Chapter 2 Why is Customer Service Automation Important?

The importance of automating your customer service may not be readily apparent. Think back to your own experiences as a client or customer. Have you ever had to stay on the phone with your bank, while the consultants transferred you from one department to another for an hour? This is just one of the many common headaches resolved by proper customer service automation.

Over half a century of accumulated knowledge and experience in the field allows us to confidently state a few facts about the average customer. Firstly, they hate waiting. When they ask a question, they expect an answer in a reasonable amount of time (read: instantly). They demand answers which solve their problems, and they must be handled with respect. Recent history has shown that no brand is safe from the “bad customer service experience turned PR nightmare” scenario.

Every modern brand, then, should strive for customer support excellence. The exact definitions of “quality” and “excellence” will vary from company to company, of course. What doesn’t change, is the fact that changes will need to be implemented both at the lowest, most detail-oriented level, as well as on a large, company-wide scale.

Many different factors contribute to creating a positive customer support experience. Presently, customer service automation is the easiest and most efficient way to improve and optimise as many of these aspects as possible.

Chapter 3 What are the benefits of customer service automation?

• Streamlined service processes
• Better, smoother customer – and agent – experience
• Lowered costs: doing more with fewer resources
• Better access to actionable insights
• Providing 24/7 customer care becomes easy
• Expanding to different markets without additional overhead
• CS agents have more time to provide quality service

Chapter 4 What are the cons of customer service automation?

• Lack of a personal connection can turn off older customers
• Proper implementation requires constant maintenance
• High initial entry cost: research and design takes time and effort
• Errors in pipeline design can derail the entire process

Chapter 5 Best practices for customer service automation

Although automation opens up an incredible amount of possibilities, these possibilities hinge on a single thing: getting it right. As with any innovation, many different businesses, brands, and companies will immediately jump on the bandwagon. Only a few, however, will implement automation properly.

How do you make sure you’re one of these success stories? There are several best practices you should follow:

Augment, not replace. Automation is great for checking the shipping status of an order, but they won’t recommend any accessories that go well with a hoodie the customer is interested in. Many critics of automation raise the issue of job loss and replacements – after all, why would you pay a human employee for a job a robot can do for free? Indeed, in many poorly run businesses, this may happen! Make sure you’re not one of them. Automation is a tool for your employees, not a replacement for them.

Research, experimentation, testing, more testing. Customer service automation is not a one-and-done kind of deal: either you’re in for the long haul, or you fall off. Introducing automation to your customer service pipeline is just the first step. One of the inherent characteristics of this process is how much data it produces on a daily basis. It’s a true treasure trove of actionable insights you can use to improve not only your product, but also your customer service process as a whole. The key to success is constantly anticipating and meeting customer expectations – and doing so before your competition beats you to the punch. Instead of throwing random ideas at the wall to see which one sticks, why not listen to your customers?

Keep it simple and think of your audience. It may be tempting to introduce automation to every single aspect of your customer service process, but many brands don’t really need a lot of automation. If your primary client demographic includes seniors, they may turn away from you if friendly customer service agents are suddenly replaced by a confusing phone menu. Make sure you know what your customers expect and use automation to only use automation for those scenarios where it would make the most sense.

Chapter 6 Common Customer Service Automation mistakes

Customer service automation is not a one-and-done deal. By far the most common mistake made by brands is to misunderstand the fundamental truth about automation: it’s a constant, repeating process.

One of the main benefits customer service automation provides is a constant stream of actionable insights and data — virtually for free. It would be foolish not to utilize this literal font of knowledge to constantly improve your product or service. Most importantly, you can use this information to continue optimising and tweaking your customer service pipeline.

Another common mistake is not fully committing to the process. Since many different factors contribute to the overall customer experience, they should all be treated with equal care and attention. Cherry-picking the easiest pipeline stages to automate may be tempting. This temptation, however, must be resisted at all costs. A holistic approach to constant improvement is the only sure-fire recipe for success.

Chapter 7 Tools for customer service automation

There are many tools available to brands looking to enter the world of customer service automation. You may already be familiar with some of these – after all, the customer service sphere is very much interconnected. As such, many software solutions offer a lot of different features to meet as many needs as possible. Because of that, instead of describing specific tools, we’re going to describe their broad categories.

