What if I told you that your contact center had its own collective consciousness? As strange as it sounds, it’s true. At least in a certain sense.
Anytime we pull up a big-picture report of average CSAT scores or employee morale, we’re looking at the minds of all our customers or all our agents; we’re getting a peek into their thoughts and feelings, as an aggregate entity. And where there’s consciousness, there’s subconsciousness. Bubbling way down there, deep beneath the surface, are all those hopes and fears and desires of your customers, which even they themselves may not be aware of.
If you, as an individual, wanted to grasp the deepest truths about yourself after deciding to work towards becoming a better version of yourself, you might go to a psychotherapist for counseling. If you’re a contact center leader looking to do the same? You might go to AI.
That’s what I want to talk about for part three of our “Customer Experience Differentiated” mini-series: How sentiment analysis, fueled by AI, can help you tap into that collective customer subconscious, allowing you to better understand their needs, and ultimately, improve satisfaction and retention.
Digging a little deeper
As we’ve seen throughout this series, the ability to gather quality, relevant feedback, and to respond to that feedback appropriately, is absolutely critical. A customer who has a poor experience only has a 43% chance of retention for the following year, so you’d better be able to know when they do have a sub-par experience, and understand why.
Surveys and direct feedback are great for this, of course. They provide invaluable insights into an individual customer or agent’s response to an interaction. But even if they’re being completely honest and thoughtful in their feedback — which we all know isn’t always a given — what you’re capturing in a survey is their conscious response. As any good therapist will tell you, if you want to see the whole picture, you need to capture what’s going on with both the conscious and the subconscious.
That’s where AI and sentiment analysis enter the room.
What does sentiment analysis really do?
Sentiment analysis has been around for a while now, but I still know plenty of contact center folks who might view it with a dose of skepticism. After all, “sentiment” is an extremely soft and mushy concept. But while that may be the case, the fact remains that relationships are the key to customer loyalty, and relationships boil down to feeling and emotion more than cold calculation.
The most revealing feelings — and therefore the most useful — tend to be those that are buried the deepest. What sentiment analysis does is dig them up, aggregate them and refine them into something we can work with.
To put it more concretely, sentiment analysis is a technique to analyze linguistic artifacts — in our world, a whole bunch of aggregated customer conversations — using a form of machine learning called natural language processing (NLP) to identify emotional undercurrents and trends. At the simplest level, this usually determines whether the content expresses a positive, negative or neutral sentiment.
Sentiment analysis takes those raw conversations (which hopefully, as we discussed last week, you’re transcribing automatically) and surfaces the patterns of diction, tone and syntax most indicative of your customers’ emotional reactions. Then AI processes them into valuable insights at scale via information extraction, theme classification, and emotion and intention analysis.
The point of all this? To identify trendlines and inform strategic decisions that improve CX delivery and minimize customer churn.
The path to CX self-actualization
Just like conversational AI and agent assist have become vital parts of optimizing the pre- and mid-interaction phases, respectively, sentiment analysis has become an invaluable tool for the post-interaction phase.
Customer conversations are a gold mine of valuable information, data and insights. The gold’s down there; you just need to go extract it. With sentiment analysis, you can.
You can get a much deeper understanding of your customer’s needs, problems, opinions, motivations, preferences and perceptions of your brand — because unlike with surveys, limitations on time or attention span aren’t a factor. You can provide real actionable insights — for example, by identifying which words, topics or decisions tend to trigger positive or negative reactions — to better inform both your agents and your other AI systems, such as agent assist.
Though the idea of “sentiment” may still feel squishy, the results of sentiment analysis are anything but. After all, the cost to acquire a new customer versus the cost to retain a customer is significant — usually 3:1, and in some industries, as high as 10:1. By making better strategic decisions about how to talk to customers, informed by a deeper emotional understanding, you can build stickier relationships and make massive, measurable gains in retention.
For example, on this week’s episode we touched on a certain client that was experiencing a churn issue. After implementing sentiment analysis, they uncovered a range of surprising facts about their customers’ collective emotional response that shifted their overall approach.
Most notably, they found that encouraging agents to stay on the phone with customers as long as the customer might want — rather than minimizing call times, as popular wisdom often suggests — created so much more positive sentiment that gains in customer retention far outstripped the drawbacks of slightly higher call volume or average handle time.
Here’s a chart they shared comparing their improved monthly churn rate (red) to the industry standard (teal):
The lesson? If you harness sentiment analysis to find, extract and process customers’ aggregate emotional response — then make smart decisions regarding what to do about it — the payoff will be not only tangible, but impactful.
The 3 key takeaways
As we’ve seen, we can leverage sentiment analysis the way we might lean on a therapist or psychiatrist in our personal lives: To help us see what’s really going on under the hood.
With contact centers, as with therapy, self-improvement is never going to just solve itself. But with AI and sentiment analysis, you have the tools you need to put together a kind of behavioral therapy plan for the contact center.
What’s happening in these conversations to make customers feel happy or frustrated? Where does the tone change? Where, based on their emotional state, are there opportunities to upsell, cross-sell or mitigate a negative response? Is there a strategic change you can make to rethink how you gain and create value from better connecting with customers or differentiate your CX? Here are some high-level takeaways from your first therapy session:
- Your contact center is conscious; therefore CX results are driven by its subconscious!
- You can use AI and sentiment analysis as a kind of psychoanalyst to unlock a deeper customer understanding that informs a more successful CX approach.
- The use of AI in the CCaaS cloud can create a competitive advantage that’s unique to your organization’s USP.
For more insights, watch the full recent LinkedIn live stream episode, “Your Contact Center’s Subconscious Made Conscious: AI as Your CX Psychotherapist.”
For discussion of similar topics, tune in for the “Customer Experience in the Cloud” live stream series with Valur Svansson, every Wednesday at 9:30 a.m. CT on the Lifesize LinkedIn page. To watch past episodes on-demand, visit our YouTube channel.