Stop Bribing Your Employees for Good NPS Scores


Share on LinkedIn

We’re early in Customer Experience (CX) capability development, and I absolutely love it! We’re discovering the best practices that our successors will take for granted; “of course that’s how you do it.”

Unfortunately, being in this early stage means that some “best practices” aren’t. Some actually hinder the goal of improved CX – to create loyal customers who love your brand and come back time and again.

One “best practice” that can create a terrible customer experience is paying employees to achieve good NPS, or Customer Satisfaction, scores. This needs to stop.

The idea comes from a good place. We know that incentives drive behavior. The research is clear that, even though we all feel we’re immune to it, pay sure seems to drive everybody else, so why wouldn’t you pay people to improve customer scores? Especially if we know (or believe) that these scores link with true loyalty.

It’s the law of unintended consequences. Paying people for better NPS scores can result in a downgraded customer experience even when those scores improve.

Paying for scores is using a sledgehammer – a big, heavy tool without nuance. Incented employees will go to great lengths to receive that pay. We’ve all had this car dealership experience: “is there any reason you couldn’t give me a 10?” It gets worse! I’ve even been told, “we measure our success through survey scores, and I get in trouble with my boss if I don’t get a 10,” and “my performance is measured by your scores for overall satisfaction, so please consider that when filling out your survey.” This last example was from a Microsoft support rep – it happens everywhere!

But none of these are as blunt as the tactic used by our local ABRA Auto Body, who gave me a “sample survey” so I’d know how to fill mine out.

The problem is that, while there may be a link between NPS scores and business outcomes (which is an open question), it’s difficult for an employee to draw a straight line between specific behaviors and a customer’s likelihood to recommend. For example, should a call center rep ask about the weather and the customer’s kids to build an emotional connection? Or should they focus just on the matter at hand to decrease effort? Either is a valid approach, which means reps will react differently, and your customers will never know what to expect.

Paying for survey scores drives action. But is it the best action for your customers? Assuming your goal is to improve business outcomes like repeat customers and increased share of wallet, then incenting employees to beg for scores isn’t the best method.

What alternatives do we have? Here are three:

  1. Remove incentives altogether.

    Of course, this probably isn’t your call. But it needs to be listed as an option. If you do this, you need to be careful. You don’t want to send the message “Customer loyalty is no longer important to us, so we’re going to stop paying for it.” You also don’t want to create a big pay cut. Instead, reallocate the money towards items that will reward employees for improved customer results, perhaps training or finding something to pay them for instead of a comprehensive, high-level CX score. For example, you could

  2. Incent based on business metrics.

    You have to be careful to select the right metric and provide the management structure to ensure you get the right behaviors (see Wells Fargo). The wrong business metric drives the wrong behavior just as bad survey incentives do. If you’ve ever been on the line with a call center agent who’s managed to Average Handle Time, you know what I mean. Incenting somebody to get off the line as quickly as possible ensures that some will hang up before the problem is fully resolved. This creates both an annoyed customer and increased call time as the next agent has to read through what was done the first time. Higher costs and more frustrated customers are the risks of doing this wrong, so be sure the business metrics align with a customer’s interest. Number of calls isn’t right – but perhaps number of repeat calls on the same topic is. Of course, ideal business metrics vary by role, which also makes it more challenging. If you do want to use survey scores, you could

  3. Use a different score.

    I’m not wading into the “Is NPS/Customer Satisfaction/Customer Effort Score the right metric” discussion, although I’ve written on that plenty. Instead, I recommend focusing incentives on more actionable scores. Do you do a driver analysis on your surveys (this is a statistical methodology to see which questions most drive the item of interest, such as NPS)? If so, look at these to find which are most actionable. For example, if the top factor driving NPS is “sales rep knowledge,” then incenting on this will ensure reps learn your product well. Yes, there may still be some unintended consequences, as reps may go into more detail than clients want. But at least this is aligned with what customers want and need, and it’s easier to manage.

On this last front, when I worked in the field of employee engagement, we knew that incenting managers on their team’s engagement scores drove all kinds of bad behaviors. Stores would put up “5” balloons during the survey time, for example. Incenting them instead on “in the last year, did your manager take action to improve based on your last engagement survey” was harder to game and created the right kinds of behaviors. Can you do something similar?

Paying people to make customers happy makes a lot of sense and can certainly work in some situations, but you risk putting your teams in situations where it’s a whole lot easier to take shortcuts and just ask for the scores.

Republished with author's permission from original post.

