The landscape of tools and software for support teams is vast and ever-changing. Get up to speed with some of the emerging trends that can help give your support team the edge.
Guest post by Sarah Chambers, Founder, Supported Content.
Today, there are many, many tools emerging that promise businesses quicker, faster, more effective ways to work smarter and serve their customers. To keep up with the competition, customer support leaders need to keep up with the latest technology on the market.
But how do you choose which ones to experiment with? Some are more useful and more interesting than others. Some overpromise. Some are hidden gems that make a big impact with little setup. How do you know which tools will provide a big return on investment for your team?
Staying on top of emerging tools is important because, at the end of the day, customers want to solve their issues with as little effort as possible. Anything you can do to make their experience as efficient as possible will help build loyalty and, in a lot of cases, beat your competition at customer service.
In this section, we’ll cover three emerging trends in customer service tools, to help you understand the opportunities to seize and the pitfalls to avoid: Artificial Intelligence, Customer Experience Platforms, and Contextual Tools.
Key Artificial Intelligence technologies
Much has been hyped about how AI, machine learning and automation will impact customer service in the future. Will we all lose our jobs to the robots? It looks unlikely. But AI is already having a big impact in how teams scale their customer service efforts. Let’s look at what tools are currently available, and how your team can harness the power of AI.
First, a quick primer to AI for those new to it. There are three big terms you need to know:
Artificial Intelligence is the ability of machines to simulate the mental characteristics of humans, such as the ability to reason, discover meaning, generalize, or learn from past experience. AI systems you have probably encountered before include Spotify’s recommendation engine and Tesla’s self driving cars.
Machine Learning is the ability for a computer to automatically learn and improve from experience without being explicitly told what to look for. Machine learning is a technique to develop AI. It’s based on algorithms that crunch large amounts of data to find patterns and meaning. For example, Google engineers programmed their AI to recognize photos of cats, without telling the AI to look for whiskers, ears or cuteness. They just showed the algorithm many pictures of cats, and the AI understood ‘cat.’
Natural Language Processing is a way for computers to process and derive meaning from human language. Understanding the meaning behind human language is one of the most difficult tasks AI has been given. While it’s easy to train a computer to recognize phrases or words, developing an AI that can truly read, write and analyze tons of unstructured content is a lot more difficult. For example, a human would understand that crashed, downtime, outage, and blue screen of death are all talking about something similar. An AI would need to learn each one individually.
Most AI customer service tools use NLP to understand what customers are saying, along with other machine learning algorithms to automate a main customer support function. There are three main places where AI is already making an impact for future-looking customer service teams:
- Customer facing automation
- Smart agent assistants
- Insight analysis
Let’s look at each one of those in a little more detail to see how they might help your team deliver even more amazing customer support more efficiently.
When automation is done right, it helps customers get help quicker than if they had to talk to a human, and customers love that. Nobody wants to pick up the phone or wait for an email reply if they don’t need to.
In corporate speak, this is called ‘ticket deflection.’ Every time a customer finds what they need through self service, it means one less ticket for a human to answer. Ticket deflection doesn’t just make things easier for the customer, it also reduces the cost of support for the business.
AI automates ticket deflection by suggesting answers to customers when they write into support. A tool like Solvvy connects into your existing knowledge base of help articles to suggest relevant content to customers submitting a ticket. It uses NLP to understand what a customer is asking (instead of just searching keywords) and provides snippets of text from your help center when it believes it can help. If Solvvy recognizes the issue needs a human, or if the customer says they still need help after reading an article, it doesn’t get in the way. Customers get helped by the next available human agent. Other examples of customer facing automation include Zendesk’s Answerbot and Intercom’s Operator.
However, there’s danger in using automation as a shield between customers and your support team. If customers are sent help center articles that aren’t helpful and are forced to wade through robotic emails before receiving human help, they get frustrated. That’s why it’s important to only automate a response when you’re very certain it will help, and to always leave a way for customers to connect with a human support agent.
