How Much Are Automatic Customer Service Responses Worth?

As you can probably tell by our blog and social media channels, we are big fans of chatbots- that, and any other tools that help retailers automate their e-business, in general.

Chatbots are still a relatively recent trend. The big boom has come thanks to Facebook Messenger chatbots, after all. And commerce through Messenger is just getting started. Although we are convinced that the future of e-commerce lies in more direct and conversational ways of communication, we had to ask ourselves: 'What's the real ROI of chatbots and CSRs?' Or, to put it in another way, what are automatic responses worth?

Chatbots haven't met the expectations of the public yet, but audiences state that they do want to use them, and consider them the future.
According to the Aspect Consumer Experience Index, 61% of consumers agree on chatbots being the future of customer service.

There are over 30,000 chatbots on Facebook Messenger at the moment, and they keep adding interesting features to them. For example, Facebook Messenger bots can now accept payments.

Messenger Native Payments from Facebook on Vimeo.

According to "The 2016 State of Digital Transformation", a report by the Altimeter Group, one of the main challenges for digital transformation is ROI. Almost 70% of leaders lack data to justify the value of digital transformation, and marked this effort as "challenging" or "very challenging" in the survey.

So, how does a chatbot demonstrate a return?

What's the ROI of automated customer service?

Investing on customer engagement is a good idea, that's beyond discussion. Based on the 2016 Aspect Consumer Experience Index, 49% of consumers stopped doing business with a company during that year after a bad customer experience. And general retail is the industry most susceptible to losing business because of bad experiences.

Here's an interesting data from the same Index: Two-thirds of consumers feel good when they can handle a customer service issue without having to talk to a person. Millenials in particular. And messaging keeps growing as the preferred communication system.

So, does talking to a chatbot translate into a positive customer experience?

Almost. There's still a lot of work to be done, but according to the numbers, consumers rate chatbots as friendly and nice to use, even if they aren't always accurate, which makes the overall experience of talking to a chatbot positive in most cases.

43% of consumers surveyed by New Aspect Research even said that, in a perfect world, all customer service should be done through AI assistants or chatbots.

Conclusion: The customer experience using chatbots is generally a positive one. Check.

But, are AI assistants cost-effective?

If your first point of contact with the consumer is a chatbot, you reduce the number of costlier voice calls. Your contact center will have to pay less even in outbound calls, whenever those are necessary.

But that's not all. A huge portion of customer service questions are easy to answer, and they repeat over and over, which is exhausting and boring for employees.

"McKinsey recently estimated that 29% of customer service positions in the U.S. could be automated through chatbots."

Simple and repetitive questions can be easily addressed by artificially intelligent assistants. Chatbots don't get tired, they don't call in sick, they don't feel demotivated for repeating the same answers again and again, and they can be as accurate as human beings if properly designed.

Another factor to consider is speed. Chatbots can handle more parallel conversations than any human agent, and they're always ready to answer. It's automatic, clients don't need to wait. Speed is highly valued in customer support.

Based on a study analyzing 65.6 million live chat conversations in over 14,000 companies, speed of response has a big effect on customer satisfaction. The current waiting standard is one minute, and the lower, the better. Even when the chatbot isn't enough and you need to pass your client to a live person, this process should take 2 minutes or less. Most clients think it usually takes too long to reach a live agent.

Read more: 3 questions your support team is sick of hearing.

In general, chatbots are cost-effective and they reduce labor costs in the long run. You can have a smaller, more motivated and efficient team of representatives focusing on the really complex issues, while the bots handle the rest. Resetting passwords, validating discount coupons... Or answering typical support questions on social media channels, not only on the site.

"Companies who ignore support requests on social media see an average churn rate that’s 15% higher than companies who don’t." via Gartner.

Read more: Is 6 minutes too long to wait for a social media reply from a company?

"And on the flip side, a study by Bain & Company suggests that when companies engage and respond to customer service requests over social media, those customers, on average, end up spending 20% to 40% more with the company."

Still, not everything can be handled by automated responses. Here are some tips to make the chatbot experience as enjoyable as possible:

  • There should always be a clear option to talk to a human being, no matter what, to avoid uncomfortable customer experiences.
  • The best chatbots are custom made to match the brand's personality and engage with customers in a direct and natural way, as natural as possible.
  • No matter how natural your chatbot sounds, it will never sound like a real person. Always let customers know they're talking to a chatbot, from the start.

Read more facts: 75 customer service facts.

Some questions to ask yourself

These are some of the metrics you should find out in the first place for chat success:
Source: ICMI.com

  • What's the average time customers have to wait before a representative answers?
  • What's the average time your agents take to type a response on the chat?
  • What's the number of your agents vs. the number of incoming calls? What's their availability ratio?
  • What's the average number of parallel chats?
  • How effective is each representative? How many chats can they handle at the same time?
  • How long does it take to resolve the entire issue?
  • How many issues are solved by just one agent, without having to pass the customer around?
  • How often can agents speed up the chat using canned responses?
  • Do you have a FAQs section or similar, so customers can find the information elsewhere without having to contact support?

Calculate these metrics, and if you think that an automated response system will help you and your team, give us a call. We'll be happy to assist you in the process building a custom chatbot with Ebba.ai.

Your e-mail address is required.

Your e-mail address is invalid.