Use Real Emotion with Artificial Intelligence for Positive Customer Experiences
Just yesterday my partner and I hit a snafu: Our bank had not paid our homeowners insurance, resulting in a panic-inducing email titled “your policy has expired.” Adulting in overdrive ensued.
The bank’s call center was a byzantine maze of pre-recorded messages, and it took three calls just to navigate it to the point of talking to a human being. Just when I was contemplating slamming my phone onto the pavement, I finally reached a person. A person who was clearly chagrined that I’d made it through the labyrinth undeterred. What a grump.
Not finding any help there, I then called my insurance company, which connected me directly to a person — a real, live person! — who cheerfully told me she’d contact my bank, sort out the mess, and call me back. And she did.
It was glorious.
This, friends, is why customer service, and in particular automation, has earned such a loathsome reputation.
Customers don’t want to be pitched from bot to bot, like projectiles in a pinball machine.
That doesn’t make us feel like valued customers. That doesn’t make us want to work with the company again, if we have any other choice. And forget about recommending the company to anyone else (at least, anyone we like).
But what if we could change that paradigm? What if we could create automation that was intelligent enough to give us the answers we need, and send us, quickly and efficiently, to the very best human agent capable of solving our problems?
This is the future I see as imminently possible, at least if we use automation intelligently to create more positive, relevant, and enjoyable user experiences.
Forget bots for a moment — let’s talk about people
For automation to be an integral, genuinely helpful, part of customer support (and customer success — we’ll get there), the customer support process needs to be grounded in a basic understanding of what humans need to be happy — and what customers need to be successful.
The first thing to know is: Every problem is emotional.
We tend to take people at their word. They tell us the problem; we logically try to fix it. But, whatever they say the problem is, and however logical the solution, there is always an emotional component. We’re human; emotions are part of everything we do.
When neuroscientist Antonio Damasio studied people who had sustained brain damage to areas of the brain that generate emotions, he found that the subjects were unable to make even the smallest of decisions. Their logic and reasoning abilities were fully functional, but if they were asked to choose between pasta and risotto for dinner, they couldn’t do it. They couldn’t feel one way or another.
The conclusion: Almost every decision is an emotional one.
What this means for customer service is huge:
If your customer service interaction produces positive emotions, you have the power to generate positive decisions.
Think: making sales, upsells, generating referral traffic — you basically turn customer service into a marketing, sales and retention engine.
All of that potential income is what’s at stake in every customer interaction.
Not sure emotion holds that much purchase power?
A study out of Missouri University of Science and Technology reported that “consumers’ emotional responses” while on e-commerce websites were predictive of purchases. It might seem obvious, but they essentially proved that we buy from stores we enjoy. And there’s no better place to create a joyful experience than customer service.
“Positive experience is the start of a positive association, which builds upon itself over time. One transaction or interchange turns into a relationship. Zappos, Wistia, and MailChimp are three companies that have a business approach which accentuates the positive, and, as a result, their customers are both passionate and loyal.” — Walter Chen, co-founder of iDoneThis, for Kissmetrics
Eliminate Pain Points
So how do you create positive emotional experiences? First of all, don’t add to the customer’s pain by forcing them to run the gauntlet of automated options they neither need nor want.
Pain is emotional, and reducing the pain your customers feel will go a long way towards creating a positive experience. Just think how happy I was to find a HUMAN BEING on my first try with my insurance company!
The worst pain is caused by a-thousand-cuts annoyances, and when you can relieve those small irritations, the customer’s experience will be more positive — and studies show those positive experiences are directly linked to customer loyalty and repeat purchases.
That isn’t to say that you should do away with automation altogether. Most companies need automation of some sort to economically scale customer service. If FAQs can be handled by a bot, pop-up,or FAQ page, it makes no sense to take up the valuable time of an agent.
Or, at least, that’s the prevailing idea.
But most of these customers aren’t coming to you because they have easily-answered, “dumb” questions. This frame of mind makes customer interactions into inconveniences. But they’re really opportunities.
