Instant Messaging

The Power Of Two Way Conversations

By March 22, 0208 February 14th, 2019 No Comments
The Power Of Two Way Conversations

The tech community conditions us to look towards the future – to have a vision and to change the world in light of that vision. I tend to think that looking backwards is just as important. So many problems exist today because ideas intended as solutions turn into new issues. The state of customer service is a place where this idea is relevant.

I grew up in the suburbs with plenty of small locally owned shops. Bakeries, supermarkets, corners stores, restaurants and even sports stores all owned and operated independently. The owners of those shops were able to remember hundreds of people by name and more. Did they have a CRM? No! And it was not a problem.

This may sound like a silly question, but if these small shops with almost no resources can make every single customer who walks through the door feel special, why can’t multinational corporations?

Why is developing and maintaining customer relationships so tough?

Sound like an unfair question right? Corporations can’t make people feel as special as a local establishment. It’s just not that easy to scale relationships.

Do they need to though? It seems that consumers don’t have the same expectations of corporations that they have when entering a small business. An American Express Report found that consumers defined a good customer experience as, providing a satisfactory answer to a question from someone who was knowledgeable. That person should also be to handle the question or transaction efficiently without passing the customer off to a further person.

Those descriptions are a far cry from the reasons people like their local sandwich shop. It can be argued that the expectations set for corporations are actually lower. A report from this year indicated that Gen Z is likely to leave a service or a brand if it is slow to respond. That is not much of an increase in expectations over the last 5 years.

And yet year after year we are reading about how companies fail to meet customer expectations. The above studies indicate that a large part of this is due to being slow to respond to individual customer inquiries. 89 percent of people surveyed by the Customer Experience Impact Report (PDF) claimed that they started doing business with a competitor after a poor customer experience. A seemingly simple problem, but with no simple solution.

A seemingly simple problem with no simple solution?

Reducing the amount of Customer Service Needed

One answer is to get more customer service staff or agents on the problem. But this is expensive. Another way is to direct customers to pages within the website or ask them to contact a hotline to answer their questions, but usually, this is not enough and requires the customer to make significant effort to reach out in an increasingly text-focused age, which is not good.

Good customer service can be defined as; flexible, fast, and efficient.

Flexibility is a very important component and it starts to become tricky when a business becomes global. As a business, you are faced with time zone differences. If a consumer is in a country that has a 12-hour time zone difference to the company’s headquarters, your company would need to run a customer service that runs 24/7 in order to cater for the consumers abroad who may have problems with the goods/services being provided.

Another important component of good customer service is how fast the actual service is. Waiting for over an hour on the phone to get connected to an available agent isn’t the greatest part of a customer’s day. Having experienced this before even talking to a customer service rep can hinder a customer from having a good experience with the service, regardless of whether or not the issue was resolved.  

Efficiency is also very important as customers generally want their problem fixed in a timely fashion. Being transferred over to other departments to resolve your issue, unfortunately, takes up a lot of time.

With the increase in companies globalizing, chatbots have increasingly become a good option to save money and reduce the customer service required of global corporations. Chatbots, unlike humans, are able to work 24/7, increasing flexibility in customer service as they can work all hours and reach the customers abroad. One chatbot is able to process thousands of questions per minute. Meaning, a chatbot can efficiently multi-task between customers, therefore almost entirely getting rid of wait times.

Since chatbots have access to databases, one chatbot can deal with someone’s laptop not turning on and another customer’s laptop not connecting to the wifi at the same point in time. As previously mentioned, humans don’t have the capacity to deal with two different subjects, hence why we as consumers, tend to get transferred around to different departments. Chatbots can multi-task and very well.

The drawback of chatbots is the incredible amount of effort and investment needed to get a chatbot up and running, and the further maintenance of a chatbot due to it’s tendancy to fail. Most companies make the mistake of thinking Chatbots are the end-all be-all solution to a large and more costly customer service staff. The end result is a multi-million dollar investment into a chatbot that ends up failing, and dissapointing customers even more. The other cherry on top of this poo-cake is that customers know when they’re chatting with a bot, the conversation gets impersonal, and in the end the customer really just wants to talk with a human. 

But there is a middle ground now being found with AI-assisted Human-to-Human interactions. 

But we’re getting a little off topic, first, let’s cover the basics about why two-way conversations are so important, and where AI can assist.

Collecting Feedback

Setting up a process to regularly collect feedback is the most actionable strategy. Maybe it’s once a year or once a month–or maybe it’s after every transaction. But plan on it, and make it part of your regular routine.

