How Customer Service Surveys Are Eroding Workers’ Rights

The use of customer reviews to evaluate workers is rising across the service sector. Once you get past the shiny rhetoric of consumer empowerment, it’s a new variation on an old story: enlisting customers on the side of management to help discipline and exploit workers.

Employee hands a customer their drinks at a Starbucks drive-through. (Alex Burstow / Getty Images)

“Customer participation” systems have long been a mainstay in many workplaces. These systems are increasingly being deployed to assess worker performance in a variety of customer-facing jobs across the Canadian service sector.

Corporate bosses tout customer review systems as a way for employees to improve their on-the-job performance. In truth, their main effect is to make workplaces more difficult for workers — which hasn’t affected their enormous popularity with bosses.

This crowdsourced form of control gives managers ready-made excuses for withholding bonuses and work hours. In a set of interviews with Jacobin, workers from different sectors discuss some of their experiences dealing with this modern form of workplace discipline.

Work In the Big Five

Jamal* works as a financial advisor at one of Canada’s “Big Five” banks. To evaluate client satisfaction, his bank relies on what’s known as the Net Promoter Score (NPS), a widely used system that poses to customers, often at the end of a survey, one crucial question: “How likely are you to recommend [this bank] to a friend or family?”

Respondents are then subdivided into three categories: “detractors,” who give a score of six or lower; “passives,” who give a score of seven or eight; and “promoters,” who give a score of nine or ten Jamal reports that getting “promoters” on the NPS question is far and away the most important goal. Yet it is the question that least reflects an employee’s actual performance.

It’s not uncommon, he says, for a customer to give a high score on those questions that concern the ways in which a worker actually affects an outcome, only to respond to the NPS question with a low score. In those cases, the bad review still harms a worker’s overall rating.

What makes the situation particularly frustrating for workers like Jamal is that these reviews can affect their pay. Bonuses are calculated based on a broad average of reviews across an entire department. There is no way to remove a review or pursue recourse.

Jamal notes that his department, when compared to tellers and daily banking staff, is among the more privileged. They often know in advance which clients will be surveyed. All the same, employees aren’t allowed to explain the value of a nine or ten to a customer, nor the effects that a bad review can have on a worker:

The verbiage has to be very specific. We have to play around it, like, “What can I do to meet your expectations?” or “What can I do to deserve a nine or ten?” Some bullshit like that.

Sometimes, Jamal points out, the company sends out anonymous surveys to assess lower-level staff:

One year, we were trending on a really good path, and we were going to get a massive bonus, and on the last day of reviews, we got two anonymous detractors. We were all like, “Did someone at head office do this?” But nobody was able to disclose any information to us about it.

Tellers and other daily banking workers, he notes, receive the worst of it. Customers often complain that their mortgage application was not approved, the bank’s wait times were too long, or even that the bank’s fees were too high — all problems beyond the worker’s control. “It’s not fair for us to get dinged for it,” Jamal says.

One-Star Bosses

In 2015, restaurant chain Olive Garden announced that every table at its 845 US locations would host a Ziosk: a tablet that offers “browsing, ordering, payments, survey, e-club/loyalty, entertainment and more.” Olive Garden president Dave George claimed that introducing Ziosks would enable workers “to make more meaningful connections and provide more personalized service.”

The jury is still out on whether the introduction of Ziosks has made the connections between customers and staff more meaningful. One thing, however, that has increased is the prevalence of employee ratings. As writer Clint Rainey has pointed out, digital ratings “have replaced the old-school, paper-based comment card.” That data, Rainey writes, can then “be used to assess job performance, which servers say renders an already challenged job sector all the more abusive, inaccurate, and unfair.”

Like other tablets and review systems, Ziosks provide servers with scores. When the ratings of workers take a dip, management often steps in. Some servers in the United States, for example, have seen their scheduling slashed as a result of low scores.

In 2017, Forbes noted that “a full 43 percent of companies base some portion of frontline pay on customer feedback ratings.” The digitization of this process has led to more customer feedback than ever before. This is a selling point because customers are far more likely to complete digital customer service surveys.

Ratings systems result in workers often feeling pressure and anxiety about their performance, sometimes leading to awkward conversations or pleas to customers for high ratings. Meanwhile, customers — wittingly or not — function as middle managers by helping companies control and surveil workers.

Did the Barista Get to Know You?

Valerie, who works at Starbucks, explains the company’s customer review system. Surveys are distributed by email to customers who use Starbucks’s mobile order system or their Starbucks card to pay. The questions cover the speed of the service, the cleanliness of the store, the taste of food and beverages, safety, and concludes with the question, “Did your barista get to know you?”

