Three hundred and sixty-two billion. That’s roughly how many emails are sent every day globally. Despite numerous apps promising to replace email over the years, that number is set to keep growing, reaching 392,5-billion by 2026.
That’s because, for all its flaws, email remains an incredibly powerful form of communication, writes Ryan Falkenberg, co-CEO of Clevva.
It can facilitate intimate conversations between two people, is the default for professional communication (particularly between organisations), and can broadcast messages to masses of people. But there’s one area where many organisations wish email didn’t exist: customer service.
Don’t get me wrong; they know it’s important. But keeping up with customer emails can be a massive drain on time and resources. Fortunately, artificial intelligence (AI) is reaching the point where it can authoritatively answer those emails as articulately as any human agent and in a fraction of the time.
The trouble with traditional customer email
Before looking at how AI could help solve any issues around customer emails, it’s worth understanding why they take up so much time and so many resources in the first place.
Part of the problem is simply volume. Many companies, particularly those with consumer-facing products, receive huge volumes of emails every day. As the company grows, that volume keeps scaling up too, meaning that even more resources are needed to deal with it.
Unlike phone calls or chat, email also isn’t an immediate form of communication. That means back and forth between customers and agents, further adding to the volume of emails and increasing the time before the customer gets a satisfactory resolution. And if an agent doesn’t have the knowledge to answer the query, they have to flag it with someone more senior, further delaying any resolution to the query.
That’s to say nothing of the time it can take to figure out exactly what a customer’s trying to ask for. That can be difficult enough at the best of times but is made even more so when the customer is writing in their second language or isn’t particularly articulate. Not incidentally, that’s been one of the significant barriers to automating email responses.
Time to automate through AI
As an example, let’s say a customer has lost their credit card. They send an email to their bank saying, “Hi, My name is xxxxx, with account number yyyyy, I’ve lost my credit card and need a replacement sent to the below address.”
Any automated system would have to know to trigger several different actions. It would need to verify the customer, register the credit card as lost, issue a replacement, ensure that the supplied address is the same one as on the company records, and send it to that customer.
Historically, that’s been a massive challenge for automated emails. But things are starting to change. Virtual agents, for example, can now more effectively understand the full context of the customer’s email before applying the right business rules and processes. This could include an automated outbound verification call or WhatsApp conversation, the gathering of missing data, the triggering of required workflows, and then shaping the required e-mail response. All this, without requiring a human in the loop.
Email rebooted
The ability to answer quickly and accurately has massive implications for the future of email. Far from being something painful but necessary which uses up a lot of time and resources, it becomes genuinely useful again. Virtual agents with the capability to perform like human experts and ensure consistency and compliance across channels really do change the game within customer service. And it certainly liberates human agents from being trapped by this mind numbing, repetitive work.