Ask most owners of a 20 to 150 person company in Bangladesh when their last performance review ran, and you’ll get one of two answers. Either “we don’t really do that” or a long pause followed by “sometime last year, I think.” That’s not a judgment on the owner. It’s what happens when hiring, payroll, and client fires eat every hour, and reviews get pushed to “next quarter” until next quarter never comes.
Why small teams skip reviews, or fake them
Two failure modes show up over and over, and they come from the same place.
The first is the team that skips reviews outright. Raises and promotions happen based on who asked, who’s been around longest, or who the owner happens to like, and everyone quietly knows it. Nobody owns the process, so it doesn’t exist.
The second is worse in its own way: the team that holds reviews once a year, and every single person gets a 4 out of 5. Managers hand out safe, middling scores because a low rating invites a conversation nobody wants to have, and a high rating sets an expectation for a raise nobody budgeted for. Everyone lands in the same comfortable middle, the review takes twenty minutes, and nothing changes. Employees can tell. A review that produces the same score regardless of what actually happened during the year isn’t feedback, it’s a formality people tolerate.
Both patterns share a root cause: nobody built the scaffolding a review needs to hold up. You don’t fix “everyone got a 4” by asking managers to be braver. You fix it by giving them something concrete to rate against.
What a review cycle actually needs
Three things, in order, and skipping any one is why most homegrown review processes fail.
Something to review against. A review without goals set at the start of the period is an opinion formed after the fact. If nobody wrote down what “good” looked like in January, rating someone in December just means recalling whoever made the strongest recent impression.
A rating scale that means something. “How did they do this year” only has a useful answer if there’s a defined scale behind it, with labels that mean the same thing to every manager using them. A 3 should describe the same level of performance whether the manager writing it runs sales or finance.
Calibration. This is the piece almost nobody at this size has heard of, and it’s the one that actually fixes “everyone’s a 4.” Calibration is a session where HR sits down with managers and lays every rating side by side: does a 4 from you mean the same thing as a 4 from her? Usually it doesn’t. One manager rates generously because they hate conflict, another rates harshly because they think that’s what “high standards” looks like, and without calibration those two teams are being judged on completely different scales while everyone assumes it’s fair.
Calibration doesn’t mean forcing a curve where someone has to get a low score. It means making sure the scale itself is applied consistently, so a rating reflects performance instead of which manager happened to give it.
Not every company needs all of that for every role. A 25-person company running self-assessment, peer, upward, manager, and calibration for a junior hire ends up filling out more forms than managing people. What matters more is that whichever phases you pick stay consistent across the company for that cycle, otherwise comparing one team’s results to another’s later becomes meaningless.
The review shouldn’t be the only feedback someone gets
A formal review, even a well-run one, has a structural problem: it’s often the only real feedback moment an employee gets, and it lands months away from the work it’s describing. Someone who did something worth noting in March is hearing about it, if at all, in December, by which point neither of them remembers the details clearly.
The fix isn’t more formal reviews, it’s giving people a way to recognize good work as it happens. A quick note of thanks after someone covers for a colleague, a manager flagging a well-put-together report, a peer calling out help they got on a tight deadline. None of that needs a rating scale. It just needs to be visible and easy to give, so the annual review becomes a summary of a year people already had a sense of, not the first time anyone said anything.
Where this fits with the rest of HR
Performance reviews don’t sit apart from the rest of how a company runs its people. Goals set at the start of a review period are more useful when they connect to something the whole department is working toward, not each person inventing their own targets in a vacuum. And a manager writing fair, evidence-based reviews for eight people twice a year, by hand, in spreadsheets, is fighting the format as much as the work of managing.
This is the part of the problem Utso’s performance management module is built around. Goals cascade from company level down to department and individual. Review cycles are configurable, so a company can run Self, Peer, Upward, Manager, and Calibration phases in whatever combination fits a given role. The rating scale is a 5-point scale with labels a company sets itself. Calibration is a built-in session where HR can see rating distribution across managers and normalize it, with an audit trail showing what changed and why. And outside the formal cycle, peer recognition and kudos let people acknowledge good work as it happens, so the review isn’t the only time anyone hears anything.
None of this replaces a manager who actually knows their team. It stops the review from collapsing into either “we didn’t get to it” or “everyone’s a 4,” which is where most reviews at this size end up without some structure holding the process together.
Where to start if you have nothing in place
Don’t start with the full version described above. Start with goals: get every manager to write down, in one sentence per person, what a good quarter looks like for that role. That alone gives you something to rate against later, which is the piece most homegrown processes never had. Add a rating scale once goals exist. Add calibration once you have more than two or three managers whose ratings need to line up. And start giving out small, informal recognition now: it costs nothing, and it’s the easiest habit on this list to build early.