The chatbots of the past had limited capabilities. They could answer a predetermined set of questions, but no more—which often left customers frustrated. But conversational artificial intelligence (AI) changed all that.
Conversational AI gives bots a greater ability to understand human language, sentiment, and intention. AI-enhanced bots can customize messaging based on real-time data and complete complex tasks and transactions.
Today, conversational AI is the perfect addition to customer service teams, giving companies a competitive advantage by increasing agents’ productivity, delivering a better customer experience, and increasing customer satisfaction.
Want to add conversational AI to your customer communications but don’t know where to start?
On the Let’s Talk CX podcast with SKWeston, I sat down to discuss the impact conversational AI has on customer satisfaction and agent productivity, along with the best practices I’ve identified after being in the customer experience field for over fifteen years. Of course, understanding the trends and technologies that will deliver exceptional customer experiences is key. And that’s where our partner, SKWeston & Company, provides guidance on identifying the best path for our clients’ unique needs. They’ve identified and implemented innovative solutions in areas such as:
- Conversational AI platforms
- Omnichannel knowledge management system
- A reimagined BPO with unbound agents and technology
- …and many more such a RPA, blockchain, etc.
What are some best practices to increase survey response rates?
Ask the right questions.
First off, determine which type of survey applies to your business. For example, is it customer effort, customer satisfaction (CSAT), or Net Promoter Score (NPS)? Each survey measures something different, so deciding which measurements are essential to your company will determine the type of survey to use and which questions to include.
Keep timing in mind.
The next thing to keep in mind is timing. Let’s use CSAT as an example. In the past, it was standard practice to conduct CSAT surveys once every six months. In a world where your customer can find another product or service with a click, you can’t afford to wait that long to know if they’re happy.
I believe brands should immediately trigger a survey at the end of a customer conversation to make sure the customer got what they wanted. And because it’s just an extension of an exchange the customer already had, they’re more likely to respond and complete the survey.
Trigger the survey in-channel.
Make sure you ask for feedback in the channel where the interaction took place. So no matter which channel the message is coming through, SMS, web chat, Apple Messages for Business, etc., make sure that you’re asking those CSAT questions in the channel immediately after the interaction when it’s fresh in the consumer’s mind, and they’ll be more likely to participate.
Aim for more than just a response.
Completion rates are just as important as response rates. Asking your customers too many questions can lead to a lower completion rate, which is almost as bad as having them not respond at all. Stick to between 3 and 5 questions, and make them brief and meaningful.
Test, test, test.
I always recommend A/B testing. You can test images, topics, tone of voice, number of questions, length of questions, and even incentives. Using the results can help you refine your surveys for higher participation.
The other thing I recommend is to offer incentives or promotions to garner a response. The National Center for Biotechnology Information (NCBI) found that achieving adequate survey response rates is an ongoing challenge. Their study found that monetary incentives increase response rates. Examples of effective incentives I see organizations offer are promotions or coupons, such as “Complete this survey and get 10% off your next order.”
Keep it conversational.
Conversational AI lets you use different flows and responses according to how the consumer interacts with your survey—taking it from boring to a more engaging way to communicate with customers. Having insight into who your customer is, what products they’ve purchased, and the next best product for them helps contextualize the conversation and allows you to customize the CSAT survey to drive a higher response rate with higher satisfaction.
How can you improve your CSAT score, and how does conversational AI help?
CSAT is all about the customer being able to get what they need in that moment of truth. So when it comes to conversational AI, keep your automation practical and data-driven to the fullest extent possible. With companies like Quiq, you can enable your automation to find the answers consumers want and be consistent across all your messaging channels.
Start with what you know.
Conversational AI chatbots can answer common questions and set customer expectations. When building your automation, use the 80/20 Principle: 80% of the answers you give out on a day-to-day basis can usually be answered by 20% of the knowledge. So start with what you know, don’t try to boil the ocean.
Asynchronous communication in your customers’ channel of choice.
Make sure you’re communicating through your customer’s channel of choice and keep it asynchronous. Customers are so busy these days. You and I and everyone else are constantly moving, and it’s tough for us to get on the phone and have a conversation that lasts 15-20 minutes. So having asynchronous channels where you’re communicating at your consumer’s speed of life is a very convenient way to communicate. And it will typically generate a 15-point higher CSAT score than traditional channels.
A Forrester report commissioned by Google shows that:
- 65% of consumers find asynchronous communication more convenient
- 61% of consumers prefer asynchronous communication
- 60% of consumers trust asynchronous communication
- 56% of consumers are already using asynchronous messaging for other purposes
- 54% of agents are more efficient when using asynchronous chat
Take a survey response from negative to positive.
An article by the Harvard Business Review shows that engaging with an angry customer to resolve issues can turn unhappy customers into loyal ones.
One benefit of triggering a survey at the end of customer interaction is that if a customer responds negatively, you can reopen that conversation with an escalation path to someone who can immediately address the issue. So you can ensure that the customer, who may have started with an unpleasant experience, ends with a positive experience.
Make agent access easy.
You always have to make sure that you give access to agents in case your automation can’t help. Consumers need to know when they are working with a bot that there’s an easy and very obvious way to get to an agent if they need help.
The bot automation should include a warm introduction with instructions like, “If you need to speak to a human, just type the word agent or hit the button at the bottom, and you’ll immediately be transferred to the next available agent.”
Quiq’s platform combines conversational bots and the channels with the agent desktop—so it’s a smooth transition between bots and humans, humans and bots. As a result, we often see conversations move back and forth between humans and bots seamlessly.
Support agents with AI.
AI chatbots can help agents do their job better and provide a greater customer experience. I strongly believe it’s all about empowering the agent to support customers in the most efficient and effective way possible. And sometimes, that means automating negative-value redundant questions that enable the agent to handle more valuable interactions, like presale conversations.
I often see agents using bots in the middle of a conversation to help with redundant activities, like data gathering in an RMA process. And so, the ability to transfer from an agent to a bot and then back to an agent allows you to do more with less and keeps your agents focused on the most valuable activities.
How can serious executives unlock customer metrics?
Conversational design is an art and a science that requires good data. So one thing to look for is a conversational platform with a robust set of metrics and analytics that cross not only the bot but the human conversation, including:
- Visual analysis of where customers navigate to, where they gain success, and where they abandon
- KPIs related to customer satisfaction, such as customer intent and sentiment
- Real-time and historical understanding of agent performance
- Average and first response time for individual conversations aggregated over an agent’s shift
- The ability to search past conversations by the customer, queue, agent, date, time, channel, and more
- APIs that export conversations and blend them with other business data for better decision making
Executives can take that data and push it into business intelligence tools to inform decisions, including how to design conversational AI to meet their company’s and customers’ unique needs.
Executives can also save time by partnering with companies such as Quiq that own both the bot and the human conversation, so a 360° view of the entire interaction can be had. All that being in a single platform with solid reporting and analytic data so they can make real-time decisions on how their customer journeys are operating.