Jim Tincher
Jim sees the world in a special way: through the eyes of customers. This lifelong passion for CX, and a thirst for knowledge, led him to found his customer experience consulting firm, Heart of the Customer (HoC). HoC sets the bar for best practices and are emulated throughout the industry. He is the author of Do B2B Better and co-author of How Hard Is It to Be Your Customer?, and he also writes Heart of the Customer’s popular CX blog.


  1. Well articulated. It is generally true that we can ‘expect what we inspect’, but as soon as incentive is applied, expected behaviour can fly out of the window. As Dan Pink said, “When the profit motive becomes unmoored from the purpose motive, bad things happen.”

    Using NPS or similar tools to measure or encourage specific employee behaviours is a fundamentally flawed practice. Similarly, incentives should never be applied to outcomes – but instead to behaviours. Drive the right behaviours, and outcomes will look after themselves.

    Interestingly, the best results I have ever seen were not driven by financial incentives, but by relentless support from coaches (and I mean Relentless – much in the way that a personal trainer interacts with a client). There were, ultimately, financial rewards as teams began to consistently hit their targets. But these served only to reinforce the growing pride and sense of accomplishment within the team.

  2. We’re actually fairly mature, and continually evolving, in CX. Creating artificial performance results, among companies still doing it, is a sign that the organization is immature and unevolved where CX is concerned. Leaving aside the reality that NPS is a highly flawed performance metric, jawboning employees to ‘game’ any CX measure through bribery or coercion renders resulting data subjective at best, misleading, and perhaps non-usable at worst. I’d submit, as well, that having employees do this sends negative signals to customers about the company’s honesty and objectivity. It also suggests the company is so insecure about how value is being delivered that they have to cheat their way to a good score. Further, using employees in this regard is unfair to them – – it makes them complicit in the enterprise’s manipulative treatment of customers.

  3. I think the degree to which employees seek incentive pay depends on the magnitude of the incentive, and the effort required to get it. There’s always a calculus involved – whether explicit, or tacit. Over the years, companies I’ve worked for have offered incentives (pay and non-pay) for many different things, and it’s not unusual for some to fail to achieve their intended purpose.

    The law of unintended consequence applies to any pay plan, not just variable pay. When salaries are set too high, employees can get complacent. When they are too low, employees more readily move on and take jobs where they can achieve higher incomes. For the latter, this can be an intended consequence as an alternative for firing. But mostly, both are unintended. The point is that the possibility of unintended consequences exists in any pay plan, and should not, by itself, discourage implementing the program. The key is recognizing what the unintended consequences might be, and then mitigating any consequential risks.

    Similarly, I cannot comment on the efficacy of NPS. But incentive pay remains an essential tool for businesses. Because it’s sometimes misused or misapplied in one situation does not mean it’s wrong elsewhere. The best discussion I’ve read on this topic is by Cornell professor Rob Bloomfield: What Counts and What Gets Counted: Seeing Organizations Through an Accountant’s Eyes. Bloomfield offers a clear and concise explanation regarding the purpose for incentive pay. I have described his points in an article I wrote, Teach Your Sales Force Well: Learning from Pay for Performance. (please see: The recommendations I included in my article extend elsewhere in the organization.

  4. Good discussion on a critical topic. How employees are recognized for performance is the essence of management.

    Of the options that Jim listed, I think that incenting for behaviors that link to business outcomes (perhaps measured by NPS etc.) is less likely to drive unintended consequences. And easier to fix. If sales rep training ceases to drive scores (or is gamed) then shifting to something else is easier than changing pay on a metric that the CEO is tracking.

    And a side benefit, to Michael’s point, is that the score’s credibility won’t be undermined.

    With the right culture, it’s possible to get away without formal incentives. I remember talking to an Amica insurance agent years ago and asking what management did to motivate them to give great service. Answer: “nothing.” They hire people that want to please customers, and let them do it!

  5. What Amica is doing in their hiring (and, I’d assume, mentoring, training, and reward and recognition) practices is a pillar of employee ambassadorship.

    When no carrot-on-a-stick incentives are needed for employees to consistently deliver top-end customer experience and value, it means a) the right kinds of performers, representing the enterprise in the right way, are in place, b) an objective program of employee reward and recognition is in operation, and c) virtue in the form of linkage to customer behavior is, indeed, its own reward.

    I’m also reminded of how MBNA’s people-first, stakeholder-centric culture emphasized how employees – all employees, and not just front-line service folks – should always put themselves in the customer’s shoes: The company, as one result of their cultural practices, never needed to be concerned about the credibility and actionability of their KPIs.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here