care support line, and you essentially reach a bot first.

“It tries to do self-resolution after which it will transfer to a human agent with complete context of the conversation,” Rashid said.

Initially used largely for sales and marketing use cases like lead collection, lead verification, and lead qualification, Rashid added that recently there has been an uptick in VAs deployed for other use cases.

“So, the largest organisations are also leveraging this technology to enable their employees’ productivity.

Industries like FSI and utilities adopt VAs to help them push margins up and offer customer support.Yet other industries like fast moving consumer good (FMCG) also adopt VA to create a much more natural way for customers to discover experiences.

How are employees are using VA

Rashid recounted about one of the largest oil companies, as their customer, which had built out multiple excellence centres.

“Here employees can raise support requests and especially for repetitive queries about leave policies or IT support or work travel. So, we have built out multiple bots or virtual assistants which integrate with their downstream HR and IT applications.

“If you have a query about leave policy in Cambodia, you can query it in natural language, the virtual assistant will figure out which country you are asking from, it goes and refers from the right knowledge base for the answer and saves a lot of time for HR, who would have done this manually otherwise.

“Employees have access to this kind of support for 24/7. For most of them, the lead times for getting response from HR is anywhere between 24 to 48 hours. With a virtual assistant technology, the response can be immediate.

The same applies to transaction type of requests, say for a new laptop. Employees can put forward their requests in productivity apps like Slack or Microsoft Teams, and a lot of complexity with application forms, is taken away as a result.

VA adoption Drivers

Rashid said, “Based on our analysis, we realised that any organisation which has more than 2000 employees will benefit from deploying virtual systems for their employees and obviously as the number of employees grows, the impact of the virtual assistant increases exponentially.

Moving forward, Yellow.ai is working towards putting together guidelines for better virtual assistants today.

“A lot of research we do is also largely going in the direction of humanising those conversations,” he concluded.