A recently opened restaurant is well into its dinner seating. The owner visits several tables to ask how everything is. One table compliments the décor, another raves about the main course while a third table agrees enthusiastically that they’ll be back again. Back in his office, the owner reflects on his success and high customer satisfaction. But observing their behavior would have told a different story altogether. That same night, the first table posted to social media that the service was poor, the second table ate the main course but left the rest of the plate untouched while the third paid with a coupon and quietly congratulated themselves on a frugal date-night dinner.
Why is there more credibility in what a customer does than in the words they use? It comes down to simple human psychology. The explicit information taken in through surveys and by talking to customers is often shaped by these four things:
- Limits in their motivation (the first table didn’t tell the owner face-to-face about the poor service because they didn’t want to be rude).
- Limits in their opportunity (the frugal couple would have offered up that they mostly came because of the coupon but they weren’t given the chance).
- Limits in their ability (the patron liked the main course but had a hard time adding in but the potatoes were cold).
- Limits in their awareness (the social media poster was actually influenced by the fact that her waiter resembled her ex-boyfriend).
Taking in customer behavior (i.e. where they clicked, when they called, why they left) and turning it into data points allows companies to draw conclusions, free from the influence of the above-described limits. The secret to getting and applying this valuable behavioral information quickly is incorporating predictive analytics comprehensively into your business strategy.
Before predictive analytics, companies could loosely identify a pattern of customer behavior by deep diving into quarterly reports, attrition rates, renewals and purchasing trends. Still, there was a limited degree of certainty about what data said, and even less certainty about what it meant.
Predictive analytics looks at data from customer behaviors across an entire organization (sales, support, marketing etc.) and finds patterns in those behaviors that increase the accuracy and impact of future decisions.
Predictive analytics powered by machine learning increases a company’s ability to engage more intelligently with its customers as the engine is constantly learning, finding patterns and making new predictions every time a customer acts, regardless if you ask for their opinion.
As an example, let’s say someone under the age of 25 goes to the website of a telecom company in search for help on a specific topic. He can’t find any information, so he goes to Twitter to ask if anyone can answer his question. Five minutes later, his question still isn’t answered so he calls support with the intent to cancel.
In this case, the telecom company uses predictive analytics, which picked up on a pattern showing that millennial-aged customers phoning in billing issues and not getting resolution within a certain number of minutes are cancelling service. They used this information to create a plan to improve resolution time targeting that specific demographic.
Because of predictive analytics, the support department now has a script that speaks directly to millennial-age customers, making them feel valued and important. Using machine learning, they can also match incoming tickets with the best agent. So, when this 25 year-old calls in, his call is automatically directed to another 25 year-old in the department who knows the script, can identify with him, feel his pain, resolve his issue, and keep him as a customer. All before the customer can voice a complaint.
In this world, the most helpful information we receive is from unbiased sources. If you want to know how those glasses look on you, don’t ask your mother. If you want to be better, don’t ask your customers what they think—just watch what they do.
Find out how you can leverage machine learning to learn what your customers really think.