Analytics. Broadly speaking, these tools go through large sets of data looking for statistically significant patterns. Modern marketing lives and dies by the correct interpretation of analytics data. In the context of customer service automation, analytics are used to identify patterns and develop insights about a product, service, or app. The conclusions from this process can then be used to improve certain elements of the user experience to reduce the amount of enquiries.

Chatbots. Chatbots are exactly what it says on the tin – bots the user chats with. Ever since the first chatbot was released in 1966, and even more so since the internet boom of the 1990s, their potential in customer service and marketing has been explored. Thanks to recent advancements in natural language processing, chatbots are now able to detect user intent with a very high degree of accuracy – which makes them ideal for performing many tasks. Customer-facing bots can be used to answer the most common questions and perform various tasks, such as checking a bank account’s balance, looking up the status of an order, or listing the nearest store location.

Social media management tools. Ever since the modern internet concentrated around several high-profile social media platforms, tools have been in development to help brands manage their online presence. After all, who wants to have to switch between six different apps to make sure you’re answering every single message? Social media management tools integrate all of your different platforms into a single interface for easier management.

Routing tools. Although technically this is a feature of some social media management tools, the impact it provides is significant enough to warrant its own section. Simply put, routing tools are used to automatically assign support tickets to available customer support agents. This ensures no enquiry “falls through the cracks” and remains unanswered. They can be used in conjunction with natural language processing algorithms to assign customer questions to the relevant team based on the message’s content. That means support questions would go to the support team, while sales enquiries would go to the sales team. This approach is known as smart routing.

Chapter 8 How to automate your customer service pipeline

1. Do some research: gather historical data from your recorded calls and message logs. What questions are most commonly answered? Can you improve your app / service / product to address these issues?
2. Identify patterns and segment incoming requests into different categories: support requests, product questions, marketing enquiries, et cetera. 
3. Identify areas in which your product, service, or app could be improved in order to reduce enquiries. Go through your how-to and help pages and make sure they are relevant and helpful to your users, based on data from step 1.
4. Establish different customer service teams based on competencies. Different and separate teams should handle billing questions and tech support, for instance.
5. Integrate all of your communication channels into one tool for easy access, monitoring, and analytics. This not only saves your team valuable time which would otherwise be spent jumping from app to app, but it also allows for easy expansion to new services and platforms.
6. Set up automated routing so that every request is assigned to an agent. This ensures no question “falls through the cracks” and remains forever unanswered. Since enquiries are automatically assigned to agents, the time to first response is drastically reduced.
7. Prepare response templates so that your agents can quickly resolve the most common issues.
8. Implement an AI answer suggestion algorithm to help your agents pick appropriate responses even faster. This is another timesaving measure – why waste time scrolling through a list of templated responses, when an AI can do it almost instantly?
9. Deploy a chatbot interface to take care of the most common requests and enquiries, thus allowing your agents to focus on what’s really important – delivering high-quality customer service even in the toughest support cases.

Chapter 9 Customer service through different communication channels

We are currently in the middle of a major shift in the customer support space – the baby boomers get overtaken by millennials and Gen Z as the biggest parts of almost every customer base. With this generational shift comes a change in preferences. While the customers of old preferred communicating with companies primarily through telephone calls and physical letters, these days your clients are overwhelmingly more likely to try and reach you over the internet.

This, unfortunately, opens up a rather unattractive can of worms. The internet is balkanised – which means your client base is split across dozens of services and platforms. While it is a rather safe bet that most of your customers are on Facebook or Instagram, wouldn’t it be better to be available to as many internet users as possible?

Of course, your agents are only human. Asking your team to handle a dozen different apps and services is a tall order – there are only so many hours in the day, and constantly switching from Facebook to Twitter to Instagram to WhatsApp and back to Facebook will eat into that time very quickly. As a result, the quality of your customer service will suffer. Sure, you could hire additional personnel and divide the work more evenly, but that means additional costs. You can have it done well, or you can have it done affordably. It’s one or the other, but it can’t be both.

…or can it? Omnichannel customer service is a very fancy term referring to integrating as many communication channels as possible into one tool. It solves both aspects of the problem: your team doesn’t waste time juggling apps around, which gives them time to provide a consistent quality of service. The amount of time saved is so great, that even a relatively small team can take care of ten communication channels without a problem.

