The dangers of tying compensation to customer feedback

If you are really serious about creating a culture of customer advocacy, then you will certainly tie incentive compensation to customer feedback.  Right?

Wrong.

Inherent risks and unintended consequences

Anyone who has ever experimented with incentive compensation, knows it can be a powerful way to create organizational focus.  Done right, incentive compensation can reinforce the company's values and motivate specific actions designed to meet important goals.  Done wrong, the unintended consequences can be devastating.  Some of the most important risks: 

• Extreme focus on the score:  With incentive compensation, the saying goes, you get exactly (and only) what you pay for.  If you give your team goals based on scores, they will focus on those scores.  How will this manifest itself?
  • They may criticize the data collection methodology and score calculation.  
  • They will probably spend way too much time explaining and rationalizing the scores.  You will hear things like "problems with our vendors impacted our customers and depressed our scores" or "our score was low last month because it rained a lot and made the customers depressed."
  • They will complain that the scores measure things outside of their control, and therefore the feedback collection methodology should be modified to focus on things only under their direct control
• Gaming:  The logical extension of extreme focus on the scores is active gaming -- cheating.  This takes several forms.  The two most common:
  • Selective surveying.  In some companies, front line employees can "bury" customers who they know will give negative feedback by recording inaccurate contact info or flagging them as having opted out of surveys.  Similarly, they may explicitly ask only happy customers to answer the survey they will get.  At one retailer, I noticed a clerk would use a Sharpie pen to circle the survey invitation at the bottom of the receipt, look the customer in the eye, and say "It would really help me out if you would fill out our survey.  If you mention me by name, I'll get an award."  She only did this for a subset of the customers, however.  Guess which ones.
  • Misguided attempt to influence the SCORE instead of earning a customer's loyalty (click to enlarge)Begging for scores.  The classic example of this is the car dealership service department that posts a huge sign next to the cashier showing the satisfaction survey with a big bold circle around the top score.  Or the car salesman who, after what was an otherwise good experience, pulls out the survey and tells you that if you don't give him a 10 on these two key questions, he will lose his top status and his kids will starve.  (For a great example running all throughout a company, see the related post: Begging for scores at Hilton.)
• Disaffection and disengagement:  If you don't have a stable, reliable metric before you begin working it into your compensation system, get ready for a real firestorm.  Two major issues can create this sort of situation:
  • Score instability.  Problems with sample selection, changes in the mix of customers by segment or touchpoint or product, differences in response rates caused by survey methodology changes, or even inadequate sample sizes can create instability or even plain unreliable scores. 
  • Inadequate or immature understanding of loyalty drivers.  Early in a typical closed-loop feedback process, front line employees, supervisors and the management team struggle to understand what creates positive and negative changes in scores.  Frankly, it's pretty unusual for employees to trust the scores early in the life of a new metric system.  Until your team has some sense of control and influence over the outcomes, a compensation system based on a new score may seem unfair or arbitrary.  Result:  anger, finger-pointing, defensiveness.  Anything except the intended focus on the fundamental goal of creating more Promoters and fewer Detractors.

Conclusions

Please don't misinterpret my objective or the implications here.  I believe most companies should ultimately work customer feedback scores into incentive compensation systems.  Based on our experience at Bain and with the NPS Loyalty Forum, we have found that if the senior-most executives are explicitly and publicly on the hook for improvement, they pay more attention to the issues.  It also sends a great message to the organization about what the company values.  If not, lots of important opportunities will be missed and the company's efforts may even be at real risk of failure.

Nevertheless, Bain clients and others have had experiences that demonstrate the need for caution.  Before including any form of customer feedback scores in compensation, especially for front line employees -- the ones interacting daily with your customers -- you need a stable, reliable metric that people understand and trust.  Moreover, gaming is serious stuff, and so you need protections against it.  It erodes customer trust and ultimately undermines the entire intent of your efforts to improve customer loyalty using closed-loop customer feedback.

Next time: Some constructive advice about HOW to use feedback scores in incentive compensation

Comments

[Oliver Bendzsa]

Hi Rob:

Excellent post on the many issues of tying compensation to customer feedback. I've encountered a couple of other problems too:

1. Misalignment to the business model - a company can get stuck on pleasing all customers rather than focussing on those that drive their business. Aligning the customer measures with the financial measures and the service delivery model is also important.

2. Inflexibility of the program - once established, any changes to the program can introduce bias, for example moving from telephone ot online or switching research vendors. With large sums of money at stake it can mean it can be hard to improve the program or test new data collection methodologies.

Again, thanks again for the article.

[Zane Safrit]

Great post. I love how you led us readers to the Net Promoter Score.

I thought you were a loon as I read the first few sentences. I anticipated you were one of those folks who give lip service to aligning w/ customer's needs and expectations and employee motivations but who then whine "yeah, but there's so many obstacles to doing it right....so we'll continue to do what we've always done. Nothing."

Obviously, I was wrong.

And just as I'm firing up a retort that includes "Have you heard about the simple, effective, Net Promoter Score/Ultimate Question Survey...." there you are, already doing it.

I am very anxious to read part 2 of this post.

Thanks for this post.

[Rob Markey]

Oliver and Zane: thanks for your kind words. Oliver points out some additional risks I think are spot-on. Failing to align the incentives with what really drives business results always comes back to haunt you, and "on average" customer feedback is risky in just this way. If your best customers (most profitable?) have a very different interaction profile than the rest, and if they account for just a small part of your feedback because there are so few of them, this can create a terrible distortion. This issue of customer segmentation is complicated enough to merit its own post, so I'll save that for another day.

[Andrew Panariello]

Great job with the "don'ts". So when are the "do's" coming?

[Brian Kaznova]

Rob, we solved this at The Ritz-Carlton Hotel Company with extraordinary front end planning in the critical aspects of survey design and analyses...independent, objective, properly scaled, 3rd party, statistically reliable random sample, validity, follow up (if necessary on non-respondent bias),comprehensive breadth/scope research into drivers of loyalty, sample by hotels statistically representative of customer mix (Individual Guests, Meeting Planners), statistically signifigantly improvement goals set by Hotels, drivers of Loyalty researched/explained to all Hotels...result=HIGHEST POSSIBLE SCORE by U.S. Dept. of Commerce Baldrige Examiners when we won the Baldrige Award in 1999 and established a benchmark fior satisfaction/loyalty in the World that has yet to be exceeded, albeit I may well be bereft of satisfaction/loyalty #'s since 2006 that may well be better then the RCHC World's Best Benchmarks.

[Steve Bernstein]

To Andrew's question, here's a recent article on some of the "do's" for linking comp to customer loyalty:
http://waypointgroup.wordpress.com/2010/12/03/incentives-for-customer-loyalty/

[Adam]

I don't like the gaming part. It is one of the reasons why I don't trust surveys most of the time. Especially the selective surveying part, I guess that is not fair.

Adam
Survey Storage