Every January, thousands of small companies perform a ritual they inherited from companies fifty times their size. The manager blocks a weekend to fill in forms about things she half-remembers. Employees write self-assessments calibrated to sound humble but not too humble. Everyone sits through an hour that both sides quietly dread, a number gets attached to a year of work, and then the form goes into a folder nobody opens again until the next January.
At a 400-person company this ritual at least feeds something — calibration committees, compensation bands, succession charts. At a 12-person company it feeds nothing. There is no committee. There is no band. There's you, a person you work beside every day, and a form that makes you both stiffer and less honest than you were at lunch. The annual review at a small company is theater, and the audience knows it.
The problem isn't that reviewing performance is a bad idea. It's that the annual format fails for a structural reason no amount of better form design can fix: it collects one data point a year, and it collects it from memory.
Why the annual format can't work at your size
Memory is the whole problem. Ask a manager in January what someone did in April and you'll get whatever survived nine months of overwriting — which research on rater behavior says is mostly the last eight weeks, plus one or two vivid incidents. That's recency bias, and at annual cadence it's not a flaw in the process, it is the process. The review claims to measure a year; it actually measures a quarter, dressed up in a year's authority.
The second failure is surprise. When feedback is stored up for a single annual event, problems arrive at the employee eleven months late, and by then they've compounded — in the work, and in your frustration. The employee's entirely fair response is "why didn't you tell me in March?" There is no good answer to that question. A review should never contain information the person is hearing for the first time.
And the third is what the format does to the conversation itself. Attach a rating to an annual verdict and the employee's rational strategy is defense, not reflection. You wanted a conversation about growth; you built a hearing about a number.
The replacement: one hour a quarter, four questions, zero surprises
Here's the system that actually fits a small team. It has two components: a running evidence file you keep all quarter, and a one-hour quarterly conversation built on four fixed questions. That's the whole apparatus. No self-assessment essays, no 1–5 scales on "communication," no forms with drop-downs.
The evidence file
For each person, keep a note — a doc, a card, anywhere you'll actually use — and add two lines a week: something concrete they did, good or concerning, with enough detail that future-you knows what it means. "Caught the billing bug before the client did (Tues)." "Third deadline slipped this month — the Hendricks deck." Twenty seconds per person per week. This is the same discipline that fixes hiring — evidence written down at the moment, not impressions reconstructed later — applied to the people you already have.
The file does two jobs. It makes the quarterly conversation specific instead of vibe-based. And it quietly protects everyone: if you ever do face a hard call — a promotion, a termination, a dispute — you have a contemporaneous record instead of a memory contest. Ask any employment attorney which one they'd rather defend.
The four questions
The quarterly hour walks through four questions, the same ones every quarter, both of you knowing they're coming:
- What went well this quarter — specifically? Open with evidence from the file, not a generic "good quarter." Named wins compound; vague praise evaporates.
- What was harder than it should have been? This question surfaces friction — broken processes, unclear ownership, missing tools — before it curdles into "performance issues." Half of what looks like underperformance at small companies is actually process debt the person was too polite to escalate.
- What's the one thing to change next quarter? One. Not a development plan with five competencies. A single, concrete, agreed change you'll both be able to check in ninety days — because you wrote it down and it goes first on the next quarter's agenda.
- What do you need from me? Asked last, answered honestly, acted on. This is the question that makes the whole ritual worth an hour of their time — it's the one that treats the review as a two-way instrument instead of a verdict.
Notice what's absent: a rating. If you need numbers for compensation, do compensation separately — on market data and role scope, on its own calendar. The moment a number enters the room, the conversation dies in its defense. Decouple them and both get better.
The no-surprises rule
One rule governs everything above: nothing said in the quarterly conversation may be new information. If a problem is serious enough to raise in the review, it was serious enough to raise the week you noticed it — and the evidence file will show you did. The quarterly hour is where patterns get named and priorities get set; it is never where bad news debuts. Hold yourself to that and reviews stop being dreaded, because nobody walks in braced for an ambush.
Where the tool fits
The honest failure mode of this system isn't disagreement — it's the file. Two lines a week per person is a twenty-second habit that dies the first genuinely busy month, and once the file is empty the quarterly conversation slides right back to vibes and recency. The mechanics are exactly the kind of thing software should carry so the judgment doesn't have to.
That's where HRByDesign earns its seat: it scaffolds the quarterly cycle — the evidence log per person, the four-question agenda, the agreed one-thing carried forward to next quarter's opener — and drafts the conversation summary from your own notes so the record writes itself while it's fresh. It doesn't rate anyone and it doesn't decide anything. The conversation, the judgment, and the call stay yours; what the tool removes is the administrative drag that kept pushing the whole discipline to "next quarter."
The bottom line
Small teams don't need a lighter version of the big-company review. They need a different instrument entirely: evidence collected in the moment instead of reconstructed from memory, conversations at the cadence problems actually develop, one agreed change instead of a development plan, and a hard rule that nothing in the room is a surprise. An hour a quarter, honestly run, will tell you more about your team — and do more for it — than any January ritual ever did. The form was never the point. The conversation was.
— Tom
Run the quarter, not the theater
HRByDesign scaffolds the quarterly review cycle — per-person evidence log, the four-question agenda, and a drafted summary from your own notes. No ratings, no forms. The judgment stays yours.
See how HRByDesign works →