Why Chatbots Don’t Solve all Customer Service Problems

Technology “of the future” that previous generations wrote about in science fiction books and movies is here with us now. Yes, even flying cars. However, one example of advanced technology that might seem simple to us now is the humble chatbot.

These days 23% of customer service organisations are already using AI chatbots in some form. Yet, 75% of customers still expect companies to use new technologies to create better experiences for them.

 

customer expectations are driving digital transformation

Image source: Salesforce

So why the huge gap between existing tech and expectations? This quick guide will cover what chatbots are (in customer support contexts) and why companies currently use them. Then, what problems they still can’t solve and solutions to these problems with examples.

What are customer support chatbots?

Customer support chatbots are conversational software solutions designed to help customers with frequently asked questions (FAQs) and customer issues instead of, or in tandem with, human agents.

To the customer, they usually appear like a bot that provides answers to a few key questions (think “please select an option”). Under the hood, these basic chatbots are programmed with specific messaging flows. These flows often take the customer to an online knowledge base for self-service or redirecting to a live agent.

Chatbots that are more advanced use some level of natural language processing (NLP) which can pick up keywords in conversational context and choose appropriate responses that way, instead of relying on pre-programmed message flows.

Why do businesses use chatbots?

Despite the title of this guide, there are real benefits to using basic chatbots, especially as a company just beginning to consider automated support solutions. Here are some of the main reasons why companies currently use them.

24-hour response times

As customer experience goes, one of the top priorities is quick response times. When surveyed, 90% of respondents from Hubspot Research rate “an immediate response” when they have a customer service question as important or very important to them.

To them, an “immediate response” is 10 minutes or less. In a globalised world, that can be a problem if you only have a contact centre in one time zone and limited availability.

Chatbots can help to provide the initial response in real-time and in some cases complete the customer interactions at any time of the day.

consumers are impatient

Image source: Hubspot

Answers common customer questions

Since the global pandemic has changed the way customers and businesses interact, companies have had to adapt and handle higher volumes of tickets for customer service teams.

Digital Genius’s own research shows that 40% of support tickets are mind-numbing and repetitive. Being bogged down with simple questions slows down a support team’s ability to deal with more complex queries.

Knowing this problem, companies build chatbots to deal with these higher volumes to answer common customer questions — which in turn, leads to faster case handling times.

Frees up customer service agents

Each of the previous reasons leads up to this one — chatbots help to filter out common questions which free up human support agents to work on cases that need more help.

In fact, 70% of consumers surveyed in 2020 suggested they either already use or are interested in using, chatbots for simple customer service needs.

 

extent of customer use

Image source: Salesforce

In theory, this means you could lower your human customer service personnel and still have high customer satisfaction, win-win, right? Unfortunately, it’s not always that simple.

Problems customer service chatbots don’t solve

For all the good chatbots do for businesses looking for solutions, they aren’t perfect. At worst, they can impinge on the customer journey and create a bad customer service experience. Let’s take a look at why this can happen.

Chatbots don’t take actions

Remember how we said chatbots help to answer common questions? One of the struggles of simple chatbot technology is the inability to complete backend processes. Instead, they rely on the end-user to take further actions themselves.

Part of the reason for that is in many cases, contact centre agents use between 5 and 8 different systems to resolve inbound queries.

The result is that the customer is frustrated because they either have to specifically request a human agent or give up.

They’re less personal than they seem

While a chatbot is usually designed to have an on-brand personality, that doesn’t mean having them leads to a more personalised experience.

The fixed responses in pre-programmed workflows can actually have the opposite effect. According to Acquia, 75% of customers feel these experiences are impersonal.

The key here is that personality doesn’t always mean personal — without customer relationship management (CRM) information and process integration, customers can feel they aren’t getting a better customer experience

They’re limited by keywords or fixed responses

This problem feeds into the previous one — fixed responses make it so that the customer is limited in their ability to describe their issue. These can lead to frustration and customers ultimately feel the need to speak to human beings.

Having a chatbot adds friction to the process for queries that aren’t simple. You’ll often find customers are already in a bad mood having wasted time before they get to the human interaction.

The solution: artificial intelligence, machine learning, and process automation

The problem with chatbots is that they fall short in providing meaningful solutions to issues customers face. Even with simple requests. So where do you go from there?

Enter: AI, machine learning, and process automation.

The majority of failings chatbots have can be resolved while keeping the benefits by using these solutions:

AI and machine learning can use historical and new transcripts and learn how best to solve queries from them — goodbye fixed responses.
By connecting CRM and customer service software, process automation can help take actions, not just tell customers where to find more information.
Advanced natural language processing can detect frustration and route the query directly to an agent — no more wasting time.
Let’s take a look at a couple of examples where companies have used these techniques.

Examples of AI successful use cases

With each of these examples, we’ll briefly explain their goal with AI automation and their results:

Freeletics

EXAMPLES OF AI SUCCESSFUL USE CASES Freeletics

Digital fitness coaching company, Freeletics, came to Digital Genius with the issue of long case handling times (days in some cases) after seeing significant growth in customers. After implementing Digital Genius’ automation services they saw:

  • 90% CSAT on all cases handled by the automation.
  • 3 minutes saved on every ticket.
  • 50% decrease in average handling time.

KLM

adopting AI guided automation solution KLM

As one of Europe’s main airline services, KLM receives high traffic volumes from multiple channels of communications in multiple languages every day. They recognised that hiring more employees wasn’t the best solution for them.

After adopting AI guided automation solutions from Digital Genius, they achieve:

  • 130,000+ messages processed each month.
  • Half are supported by Digital Genius.
  • 95% accuracy of responses

Delight your customers with AI today

By now, you should have an understanding that even though chatbots can be helpful in some situations, they aren’t the perfect solution for customer service problems.

With the help of Digital Genius, you can mitigate the issues that chatbots can’t overcome by using AI, machine learning, and process automation alongside your human customer support teams.

Request a demo today.

Subscribe to Our Newsletter

Get updates by subscribing to our newsletter