A company that doesn’t use surveys often experiences more complaints about how poorly a consumer was treated over the phone rather than remarks on how great another interaction was. Thus, there (rightfully so) is a huge emphasis on active feedback through the use of surveys in order to get a more balanced idea of where your customer service is and how it is doing.

Surveys by phone have a very low response rate of roughly 8-12%. Surveys conducted via email have a much higher response rate of 24.8%, assuming all emails are opened. This does seem high, but it is still possible to further increase this response rate.

A cohesive messaging management tool could allow surveys to be sent out within the messaging platform. This way, the survey does not seem like an additional task the user is asked to do and rather something part of the conversation. This method leads to higher response rates and better quality feedback.

A big issue with having call centers is that the employees are subject to a lot of dissatisfied customers and are sometimes victims of rage. These employees report high anxiety and stress levels, rapid heartbeat, tight stomach, and negative emotions, to list a few. These issues lead to lower confidence, higher absenteeism, as well as a high turnover rate. From a business perspective, this also means more money spent on training new employees to fill in the positions.

The most important aspect of customer service is listening to your customers. What do they want? What can you improve on? Can AI help fix those problems? Employees in companies tend to have a fear of what they will hear. There is a pitfall of handling the analysis of customer feedback within an organization: the very human instinct of self-preservation is just too strong for employees with a vested interest in the status quo to not somehow shade the results as they make their way up the chain of command. Hence, another benefit of AI assistance would be that the developers sort out the analyses and provide a platform for people to see the full, unaltered results of what customers are saying about their service and relationship.

A benefit of having AI in tandem with your agents over other feedback mechanisms is that there is valuable data within a conversation. For example, if your customers frequently ask if your companies product or service includes a certain element, and your AI does not understand this request as a result of you not having that element, then this is valuable feedback. It means that your product is limited in this widely popular element and thus you may need to act on it.

As I mentioned earlier, having a conversation is imperative to understand how a customer’s needs are changing. When you know what they are looking for ‘at the moment’, it gives you an opportunity to upsell or cross-sell your solutions to them. Since they have already engaged with your business, they’re more likely to trust your offering.

Pinpoint Opportunities for Upselling and Cross-selling

As I’ve mentioned before, AI is able to cross-check multiple things at multiple times whilst considering a large number of factors. Chatbots coalesce with Facebook messenger enabling them to call a customer by name.  They explore different sources of data to show people products which may resonate with them. In essence, they make the customer service experience personal. This rapport opens up additional sales opportunities for companies – people trust the barista at their local coffee shop to suggest them a new drink or even a sandwich to go with it because that barista knows their tastes from their previous orders. AI can mimic this interaction, making informed suggestions which allow larger companies to target consumers like never before!

So you may be wondering how this would work? Well if, for example, a customer of a clothing website asked the question ‘do you sell black stilettos?’, a human operator would first have to ask more questions to narrow down a search, then painstakingly go through all the shoes on the site to make suggestions. This process is time-consuming, making it frustrating for the customer and a waste of company resource. A human operator’s time is arguably more wisely spent processing customer issues which are more complex and hold more nuance. Therefore, it is more efficient to make talking to a human the second point of call, with queries only being directed to the human operators after a chatbot has filtered out and answered the more commonly asked and straightforward questions such as the one in this example.

Following the success of the initial sale, AI can then upsell.  After the customer adds stilettos to their basket, the AI or a platform that manages messaging channels can pop up suggesting a coinciding product which the customer may have not previously considered buying, such as gel insoles to make the shoes more comfortable. The AI can even offer a discount code for the two items if bought together, creating an incentive to make the upsell more appealing. There is also the option for the AI to attempt a cross-sell, suggesting items which the customer can wear with the stilettos, such as maybe a bag which previous customers have bought and paired the shoes with. By the time the customer checks out their basket, they could potentially be buying three times the amount than they would have! It is clear to see that this way of communicating benefits both the customer and the company – the customer enjoys a more tailored shopping experience and the company gains more sales and increased customer loyalty.  Everybody wins!

AI for customer service

The power of two-way conversations in customer service is essential. AI provide the platform for this. Not only are they cost effective and more efficient, but they also allow you to collect valuable insights on what your customers are thinking and want. Customer service looked doomed as businesses failed to find a solution to scaling personable service to each individual customer. AI provides a glimmer of hope.

Gerardo Salandra

Gerardo Salandra

Gerardo Salandra is the CEO of Rocketbots ( An IBM & Google veteran, with deep expertise in growth hacking, machine learning, and chatbots. Chairman of the AI Society of Hong Kong, promoting the development and understanding of AI technologies in the Region.

Leave a Reply