From the customers’ point of view, they get the same quality of service no matter if they contact you through WhatsApp, Instagram, or Facebook Messenger. To them, your brand seems more in tune with their needs – after all, they don’t need to sign up to another platform just to talk to you. And since channel integration is actually pretty simple, if you discover a new service which is gaining in popularity among your customer base, you can set up your presence there at no additional cost.

Chapter 10 Common Customer Service Metrics

When it comes to measuring the quality of your customer service efforts, several metrics are used. Two categories of these metrics can be distinguished: operational metrics and experience metrics. Operational metrics concern the actual, tangible activity of your team, while experience metrics come from surveys and concern the customer’s opinion of your CS process.

Examples of operational metrics include:

• Average ticket count – how many tickets your customer service team receives within a day, a week, or a month. A constant increase in ticket numbers could indicate a potential problem with your app, service, or product.
• Average time to first response – how long the average customer has to wait for a response from your team. Depending on the communication channel used, there are different benchmarks to meet: customers are willing to wait 24 hours for a response to their email, but only 60 minutes for a response to a social media post.
• Average ticket resolution time – the average time agents spend working through a single ticket before resolving it. 
• Preferred communication channel – this is rather self-explanatory. It’s a simple calculation of the relative popularity of each of your customer channels. Observing this metric can help you anticipate changes in the market and adapt to your customers’ preferences.

Examples of experience metrics include:

• Customer satisfaction – although the precise definition of customer satisfaction will always be rather vague, most brands define a satisfied customer as someone who rated their experience at least at a four on a one-to-five scale. Despite its vague definition, it is one of the most important indicators to observe.
• Net promoter score – alongside customer satisfaction, this is another important indicator. It’s measured through a simple survey question in which the customer is asked how likely they would be to recommend your app, product, or service to others, on a scale of 1 to 10.

Chapter 11 The current state of customer service automation

Customer service automation has come a long way in the last twenty years – but this is just the beginning. Most large brands either use their own in-house CS automation or outsource it to third parties. The current trend is making the technology accessible to smaller brands. Easy to use tools allow every company to implement automation within their CS teams in a matter of hours.

For a few years now, AI has emerged as the new trend within the customer service automation space. More and more automation solution vendors are experimenting with adding AI features to their software: be it in the form of smart routing based on keyword detection or response suggestions based on natural language processing algorithms.

A related trend is implementing Conversational User Interfaces – or, as they are more commonly known, chatbots – to handle common customer enquiries and requests. They provide a convenient way to check your bank account balance, ask pricing questions, track orders, and more – all from the comfort of Facebook Messenger, or wherever else the bot is deployed. Although far from mass-adoption, it’s very likely that chatbot technology will soon establish itself as a mainstay in the world of customer service and marketing.

Chapter 12 Customer service automation through SentiOne

SentiOne provides a one-stop package for customer service automation. Our flagship product, React, is a complete automation solution aimed at both large brands present on multiple markets, as well as smaller companies. Its omnichannel integration feature allows you to seamlessly integrate every communication channel you can think of into the same interface – which means you can be everywhere where your customers are. Our smart routing feature will ensure no question goes unanswered. Best of all, the app is incredibly simple to set up and use, even if you’re serving ten different markets in a dozen languages.

React is augmented by two other modules on the SentiOne platform – Listen and Automate.

SentiOne Listen is a social listening and sentiment research tool. It collects data from the entire public web (including social media platforms) looking for any mentions of your brand. Sophisticated NLP algorithms provide users with highly accurate sentiment detection, allowing you to meet customer expectations based on feedback from public discussions. It also provides a detailed statistical breakdown about the people mentioning your app, product, or service.

SentiOne Automate is a complete chatbot solution which allows users to create sophisticated chatbots using our proprietary, bleeding-edge natural language processing algorithms for intent detection. We strived to make creating even the most complicated bots as simple as possible – even non-technical people can easily create an AI assistant in a couple of hours. Despite this, they are highly customisable – they can perform simple tasks, such as answering commonly asked questions, but can also be tied into external services to act as a true conversational user interface for your own platform.

If you’re interested in exploring the SentiOne platform, please visit this link and schedule a free consultation.