They’re opportunities to maximize lifetime value and customer loyalty.
They’re opportunities to gain referrals and offset cost-to-acquire expenses.
They’re opportunities to upsell, cross-sell, and basically GROW.
Do you really want to hand all of that over to a robot?
If so, you’d better do it better than my bank does (and many other businesses).
One significant improvement you can make is to get customers to the answers they need faster — which brings us to our next positivity-generator:
Most psychological models agree with Freud: Humans act on the “pleasure principle,” the instinct to gratify desires and avoid pain. When we don’t get what we want, we feel bad — a negative outcome we’ll go to great lengths to avoid.
And, the longer it takes to get what we want — especially if we expect to get what we want instantly — the more aggravated, frustrated, and angry we become.
Millennials, in particular, have grown up with mobile devices, high-speed internet and instant answers. Their tolerance for poor customer service is low, and thanks to social media, they’re vocal about it. Interestingly, the Millennial culture is so pervasive that their expectations have infiltrated every other generation.
In Nuance’s whitepaper The Millennialization of Customer Service, their past research on customer service frustrations was divided along generational lines — Millennials, Gen Xers and Boomers “disagreed on preferred channels and their expectations varied.” In just a year, they report “that gap has significantly shrunk. The research now calls for an intuitive, consistent experience — across channels, interactions and generations.”
“Customer service has overtaken price and convenience as the greater influence on consumer buying decisions. As Millennial expectations for streamlined experiences become the new normal, the emphasis on service continues to rise. Eighty-seven percent of consumers report that a company’s customer service has a significant impact on their decision to do business with them.” — Nuance, The Millennialization of Customer Service
Everyone wants what they want now.
When you’re looking at delivering positive experiences (and generating more conversions), the ability to feed that need can really help — like with an instant chat feature.
“On Twitter, over 80% of customer service related tweets are negative or critical of the brand in question, a large majority of customers expect a reply in less than an hour (unfortunately, most companies are failing to deliver)” — Touch Agency Twitter Facts and Figures
This is where automation can actually come in and save the day — you know, instead of ruining it.
When a user is on your website looking at a product, and wondering “will this work for me?” — you want to be available to answer that question right away. When a user is on your social media, hash tagging complaints about your brand in front of EVERYBODY, you want to be able to react instantly.
“Twenty-seven percent of Millennials have written a negative review online, compared to sixteen percent of Baby Boomers.” — Nuance, The Millennialization of Customer Service
The only way to deliver fast responses at scale (without hiring hundreds of full-time agents) is through automation.
Automation done wrong (the status quo)
When automation takes the place of human interaction, that’s when frustrations among customers rise.
When automation moves the customer from one useless option to the next, it worsens.
But automation isn’t the only problem with how customer service typically happens. Too frequently, customers are passed from person to person, waiting on helplines between each representative, giving the same information to people over and over before finally being passed to the one person who is qualified to help (to whom they have to explain the situation yet again).
All of this is how customer service has earned its bad reputation.
Automation done right
Automation is at its best when it works to efficiently connect people to people, responding in context and even creating some context along the way. It should act as a facilitator of communication.
Ideally, automation would collect pertinent information and pass it to the person best qualified to handle the issue — so the customer doesn’t have to repeat themselves five times. It might even recognize the customer and deliver even more information to the agent, creating some context and paving the way for a more proactive approach to service.
The agent might, for example, notice what the customer has purchased in the past, or see the number of times they’ve logged into the software, prompting them to ask not only how they can solve the specific problem the customer has now, but see ways in which the customer might benefit from additional help, services, or upgrades to reach their next “success milestone” (a customer success term for making progress towards the customer’s desired outcome).
If this sounds more like a Customer Success approach than a Customer Service approach — it is.
Of course, this sequence depends on a customer coming to you with a problem.
But you can also reach out to prospects on your website with automation to warm them up.
