Using Satisfaction and Loyalty Metrics as Employee KPIs
Introduction
What are the pros and cons of using satisfaction and loyalty metrics as employee KPIs? Is it appropriate to use them at all? How can a corporate performance management or strategy function decide whether to adopt them as KPIs? These are the questions every decision-maker should think about carefully, and they’re what I’ll address in this article. I hope you enjoy the read, and I welcome your criticism and comments — the answers to these questions depend a great deal on personal judgment and vary by case. Let me say upfront that understanding this topic is extremely important, because it’s directly tied to either being fair or unjust to employees.
Key points to consider
The type of metric used. Some metrics are simply not appropriate. For example, the Customer Effort Score (CES) cannot serve as an employee KPI because it primarily reflects the complexity or smoothness of the organization’s processes. Using NPS (Net Promoter Score) is also somewhat sensitive — I don’t recommend it at all if you want to tie it to individual employee performance (each employee separately). I personally have reservations about its use in general, let alone using its results as a KPI. There’s also the satisfaction score, which is probably the best of the bunch if you must use one.
Survey design. Some designs ask the metric question at the very start of the survey to capture the customer’s overall view, while others derive the overall average from multiple aspects related to what’s being evaluated. The latter is fairer, especially if it includes one or more questions specifically about the employee — so the average of those questions’ results is used instead of the overall average, which factors in many other aspects that have nothing to do with the employee.
Know also that using the results of several aspects of an employee’s performance is far better than using the result of a single question that evaluates and even names the employee directly. The latter approach causes bias. I have personally been challenged by a customer asking whether their response would translate into rewards or penalties for the employee. When, after some initial reluctance, I answered yes — the respondent changed their answers and inflated their rating, even when the employee didn’t deserve it.
The nature of the relationship between the metric and the employee. Does the customer’s rating relate entirely to the employee and their conduct? In other words, is the employee fully authorized and empowered to serve the customer? If yes, that usually means the relationship is direct — the higher the employee’s performance and adherence to quality standards, the higher the metric.
This relationship can be tested using statistical analysis software if both variables are available (employee performance and service satisfaction). If a relationship is established, the score can be used as a KPI, with the weight of the KPI equal to the strength of the correlation. What if the relationship is not proven? In that case I do not recommend tying the metric to the employee or using its results as a KPI at all.
Coming back to survey design: if the evaluation questions are about the roles the employee performs rather than evaluating the employee directly (to avoid relationship-based bias), then there’s no need to test for correlation — we’re asking the respondent to evaluate clear, specific aspects that directly involve the employee.
Reliable results. This is purely a scientific matter that a marketing research specialist can advise you on. Reliability is affected by several factors, including sample size, margin of error, how well it represents the overall population, survey design and question hierarchy, the data collection method, and finally the quality of the data collectors (field researchers). The topic is long, and this is not the place to detail it.
If a metric is adopted as a KPI, how should the target be set?
Most marketing research scales consist of five, seven, or ten points. Let’s assume we use the most common — the five-point scale. How do you set a KPI target? On what basis do you adopt a score of 4 out of 5 as the target satisfaction level to achieve over a defined timeframe? You have two paths: either use benchmarking indicators — for which I have a detailed article elsewhere — or measure periodically over six months, then take the average for the full period as your baseline and set a margin of 5–10% above it as your target. For example, if your average satisfaction over six months was 3 out of 5, set a target of a 5% increase over a quarter — meaning 3.25 out of 5. In the early months you can test the rate of increase to adjust the target, because it’s a mistake to lock in the increase rate — as the target gets closer to 5, achieving it becomes harder. The idea of trying any initiative for at least six months is excellent in this kind of strategy, because when applying any new metric you’ll find some resistance from members of the organization at first, but after a while it will be adopted and its importance accepted.
”If you want to be obeyed, ask for what’s achievable”
If you adopt a research metric as a KPI, this must be accompanied by enrolling your employees in training programs related to delivering distinguished service, and holding occasional workshops to brief them on the main reasons satisfaction or loyalty drops or rises. Make them feel they are part of a distinguished-service strategy. Run a competition among them, announce the employee who achieved the highest satisfaction rate across all employees, and honor them at a group event. The point isn’t simply to use marketing research metrics — their weight within the KPI scorecard ultimately won’t be large, and may be ignored entirely. I remember some colleagues at a previous employer who didn’t care about attendance and punctuality, even though it made up 10% of the annual evaluation. The lesson here isn’t to raise the weight (importance) to scare employees, but to convince them to adopt it, motivate them, and make it easy for them. Remember, making it hard may shock you with strange practices you didn’t anticipate — including employees asking customers to give them high ratings, otherwise they’ll suffer such-and-such from management. They’ll dramatize the consequences, customers will sympathize with these claims, and some will escalate to management, who will be shocked. Consider too measuring employee satisfaction, loyalty, and engagement just as you do for customers — you may not have provided them what they need to deliver outstanding service that delights your customers. Your employees, in the end, are a mirror of how you treat them; delighting your customers begins with delighting them.
Which is better — reward or punishment?
There are two approaches used when linking these results: either punishment, such as applying deductions — which I do not recommend at all — or reward, giving the employee incentives based on their performance, or having the outcome of their performance feed into the annual bonus formula. It’s important that the impact be individual, not at the department level, because rewarding the staff of a department based on departmental performance is demoralizing: the high performer is wronged by the slack of others, and the laggard is rewarded by the diligence of others. “No bearer of burdens shall bear another’s burden” — it’s better to treat each employee individually. You can also blend reward and punishment: penalize the employee whose performance falls below the average by a certain percentage to avoid total neglect — there is, after all, a minimum quality floor, and organizations have reputations to protect. Both approaches are used in strategies that link marketing research metrics to KPIs. While the use of incentives is itself a debated topic in management — there’s debate about whether it’s right to reward an employee for doing what they’re supposed to do anyway — this isn’t the place to discuss it, and it comes back to the organization and its HR policies.
Conclusion
I ask God that what I’ve written benefits you, and I’d love readers to comment and enrich the article. As social media influencers like to say: if you liked the article, share, like, and subscribe :) I’m trying to ride their wave.
Related framework notes: NPS, CSAT, CES, Voice of Customer (VoC), Customer-Centric Operating Model
← Articles MOC