A QUICK WORD ON CHAT BOTS
Are chatbots the same thing as AI? Not necessarily. Chatbots are often preprogrammed interactions that help customer support teams sort and triage tickets. They can usually only recognize a specific set of commands or inputs. If a customer asks something outside of what the bot is programmed to respond to, it breaks down. In fact, Facebook Messenger bots in 2016 returned a 70% fail rate in assisting customers. The bots just aren’t smart enough to provide nuanced customer support – there are too many edge cases.
More chatbots are starting to include some level of NLP to become more helpful. However, if they don’t provide an easy way for customers to connect with a human, most customers will find them more troublesome than helpful.
SMART AGENT ASSISTANTS
Rather than subjecting your customers to the whims of a robot, some AI tools work on the side of the agent. Tools like DigitalGenius, Guru and MonkeyLearn help agents respond more efficiently to customers by automating manual tasks. Agent assisted AI tools will help increase the speed of agents by:
- Automatically prioritizing tickets based on sentiment and content
- Automatically tagging and routing tickets to the right team
- Suggesting macros that have resolved the issue previously
Agent AI assistants work on the premise that your agents know how to provide the best support. Agents are able to filter, personalize and correct any suggestions before sending responses to the customer. The agent feedback is then used to update the algorithm for better solutions in the future. If the confidence level becomes high enough, teams can decide to automate the reply without agent intervention. As a rule of thumb, most software tools recommend a confidence interval of at least 95% (meaning that the automated reply will resolve the issue for 95 out of every 100 customers) before removing human review.
The final big use case for AI in customer support is in analytics or insights. Customer conversations contain an enormous amount of qualitative data. They tell you what customers are looking for, what they’re frustrated about and what they love, but sorting through all those conversations manually is impossible. Even if you do manage to come up with a tagging system, its accuracy is dependent on how well teams understand it and remember to tag conversations. It’s also inherently biased towards the opinions of the support team. You don’t know to look for trends you don’t know about.
For example, imagine your product team wants to know what customers are asking about in the reporting module. Each agent on your team might have responded to a few tickets about reporting that week. They have their own window into what customers are asking for, but they can’t see the big picture. It’s just not possible. They’re relying on anecdotal data based on the small number of tickets they’ve seen.
That’s why many customer support teams are looking to use machine learning algorithms to dive into their wealth of data. Tools like NomNom and Idiomatic use machine learning and NLP to layer NPS, customer satisfaction and sentiment on top of the ticket content to highlight what customers really want. Turning the qualitative into the quantitative can help organizations stay truly customer focussed when looking into customer feedback.
Implementing AI tools
Most customer support leaders picture AI as the future, meant for large organizations but not them, not just yet. But AI is actually becoming more approachable for medium sized support teams every day.
- Customer facing ticket deflection tools require very little setup, have minimal risk and pay off almost immediately. Check out Solvvy, Intercom’s Operator, and Zendesk’s Answerbot.
- Agent assistants require more historical data to train the machine learning algorithm with a high degree of accuracy. Teams with approximately 10,000 historical tickets could start looking into whether they would see any benefit from implementing a similar tool. Investigate DigitalGenius, Guru or AnswerIQ to see if your team is a good fit.
- Insight tools are still restricted to teams with a healthy budget and lots of data – mostly because the input needed to train the model and discover trends is so large. NomNom and Idiomatic are impressive tools for teams that have the data set to support them.
A last word of warning: AI needs to be used as a tool, not a shield. If customers feel like the robots and AI are preventing them from getting the human help they need, they will get frustrated. Customers don’t want to troubleshoot complex situations with chatbots that don’t understand them – they just want quick assistance. Customer support teams need to automate tasks that don’t negatively impact the customer experience, and keep the human aspect of customer support front and center.
Regardless of what size of team you are, it’s important to have a good understanding of what AI can offer. Teams that put the robots to work in a smart way will be able to scale customer support much more effectively than those who insist on doing everything by hand.
