Why email in customer service is on the way out

We live now in an instant economy, where as societies, we expect to get everything on demand. It’s almost a daily reality, where products appear on our doorstep within 24 hours and the newest episodes of our favourite series are just a click of a button away. The shift has of course not been immediate, but has started already decades ago when we moved from analogue to digital, moving to the rapid growth of the internet, smartphones and social media. Through the decades of development, each industry in their turn has tried to adopt.

The big winners of this change have disrupted billion dollar industries, be it Amazon for ecommerce, Slack for communication, or Netflix for entertainment. The latest change has been the shift to instant communication.

The natural extension of this wave of instant messaging is of course instant customer service. As Mark Zuckerberg has said already in 2016: customers should be able to message a business the same way they message a friend.

It’s no news anymore, but customers expect more and more from companies in terms of service and response times. In fact, Conversocial found out that 81% of consumers have higher expectations for their digital customer service, than they did the year before (in 2017).

What does this mean? Amongst other things, that time is the new gold and the way to the consumer’s heart in 2018 is instant gratification.

So why don’t we have 24/7 hotlines everywhere without any waiting times? Wasn’t it about instant gratification?

The reason is of course cost related. Companies simply cannot afford to provide that level of service. In most cases, customer service organisations have to balance between two contradictory requirements:

• Minimise costs
• Maximise customer happiness

Unfortunately, as we all know, this does not work too well. Customer service organisations are expensive to maintain, and the waiting times are long: the average email response time hovers around 12 hours.

Synchronous vs. Asynchronous support channels

There are two types of support channels: synchronous and asynchronous. When something is executed synchronously, one task needs to be finished before one can move on to another task. When you execute something asynchronously, you can already perform another task before finishing the first.

A great way to explain this in practise, is to look at the ways we contact customer service. As an example, calling a hotline — and getting through to an agent instantly — is a synchronous means of communication. An email, however, is not. Emails are asynchronous: when an email is sent, it does not require immediate attention / action from the receiver. Multiple emails can be sent exactly at the same time, without it interrupting the reception / availability of the recipient. The recipient can also reply to the emails in any order they like, or draft replies to multiple emails at the same time, and then replying to all, at the same time.

Yet email as the primary means of contact to a customer service is extremely common. This is because “buffering” with email helps companies manage the resource allocation problems that customer service organisations commonly have. Providing email service allows service organisations to utilise their workforce more efficiently by “piling up work” for the moments when f.ex phone service is less in demand.

And that is the reason, why the average email response times are so long, which of course is clearly frustrating for the end consumers. For companies, it’s understandable why they prefer email. So to conclude: at first glance, email is disadvantageous for end consumers, but quite good for companies.

However, that’s not the whole story.

Email vs. phone — the epic battle of asynchronous vs. synchronous

As said, from company perspective, email seems like a perfect, “cheap” option as you can create a buffer and have a higher utilisation rate of the agents at all times.

But there’s a significant caveat: the cost per contact is ultimately defined by time to resolution — of the agent.

Let’s look closer and investigate this through a simple mathematical example.

The cost per contact is defined as follows:

all contacts / all costs of a contact centre

For example: 1 minute = 0.5 €

Email:

• A customer and service agent exchange two emails back and forth.
• Time per piece (=email) 3 mins x 2 = 6 minutes = 3 €

Phone call:

• Customer calls directly to service department and both questions are answered in one call.
• Time per piece (call time) 5mins x 0.5€ = 2.5€

In simpler terms, the assumption that email is always the cheaper option, is in fact unreliable. If you fail to resolve your customer’s problem — or address their questions — in one single email, it can massively increase the service costs involved. But with telephone, it’s not possible to perfectly utilise the agents at all times, in order to avoid waiting lines.

Now this is how the loop closes: people prefer instant communication, the trend is there. If people could always — around the clock & immediately — reach a support agent via phone, it would be the fastest & “most instant” path to resolution. It might be a bit far fetched here, but in the next 5 years, customers will have understood this.

Due to its synchronous nature, instant support channels will (slowly) kill email, the same way that email has killed direct mail.

The Holy Grail of customer service: low costs & real-time service

So how would ideal customer service look like? It would always be “real-time” (like telephone), where the customer doesn’t have to wait to get serviced. The requests would also be solved in the same sitting — at least from the customer’s perspective. Ideal customer service is also available whenever you need it, 24/7. Obviously, only very few companies actually offer this. Lastly: ideal customer service would also be low cost — human customer service is very expensive. I’m actually writing a separate piece on it.

Surprisingly enough, this ideal exists. The solution is adding virtual agents to support human agents. They combine the best from all of the above requirements:

• No waiting time — replies are instant
• Instant gratification — the majority (>80%) of requests are resolved (and if not, a human gets involved)
• Always available — software doesn’t need a day off ;)
• Low cost — software costs a fraction compared to a human employee (because it’s software). See our article on ROI of support oriented virtual agent.

Now as we are in the flow, let’s look to the future. What will happen after virtual agents have taken off? Slowly, but gradually virtual agents will bury telephone support because end consumers understand that virtual agents are faster and companies see a cost benefit. Exciting times!