‘We were the first in the market to analyse unstructured data’

Ady Meretz, President, Verint Systems, Asia Pacific talks about the applications of speech analytics and how it helps resolve many customer issues for organisations. He also talks about the challenges of doing speech analytics in India.

BW:  How has analytics evolved over the years?

Ady Meretz: When we started, there was focus on the BPO  vertical and the outsourcing business in India. The industry was about call recording and workforce management. Analytics started to evolve in the late 2000's when Information was captured not just for compliance, but also for quality monitoring, and to understand customer behaviour. We were the first in the market to analyse unstructured voice data. When you talk about voice, you do not know what will be in the call. So you need to build tools to analyse unstructured data. We have been active in India for the last 20 years, but analytics started less than 10 years ago and has become very popular in the last five years.

BW: Which verticals do you target? How are your solutions being used?

Ady Meretz: We target verticals like Telecom, BFSI, BPO, government, e-commerce. The top two Indian banks are using our speech analytics solution. Banks are using it for customer care, to understand why customers are calling. They also want to understand customer behaviour for new marketing campaigns.
For telecom, many calls come in for SIM card activations. Some calls are for disputes in billing. When you listen to the customer and understand his pain-point you realise that there could be some processes that need to be fixed.
In e-commerce you look at customisation of services.
Analytics offers actionable insights for back office people to fix their processes.

BW: How do you compare the Indian market with other markets?

Ady Meretz: If I compare India to more mature markets like Australia, India has higher rate of growth due to lower levels of penetratation. Australia will have 70 - 80 percent of our speech analytics but India will have, maybe 10 or 20 percent. US would be 30 - 40 per cent.
We deliver speech analytics all over the world and throughout Asia -- in many languages.

BW:  What are the challenges of doing voice analytics in India, which has so many languages, dialects and accents? How do you get round those challenges?

Ady Meretz: We can do speech analytics for different Indian languages and in different dialects. It is not easy to do and it's even harder when you have a mixed dialect -- like Hinglish (Hindi with English words).
Our language engine is tuned to eight regional languages: Tamil, Kannada, Telegu, Gujarati, Marathi, Bengali, Punjabi, and Oriya. So our solution can be used in BPOs located in those regions, where BPO employees and customers speak these languages.

BW: What kind of insights do you get from the analytics? How does it help?

Ady Meretz: Every organisation wants to reduce the number of calls they receive. CEOs of organisations want to know why people are calling and how these calls can be made shorter, and to the point.
Phone calls are good for surveys too. You can tell a couple of things by listening to a person (which you cannot do in a written survey). For instance, you can tell if a customer is happy or not, by just listening to him. It is also about the choice of words used on the call, and the number of times these are repeated; you can tell if something is wrong. You can give a score to the call. Speech analytics can determine all this.

BW: How can speech analysis help companies improve customer experience?

Ady Meretz: Customer experience is a combination of three things. One is the customer interacting with the enterprise through its employees. The customer can interact through a contact centre or through a branch office. You want to empower the employee to interact with the customer so that he can take decisions and respond. Customer experience is about empowering the employee to use analytics to support the customer in the best way. But you need the tools to enable the employee--the tools are e-learning, and quality monitoring. We also have knowledge management tools. This pushes the right information to the agent during the call. He needs the answers immediately.

BW:  How does your solution work in the backend?

Ady Meretz: After the call, the solution transcribes the call to text. We then run this through the analytics engine to determine certain trends in the call. The call can be classified based on predefined categories. The system looks for a combination of words and then assigns the call to a particular category.
If a customer is livid (and on the verge of leaving you) the solution will raise an alarm in real-time and alert the supervisor. He gets a transcription of the call, with highlighted keywords. He can then jump in and assist. That's instant escalation.

BW:  How have you digitally enabled the Verint platform?

Ady Meretz: A few years ago, we acquired a company called Kana, which enabled us to add all the digital channels. Digital combined with voice and analytics is the right way to approach customer service today.

We also acquired Opinion Lab half a year ago. The company focuses on digital feedback. This improves the customer experience. This is opt-in feedback from the customer. While on a website, if they face a problem, they can click the [+] sign to instantly give feedback. This is in the context of the page you are viewing and is especially suited for e-commerce. You will see this on the HSBC, Avis, Holiday Inn, and Dell websites. And this is available on the mobile platform too.

So with this feedback from one customer, one can fix a problem that could impact thousands of customers.

Today there are three players in this space: the legacy CRM, the legacy ACP (Agent Communications Panel) vendors, and players that offer fully integrated solutions.

BW: With AI and RPA (robotic process automation) coming in, how does that change the game for customer services?

Ady Meretz: We also offer automation and optimisation. For instance, it could be back office optimisation of a bank's processes. It could also be on logistics. These tools are part of our solution which includes data analytics on the desktop. On top of that we have the robotics tools -- once you define a process, our solution can automate it. This can save valuable man hours on simple, repetitive tasks freeing up agents to focus on more complex matters. It can also help the onboarding of new employees by providing them a guided learning environment, significantly reducing the time to become productive.