Keep a running record, per person, of finished work, what came back for changes and why, and which goals the work moved. Add to it every two weeks, from the work itself. At review time you have evidence, not memory. It is fairer, it is faster, and it holds up.
The review built on vibes
It is December. You are writing Sam's review. You remember the last month clearly and the rest of the year as a blur. You remember the one big mistake in March more than the forty things that went fine. You write "Sam is a solid performer who could be more proactive." Sam reads it and has no idea what to do differently. Neither do you, really.
This is normal, and it is a problem. Reviews built on impressions are unfair to people who had a quiet year of good work, and they do not hold up when the conversation gets hard.
What evidence looks like
Evidence is things you can point at.
- Finished work. The report, the campaign, the ticket, the page. With a date and a link.
- What came back, and why. Not "quality issues." Instead: "The June report came back because the brief did not name the decision. The August report was approved first time."
- What moved a goal. "Sam's three onboarding articles were linked to the setup-completion goal, which went from 41% to 55%."
- What was stuck, and what they did about it. "Waited four days for product screenshots, took placeholder shots and kept the date."
Opinions are "proactive," "solid," "needs to communicate better." You can hold an opinion. You cannot act on one.
The running record
Do not try to build this in December. Build it every two weeks, in five minutes, right after the one-on-one.
For each person, a simple document with four columns:
| Date | Work | Outcome | Note |
|---|---|---|---|
| Jun 10 | Pricing handout | Approved first time | Clear brief, self-checked. |
| Jun 10 | Leadership report | Came back once | Brief did not name the decision. Fixed the brief format. |
| Jun 24 | FAQ page | On time despite blocked screenshots | Took placeholders, kept the date. |
| Jun 24 | Onboarding articles (3) | Linked to setup goal, goal moved 41 to 48% |
By December, Sam has 25 rows. The review writes itself, and it is about things Sam did, not things you felt.
Let the person contribute
Share the record with the person. Ask them to add rows. This is not a secret file. It is a shared account of their work. People add things you missed, and they stop feeling like the review is something done to them.
What this changes at review time
- Quiet good performers get credit, because the record shows forty things that went fine.
- Problems are specific and early, because you noticed the pattern in week eight, not month twelve.
- Hard conversations are about work, with examples, not about character.
- Nobody is surprised. They saw the record all year.
What not to put in it
Impressions. Hearsay. Things about the person rather than the work. "Seemed disengaged in the meeting" is not evidence. "Missed three deadlines in October, each by two or more days" is.
Where M-Tracker fits
M-Tracker builds the record as a side effect of the work. Every finished task has a date, a brief, and the finished work attached. When it was marked done, M-Tracker checked it against the brief, so what matched and what was missing is already written down. Tasks link to goals, so the "moved a goal" column fills itself. Managers can see a person's work history without a separate document. There is no rating and no score. There is the work.
Try M-Tracker free14-day trial. No credit card. No employee surveillance.
Work you can trust, without redoing it.
Your team gets feedback before work reaches you, or your client.
Copyright © 2020 – 2026 M-Tracker. All Rights Reserved. Privacy Policy