Customers respond to the questions on a scale of one to seven, as Valerie explains:

A hundred surveys are released per store per week. If the customer answers anything less than seven, their answers are not scored or recorded. At the end of each week the store is given their scores, and based on these scores, stores are ranked from most to least successful in operations.

Because only sevens count, each store’s goal is to get as many sevens possible, rather than the highest average score:

If a hundred people are surveyed, forty-five people say service was a six and fifty-five people say service was a seven, the score would come back as fifty-five out of a hundred as a final for the week. There’s no way to know how many people have scored a six or lower.

The big one that really falls on baristas is “Did the barista get to know you?” — we call these “connection scores.” As most people are aware, lots of Starbucks locations closed during the pandemic and just about all of these stores had the lowest “connection scores” in their districts.

While Valerie bristles at the idea of people losing their jobs because of store closures that stem, at least in part, from low scores, she does feel that customer surveying itself is fair. It is, she says, the delivery of these surveys and the consequences that follow from them that are unfair:

There’s so much focus on speaking to every single customer that most conversations are forced or fake. Coworkers with anxiety really struggle to rise to this challenge and those who underperform in this department will constantly be “coached” and told they need to do better. Given the circumstances brought on by the pandemic, I find this practice to be quite ridiculous.

Valerie says the whole process doesn’t faze her much. “It can be a good incentive to get your coworkers to kind of strive to be better,” she says. What does bother her is that “these scores are dictating what stores in the city stay open or close. It may be a portion of the decision, but it’s still contributing to it.”

Vengeful Assholes

Katie works at one of Canada’s major tech retailers. In addition to the NPS question, the survey used by the company that she works for asks customers to rate employee performance. It’s the relationship between the answers to those two questions, Katie says, that is most important to management:

The survey is worded fairly, but people shouldn’t use the forum to rail at a company they don’t like if the low-level person didn’t do anything wrong. It’s vengeful assholes or people who have unrealistic expectations that ruin the process.

She says workers aren’t often punished for the scores, but if a worker’s score trends downward, that might lead to a “deeper conversation.” There are few routes for recourse if workers find a rating unfair:

You can share your perspective, and it isn’t dismissed if it’s clearly a one-off incident. But you can’t change the score. Because maybe it’s an indication of how you managed that customer’s perception, even if they’re mad at something that has nothing to do with you.

Like Starbucks, the ratings method is designed strategically:

The score is the percentage of the five out of five surveys, minus the percentage of scores that are three out of five or lower. The four out of fives don’t count. The company doesn’t measure success by setting 80 percent as the goal. So when people are picky and are, like, “I’ll give them a four,” I think, damn, man, you don’t understand what that does.

A Feature, Not a Bug

The design of customer surveys plays a large part in the way ratings systems function. Systems are typically weighted strategically, yet designed with sky-high expectations, and rarely, if ever, provide accurate pictures of workplace performance. Many workers stress the fact that customers don’t understand the crucial importance of perfect scores, not just good ones, or the importance of specific questions over others.

Workers feel that survey responses — particularly bad ones — can speak more to a customer’s view of the company or organization than to the worker whose performance is actually being assessed. Poorly designed surveys fail to properly distinguish the worker from the company in the eyes of the reviewer. In the case of the NPS question, they fuse the two together.

Many workers are, of course, frustrated by this common problem. Even if a bad rating is out of an employee’s control, they are the ones that ultimately get “dinged.” Unfair as this may seem, management often justifies it, as we have seen, on the grounds that workers should be able to “manage a customer’s perception” of a company.

Of course, ratings systems may be a perfect representation of a structure that is itself broken. The imprecise pictures that surveys produce may be a systemic error that companies do not actually want to fix. Having an inbuilt vent for customer frustration is a tremendous advantage for bosses.

Review systems ensure that workers form a human shield against customer frustration. This frustration may stem from deficiencies of the company that have nothing to do with employee performance. It is the perfect way to offload normal operating shortcomings, such as long wait times, onto the shoulders of frontline staff.

The truth is, the problems with surveys, which may appear to be the result of poor design, are not really problems at all. They are solutions. Ratings systems are a means of transforming customers into tools for management surveillance, even when a manager is busy elsewhere.

During the pandemic many frontline workers have been celebrated as “essential.” This increase in status has not, however, reduced the amount of pressure placed on workers to meet the expectations created by ratings systems. “We’re all very drained,” Valerie says, “being deemed essential, coming into work every day. We’re coming into contact with hundreds of people in a single shift and putting on a smile while doing so.”