Posted on Harvard Business Review: July 13, 2010 7:58 AM
Next time you're on the phone with a call center, listen carefully to what the rep says. Chances are you'll hear your name several times, hear a tone of empathy, maybe an "I'm sorry." It would be nice to think the rep really cares — but of course she's probably just following a script. That can be a bad idea, we've found. In our recent HBR article "Stop Trying to Delight Your Customers", we explored how customer service drives loyalty, including the role of managing the emotional side of customer interactions. Here's some further insight about that delicate dance.
Most companies still suffer from the checklist mentality when it comes to managing how their reps engage with customers. Use the standard greeting...check...say the customer's name three times...check...show empathy...check...ask if you've fully resolved the issue...check, check, and check.
Most companies will tell you it's all about consistency. But, let's face it, consistency breeds robotic interactions which fail to result in a tailored, low-effort customer experience.
We've seen companies move away from this "one-size-fits-all" approach and creatively teach their reps to use simple word choice — and in some cases, approaches founded on behavioral economics —to radically shape how a customer perceives an interaction.
Take Osram Sylvania for example, a case we discuss in the article. They teach reps to simply avoid negative language (e.g., "can't," "won't," and the show-stopping "that's our policy") in their most common service interactions. This has helped them net a Customer Effort Score that is 18.5% below industry peers (the less work a customer must do to get a problem resolved, the lower the customer effort score. The lower the score, we've found, the greater their loyalty).
Recently, we ran a series of experiments across two separate groups of customers to better understand the impact word choice can have on a customer interaction:
In one experiment, the rep had to authorize a customer banking account before the customer could transfer funds. But the rep explaining "you can't transfer funds until you go through these steps to authorize the account" scored significantly lower than the rep explaining "let me walk you through these steps to authorize the account." While the language is subtly different, customers rated the latter as 82% higher quality and 73% lower effort.
In another experiment, customers were told they had to bring their new bicycles to a certified repair shop. The performance of the rep who simply stated "you're best off bringing it into a repair shop" was rated significantly lower than that of the rep who noted that they'd "pass the customer's feedback to the engineering department," "check the database to see if a simple fix is possible," and "recommend the customer bring the bicycle to the shop." The latter scored 67% higher quality and 77% lower customer effort.
Such approaches go well beyond traditional soft skills. Instead, these rely on careful language choice to frame answers in the best possible way. This isn't simply being empathetic — it's calculated and anticipatory. We call it experience engineering.
Beyond simple word choice, we've seen other experience engineering approaches work well. For instance, LoyaltyOne (also referenced in the article) practices an idea called alternative positioning. This approach is premised on learning some basic information about a customer during the interaction, and then using that information to reframe a not-so-great option as an acceptable option. The company's customer survey scores have improved 15%+ as a result of this practice.
Alternatives positioning isn't revolutionary — in fact, sales reps have been framing product features in light of customer benefits since commercial interactions began. But, applying this method to service scenarios is quite innovative and generally defensible.
But we've also seen some of these techniques get a bit murkier and more ethically questionable. One airline (that shall remain nameless) told us recently that they "caught" some of their best reps using similar techniques to avoid dust-ups with customers over cancelled flights.
For instance, imagine your 11:00 AM flight is cancelled and you need to be in Cleveland tomorrow morning. There's an evening flight that's open. Where most reps would simply say "I can put you on a flight leaving at 9:00pm" other reps, knowing full well the 9:00 PM flight was available but seeking to manipulate the customer's reaction, might say "well, I know I can put you on the 7:00 AM flight tomorrow, but let me see what I can do to put you on the earlier flight, which is at 9:00 PM tonight." This technique of experience engineering is more commonly called anchoring. A less-desirable option creates a mental anchor, making the best alternative seem more acceptable. Rather than be irritated that the 11:00 AM was cancelled, you'd probably be pleased that the rep has secured a seat for you on the evening flight.
Ethically speaking, few companies would have a problem with the idea of removing negative language or getting their reps to display greater advocacy. But does this anchoring example take experience engineering too far? We tested similar approaches in our experiments, and indeed they work, but we believe this walks a very fine ethical line.
Customer manipulation or savvy service? Let us know what you think about using techniques like these in customer interactions and what your own organization's experience has been.