An instant chat feature on a website, for example, might start off as automated, “Hey there, can I help you with anything?” And if the user responds, move the conversation over to a person. That’s how it’s used most often, and it’s a good system.
But there are some improvements to be made.
See how this company uses their automated chat window — it pops up after a few seconds of viewing any page on their website.
Notice the title of the blog post — and then look at the chat greeting to the right.
Kevin’s greeting is completely relevant to the topic of the page I’m reading. And, it’s not overtly trying to sell me anything or get me to opt-in to anything (they have several other pop-ups timed to do that); it’s just offering advice in a topic I’m clearly already interested in learning more about.
Now, did Kevin sit down and type that just for me? No!
But it feels like he did.
If I responded to Kevin and our conversation went well, several things would happen: They’d keep me, the prospect, on the site (rather than continuing my research elsewhere), and I’d feel grateful for the personal attention. That gratitude could influence me to make the purchase once I’m done chatting — it’s Cialdini’s Reciprocity principle in action.
This is automation with a human touch, a trend that companies like Freshdesk are spearheading.
Pit Crew Style
Freshdesk proposes a customer support model that uses real people and automation to act like a “pit crew” for each customer interaction — with an emphasis on everyone on the crew getting pertinent information. Rather than pinging from person to person, or rolling through an assembly line-style customer service routine, the customer moves through something like this:
The customer has automated answers to choose from, or can provide context and go directly to the right agent. Instead of ending the interaction there, additional context (brought to the agent through the wonders of automation and collaborative tools) allows the agent to personalize the interaction and find opportunities to help the customer even more.
The customer gets what they need faster — eliminating pain points and delivering that all-important instant gratification — and feels valued. Most importantly, by the end of this customer service experience, that customer will feel even closer to reaching their desired outcome (whatever they hope to achieve with the product).
Even though this approach uses big data, machine learning, bots, collaborative tools, and natural language processing (NLP), they work together to “place people over processes.” Exactly what automation should do.
And, this approach works even better in-app.
In-App Customer Support — More Immediate, More Relevant, More Appreciated
“A Freshdesk customer reported (anecdotal) a 60% reduction in call volumes from their app users (with messaging) when compared to users on their website. 5–12% of transacting users in commerce apps chat with the business for various queries. We have even seen a 50% drop in negative conversations on social media for a big customer, thanks to in-app chat.” — Freshdesk
In-app automated messages relieve some of the worst pain points customers have: Interruptions and wasted time. Having to call or contact customer service via phone, or even chat or email, requires the customer to stop the work they’re doing.
And it takes an average of 25 minutes to return to the original task after each interruption.
But when a relevant message arrives in the software the customer is already using, distractions and frustrations are minimized. When those messages are pertinent to the task, they’ll even improve productivity.
Of course, coordinating the timing on those automated messages involves tracking events that trigger interactions when needed.
In the best customer success platforms, agents will see “red flags” when customers fail to perform tasks that will lead them to their desired outcomes. They also receive alerts when customers reach success milestones, or would benefit from an upgraded account due to growth or usage. If a customer stays on a page for a longer-than-average time, it may trigger a chat pop-up window asking if an agent can help.
Whether you’re using this technology for success or support or both, the key is to understand what your users are doing, and what actions typically are associated with confusion, problems, and successes.
It’s a proactive approach to customer support that, yes, is a lot closer to customer success. But really, isn’t the best support the kind that leads to success?
What All of This Has to do with Growth
As I’ve said, every customer interaction is an opportunity. But when you’re proposing the idea of beefing up your customer support (or customer success) department to the boss, they don’t want theory, they want numbers.
Calculating the ROI of Happy Customers
“We used this data to calculate the revenue impact of CX across 20 industries. We found that a moderate increase in CX generates an average revenue increase of $823 million over three years for a company with $1 billion in annual revenues.” — Temkin Group, ROI of Customer Experience 2016
How can you measure how well you’re supporting your customers? Try these metrics.
