Predictive Analytics: A Crystal Ball Approach to Customer Satisfaction
As customer service moves online, our desire for quick resolution continues to rise. Regardless of the complexity, we want our issues resolved instantaneously.
How can online customer service providers keep up with that kind of demand?
Wouldn’t it be grand if we had a crystal ball to look in that would not only tell us when customers were going to reach out but what they were going to say?
Not superstitious enough for crystal balls? Ok. Today, we’re going to look at how predictive analytics (PA) just might give us the info we’re looking for.
Predicting Demand and Allocating Resources Appropriately
For much of my young adult life, I worked in restaurants as a waiter and bartender. One of them was a big, corporate outfit known for its 250-item menu and its daring forays into cheesecake manufacturing.
Corporate restaurants dance on a razor’s edge when it comes to profits. For that reason, staffing is literally a minute-by-minute concern. Working with various corporate models, protocols, and a splash of gut instinct, our managers worked hard to ensure that at any given moment we had the exactly correct number of cooks and waiters in house—no more, no less.
According to Stephen Timms, President of ClickSoftware, predictive analytics may provide the same kind of staffing efficiency to your customer service operation.
The example Timms uses is an HVAC technician whose predictive analytics enable him to keep his truck perfectly stocked. Using PA, he’s able to allocate precious cargo space only for the parts he’ll need, and nothing else.
The solution need not apply only to field professionals, however.
Imagine if you could pick up on the trends in your customer requests to determine when and where you’ll need the most support. You could then adjust your staffing levels appropriately. Not only would your hold time decrease—or response time via email or social media—but your CSAT would likely go through the roof.
With PA, you can make sure someone is always ready to respond to customers without paying your people to sit around when they’re not needed.
Answering People’s Questions Before They Have Them
Having the right people in place at the right time is great, but it does little more than fulfill that baseline expectation: a quick resolution for every problem. Taking it a step further, how could PA help us to exceed that expectation?
Having worked on well over a hundred contracts in my career as a real estate agent, I learned quite a bit about negotiating. On the front end of every negotiation, I would lay out the likely course the back-and-forth transaction would take.
Except in a few instances, I was always right on the money.
My clients perceived this as an uncanny ability to read minds and tell the future. I assure you, I had no such paranormal skill. What I did have was a whole lot of experience. I had negotiated enough times before to know how any given scenario would play out.
Once again, we turn to PA. With enough data—computer-speak for ‘experience’—a decent PA algorithm can parse through all of your previous customer interactions, plot repeating issues, and predict future service needs.
Nathan Gnanasambandam, a Senior Research Scientist at Xerox, explains how his company has been using PA to analyze past healthcare outcomes to predict future healthcare needs:
“Early pilots show that we can use predictive analytics to predict issues or questions patients will most likely have and proactively provide the information in a personalized fashion before the patient even has the issue.”
Xerox is still in the early stages, but the promise here is that one day a doctor will be able to call and tell you what you need to know before you start experiencing symptoms. Pretty cool, huh?
Imagine: what if you had a customer service representative dedicated to reaching out to customers and solving their problems before they reached out to you?
“Hey, we know you like to use our product in this way. Here’s a problem you’re probably going to run into and a few ways to avoid it. In case you need them, here are few more ways to solve the problem.”
If I’m that customer, then not only am I now thoroughly impressed with your foresight and initiative, but my experience with your product just got that much better. That’s the kind of experience customers like to share with their friends.
Artificial intelligence may not be a crystal ball, but advances in predictive analytics are bringing us pretty close to telling the future. Using PA, you can get proactive about how you serve your customers. If you’re looking for a decisive way to edge out your competition, then this is it.