We suggest that companies use new digital technologies, whether that’s AI and machine learning, or using natural language processing and data science, to really unpack unstructured voice data. Companies will spend a lot of money recording their support interactions. You hear this at the beginning of every call to a call center, right? “This call may be recorded for quality assurance and training purposes.” The reality is, it is being recorded, but it’s it’s compressed down and sent to some data center somewhere and kept for a few years before it’s deleted. Nobody ever goes back and listens to those calls or learns anything from them – but, today, the technology exists to mine this voice data for customer insight.
– Matthew Dixon, Chief Product & Research Officer at Tethr, Co-author of The Challenger Sale, The Challenger Customer and The Effortless Experience
Customer Experience Platforms
The second tooling trend customer support teams can expect to see is the use of cross-organizational platforms, sometimes referred to as ‘systems of intelligence.’ These platforms sync communications and data between an organization’s CRM, support software and other tools, creating a more complete picture of the customer’s experience.
We’re already seeing these large omni-channel platforms crop up with the release of HubSpot’s Service Hub, Zendesk’s expansion into the customer experience space with the Connect platform and Salesforce’s help desk consolidation under ServiceCloud.
These platforms are becoming popular because of the proliferation of data companies are collecting. Companies want to take advantage of the data they already have, but often find that it’s fragmented and siloed between business units. In order to deliver a well executed customer experience from start to finish, companies need to bring many different sources of knowledge together. Customer experience platforms help teams do this so they can see a big picture of the customer journey.
BECOMING TRULY PROACTIVE
Customer support usually won’t own the implementation of these systems, but it’s important to know what they are and the value they can provide.
Understanding the customer journey from start to finish helps customer support teams provide more proactive support. For example, knowing that NPS tends to drop after the first invoice can help customer support teams develop clearer billing resources. Access to product usage data can help customer support teams prioritize onboarding pieces to improve product education and reduce the resulting support tickets.
As customer journey data becomes more accessible to customer support teams, they have the power to develop personalized and proactive service experiences.
Tools to better understand the customer
The last section of future technology is a group of insight tools. These tools integrate into your support workflow to provide more information on the customer’s perspective. Instead of treating each ticket like a unique incident, customer support tools are starting to recognize the importance of context.
We’re also seeing this trend pop up in the form of screen-sharing tools. Median is a new option that integrates with almost every live chat provider on the market. They provide 1 click screen-sharing that makes it easy for agents to securely see what the customer sees and fix problems really quickly.
The allure of new tools
New tools are exciting and shiny but have the potential to distract from bigger goals. Tools can’t solve every problem – if your process is broken in the first place, a tool will rarely improve the situation.
Beware of losing focus. Be sparing with your choices. Don’t try every plugin, add-on or new platform or else you may experience overload, and the overall level of your service will drop.
But if you find the right tool for the job, staying up to date on future developments in customer support tools can help you scale better and keep ahead of the competition. Don’t be afraid to dive in and try something new!
Read other posts in the Center of Happiness series
Stages of a customer support team
You’re swamped, but it doesn’t have to be this way. Carving out time to think beyond the barrage of tickets, and being smart about your priorities, can turn things around.
You can’t firefight forever. It’s time to get some fundamental tools and processes in place, so you can give consistently brilliant support to your customers.
Your team is firing on all cylinders and your customers are delighted. How do you take things to the next level?
It may be tempting to offer support through many channels, but do your customers really care? Learn how trimming the number of channels you offer can help improve service levels.
Customer Support sometimes doesn’t get the recognition it deserves, but a few easy tactics can help you and your team become superheroes in the eyes of the company.
Numbers might not be everyone’s idea of fun, but they’re critical for guiding your team and keeping the rest of the company in the loop.
Finding, recruiting and interviewing quality candidates can be a daunting experience at the best of times, but stick to the fundamentals and you’ll have a great fit in no time.
Providing great support is hard, so it’s really important that managers understand how to motivate their team to keep them happy and productive.
Move beyond a simple knowledge base or FAQ and reap the benefits of a channel that’s efficient, functional, relevant and open 24/7.
As you bring new talent into your team, get your new hires off to the best possible start by helping them to integrate into your company structure, processes, culture and tone of voice.
Your team knows more about your customers than anyone else. Organize regular time out of the queue for them to work on long-term support projects that will benefit the whole company.