Lifetime Value (LTV) is the value a business derives over time from a single customer. Typically, the equation to find the LTV of a customer looks like this:
LTV = Revenue earned from customer within period defined by average churn rate
But it’s much more complicated than that. That “revenue” has to include not only purchases/subscriptions/renewals, but also upsells, cross-sells, and the value of referrals. Even if you have those numbers, there’s still the far more difficult to track ROI of positive reviews.
One way to track that is to employ a sentiment analysis tool, like Clarabridge or Brainjuicer, to track positive/negative sentiments across social media, and cross-check those trends against conversion rates.
Churn Rate is how many customers you lose in a given period of time. The equation looks like this:
Churn Rate = # of customers lost in period / # of total customers at start of period
You may also want to track customer churn in relation to the timing and frequency of their interactions with customer support. If people are dropping out just after their support experiences, that could indicate a problem! But, if you can show a correlation between customer support and retention, that data can prove the worth of your program.
Cost-to-Acquire (CAC) may be the piece de resistance, however. If you can get the data on product cost, research & development, and advertising — everything you need to convince a potential customer to buy — and compare that to the average lifetime value of your customers, you’ll be able to see the effect of all of your efforts to improve retention and generate referrals.
That equation looks like this:
CAC = product cost + research & development + marketing/ads / total new customers within a specific time period
And, don’t forget, investing in automation can also save your company money. As Freshdesk puts it:
“Bots can play a big role in boosting productivity and reducing agent count for companies by handling simple scenarios, engaging with users to fetch contextual information, or even understanding user intent and messaging them to buy time for a human agent to research and respond. We believe bots will assist human agents in customer support, not replace them, and hence help businesses scale their customer support while maintaining a positive experience for customers.
Is there money to be made? Yes! If you increase agent productivity by 20%, you are indirectly saving on customer support agent salaries, not just seats on a support product. Businesses will pay you for that.”
Each of these metrics is an insight into the health of your company overall, and an indicator for how well you’re serving the needs (emotional and otherwise) of your customers. And, they’re all tied to customer happiness, customer success, and customers spreading the word about your company.
Food for Thought
Automation — used intelligently and sensitively — is the way forward for growing companies to scale personalized service, deliver answers faster, and eliminate the most pervasive pain points associated with asking a question. But the goal of automation is much smaller, much more personal than that. It’s about creating genuine connections between people, with a little help from the machine.
Your Cheat Sheet for Effective Automation and Impressive Service
- Every problem is emotional
- Decisions are also emotional
- If your customer service interaction produces positive emotions, you have the power to generate positive decisions
- Every customer service interaction is an opportunity to increase lifetime value, customer loyalty, referrals, upsells, cross-sells, and growth
- Reducing pain points will make customers more likely to become repeat purchasers
- Reduce the amount of time it takes a customer to get the right answer by using a combination of automated information and access to qualified support agents
- Reduce distractions by offering in-app support, and support triggered by “red flag” events
- Adopt an internal communication tool that sends pertinent customer information to any agent they speak to, so they don’t have to repeat themselves
- Instant gratification is what customers expect; delivering it will make them love you.
- Give the customer self-service access to the information they need immediately, and if they still don’t’ find it, move them to a human fast
- Move from reactive to proactive support by tracking user behavior and alerting an agent if signs of confusion or trouble occur
- Automation can go very wrong if it blocks customers from talking to people
- Automation works best when it efficiently connects people
- By providing relevant information to the agent about the customer, like what the customer has purchased in the past, number of times they’ve logged into the software, whether they’ve reached “success milestones,” etc.
- By “popping up” with a relevant message to website visitors
- By appearing “in-app” when triggered by specific user events, alerting agents to “red flag” behaviors that may indicate a problem, allowing for a more proactive approach to customer service (that’s more like customer success)
- Calculate the ROI of your customer support program so you can prove to decision-makers that improving and optimizing your processes matters — a lot!