As response speed and first-call resolution have become a norm in call centre interaction (and overall bottom line), the most forward-thinking companies are beginning to deploy technologies like chatbots to support customer needs.

Amazon’s Alexa is a solid showcase of the demand for more efficient customer service through innovation, but how does the industrial sector, historically rooted in outdated systems, equip their human call centre agents with the tools to handle all issues a customer might have?

Chatbots

There are two widely accepted types of chatbots. One is AI driven and is completely based on learning and adapting its gained knowledge to “optimise” response. The technology is impressive, but can be too complicated to solve simple tasks. The second type is rule-based chatbots. It understands the most basic questions and using an “if, than” logic to respond. For example; “When is my bill due?”

>See also: 5 things marketers need to know about chatbots

A misconception with chatbot technology is that a human can rely on it to answer all questions they might have. Apple’s Siri and IBM’s Watson platforms are pushed as an “ask me anything” kind of product, but the complexity of the questions they can handle falters as more scenarios and words are introduced.

An example:
• What time is the Houston Rockets game?
• When do the Houston Rockets play today?

These are two different ways to ask the same question. Ideally, the chatbot handling in-bound calls has been designed to recognise and handle, but what if a third way is introduced; what time does the championship start?

>See also: A CIO’s guide to chatbots: Everything you need to know

The third question that hasn’t been accounted for crashes the chatbot because it doesn’t know what to do. This small, but crucial aspect from a customer service stand point proves that conversational design is needed to capture how humans interact with voice systems like Alexa and Google.

The intersection of human interaction and chatbot utilisation

Based on extensive customer service satisfaction research conducted by ChaiOne, a digital transformation agency, they uncovered the overall needs of customers and customer service providers and then identified how chatbot technology can fit into their lives.

The conclusion of the research determined that;

• If a chatbot is designed well; and
• If a chatbot is placed at specific steps/task of the customer-agent interaction; and
• If handoff between chatbot to call centre agents are smooth and seamless; then
• A chatbot can bring value to the customer and overall the business.

>See also: Chatbots should be experts, not virtual assistants

Industrial companies have entered a grey area from a call centre perspective. Their customers desire seamless issue resolution with human-like interactions, without distinctly calling attention to when those interactions are with a chatbot or an actual agent.

On the call centre side, agents want to automate and speed up certain tasks and processes. What this research concluded was that it’s the questions that customers ask that determine if a rule-based chatbot will be successful in resolving their issue.

With this knowledge, where is the opportunity for cost savings in industrial sector call centers? Eliminating “choke points”.

In a legacy system, there may be four different databases of information an agent has to sift through, load up and understand to answer a particular question. This is most likely a pre-made or “canned” response.

>See also: Chatbots: catering to the instant shopper

When large amounts of inbound calls come in and all operators are using the system, this is a choke point. A customer is put on hold, or assistance from a manager is requested, or an agent says their system is running “slow” today, it’s almost assuredly the choke point swelling. It’s a place where the process is stuck, the agents can’t handle issues efficiently and customer service dwindles.

Eliminating chokepoints with chatbots

Companies like ChaiOne can identify these choke points and design chatbots to soothe the process. They utilise the benefits of both chatbot efficiency and human satisfaction.

The start of a call to resolution is seamless. For example, a new customer calls in looking for all information on a brand new loyalty program and would like to understand the comparison to past programs.

An agent types it into the system and the chatbot finds the information. The agent is not the one looking through the library and can handle all customer issues in a timely manner.

The integration is intuitive. The platform is built parallel with everyday operations and a pilot is developed. The pilot is introduced to a certain market and/or audience and improvements are made upon each rollout to a different pilot.

>See also: Rise of the chatbot: security concerns

Most systems are running on SAP or Oracle so chatbot engines are developed in the cloud, which integrates into these systems with a messaging interface as well.

There is little training involved. With a system integrated in this fashion, two opportunities for cost savings are hiring and training of new agents. There are two different types of agents in a call centre: rookies and veterans.

Naturally, rookies have the highest turnover. Because of this chatbot, companies are taking in fewer rookies and making current rookies into veterans by investing their time into customer service training with new chatbot systems.

Now call centre agents are undergoing training with the new system to be more technical and capable of providing a higher quality of customer service, in addition to enhanced problem-solving skills.

1. At the end of the day, costs are reduced dramatically.
2. Customers get their answers faster;
3. The data can be mined and integrated into the chatbots for further improvement; and
4. This generates indicators on how to improve overall business.

The ROI is a no brainer.

This is only the beginning

There is no doubt about it, chatbots are at a cross road regarding optimising its use in business functions, but only for the moment. Amazon’s Alexa was created with developers in mind.

>See also: The virtual assistant: the banks who are deploying chatbots

It’s still with faults but is more advanced as it gives the developer a “sandbox”, or the ability to test and run their own programs freely to increase the overall usability of this technology.

The middle ground that has developed over the past couple years is actually one of the best things that could have happened in terms of exploring chatbot uses because human interaction is still valued so high.

Industrial IT leaders have the ability to equip their employees with technology that makes them better at customer service, update legacy systems to a more seamless process and all without breaking the bank or hamstringing daily business operations.

Sourced from Gaurav Khandelwal, CEO of ChaiOne