What the annual review really costs, and what to do instead
Key takeaways
- Deloitte counted close to 2 million hours a year spent producing ratings for its 65,000+ people.1
- Adobe reports saving about 80,000 manager hours a year after replacing annual reviews with regular check-ins.2
- Only 14% of employees strongly agree their performance reviews inspire them to improve (Gallup).3
- Simply dropping ratings is not automatically better. What seems to help is more frequent, developmental conversation.78
Most companies still run some version of the same ritual. Once a year a manager reconstructs twelve months of someone's work from memory, attaches a number, and the two of them sit through a conversation that both have been dreading. It is worth asking what that ritual costs, and what it buys.
The cost in hours
In 2015 Deloitte published how much time its own process took. Producing ratings for its 65,000-plus people "consumed close to 2 million hours a year".1
Adobe dropped its annual review in 2012 and moved to regular "Check-ins" between managers and their people. Two years later the company wrote that this saved "approximately 80,000 hours" of managers' time a year, and that voluntary attrition was trending downward.2 You will see much bigger Adobe numbers online, like a 30% drop in turnover; we couldn't find those in any Adobe source, so we don't repeat them.
Both figures are the companies' own counts, without a control group. They are still useful: they put a number on time that most organisations never measure.
What employees get out of it
Gallup asked employees directly. Only 14% strongly agreed that the performance reviews they receive inspire them to improve. Almost half (47%) said they got feedback from their manager a few times a year or less.3
The same report found that employees whose manager gives feedback daily, rather than once a year, were three times as likely to be engaged. That is a survey correlation, not proof that daily feedback causes engagement. Engaged people may simply notice or ask for more feedback. But it lines up with the 14%: the once-a-year conversation is not where most people say they learn.
What the rating measures
Then there's the number itself. When researchers analysed ratings for more than 4,000 managers, each rated by several people, the raters' own tendencies explained 53 to 62% of the variation in scores. The manager's actual performance explained 21 to 25%.4 Deloitte quoted this finding when it explained why it stopped relying on ratings the way it had.
A year-end score from one manager therefore says a lot about that manager: how lenient they are, what they happen to remember, who they compare you to.
What companies changed
By 2016, Peter Cappelli and Anna Tavis estimated that more than a third of U.S. companies were replacing annual appraisals with more frequent, informal check-ins.5 The redesigns share a few ideas, which industrial psychologists had been arguing for: performance is shaped by clear, ongoing expectations and informal, real-time feedback far more than by a formal yearly form.6
Deloitte, for example, dropped the once-a-year review and asked team leaders to check in with each team member every week.1
The honest caveat: dropping ratings isn't magic
It's tempting to conclude that you should just delete the review and the number. The evidence is more mixed than that.
- A 2016 survey by the advisory firm CEB (now part of Gartner) of almost 10,000 employees reported that performance fell by about 10% and engagement by 6% where companies removed ratings.7 It was a one-off vendor survey and we couldn't read the original report, so treat it as a warning, not a measurement.
- A quasi-experiment from the University of Southern California found some positive and no clear negative effects of going "ratingless", and attributed the gains to more frequent developmental feedback rather than to removing the number itself.8
The common thread: what matters is not whether a number exists, but whether people get regular, specific feedback they can use. Remove the annual review without replacing it, and many people end up with less feedback than before.
What to do instead
- Make feedback small and frequent. Short rounds during the year, close to the work, instead of one big reconstruction in December.
- Collect more than one view. Several colleagues who saw the work, averaged, are more reliable than one manager's memory.
- Keep it developmental. In a review of 24 studies, improvement after multisource feedback was about three times larger when it was used only for development and not for pay decisions.9
- Keep the record. If the feedback is written down and dated, the year-end conversation (if you still have one) starts from evidence, not memory.
That's the idea behind Reference Builder: short rating rounds between colleagues, averages instead of a single score, and comments that are kept, so by the end of the year the record is already there.
Sources
- Buckingham, M., & Goodall, A. (2015). Reinventing performance management. Harvard Business Review, 93(4), 40–50.
- Morris, D. (2014, March 10). Lessons learned with Check-in. Adobe Blog.
- Wigert, B., & Harter, J. (2017). Re-engineering performance management. Gallup.
- Scullen, S. E., Mount, M. K., & Goff, M. (2000). Understanding the latent structure of job performance ratings. Journal of Applied Psychology, 85(6), 956–970.
- Cappelli, P., & Tavis, A. (2016). The performance management revolution. Harvard Business Review, 94(10), 58–67.
- Pulakos, E. D., Mueller-Hanson, R., Arad, S., & Moye, N. (2015). Performance management can be fixed. Industrial and Organizational Psychology, 8(1), 51–76.
- Peck, E. (2016, June 29). Long live the performance review. HuffPost (reporting a CEB survey).
- Ledford, G. E., Jr., & Benson, G. (2019). Ratingless performance reviews: A quasi-experiment. Center for Effective Organizations, USC (working paper).
- Smither, J. W., London, M., & Reilly, R. R. (2005). Does performance improve following multisource feedback? Personnel Psychology, 58(1), 33–66.