Performance Review - It's Your Evidence That Gets Measured, Not You
The Anatomy of the Year-End Surprise
December arrives, performance review season starts. Your manager sends you a "self-review" form. You look at it, ask yourself "what did I do this year?" and remember the last three months. The first nine evaporated. You write: "I worked on project X, did frontend optimization, shared knowledge with the team." Three lines. You submit the form.
Two weeks later, the raise is 15%. Below inflation. You're surprised: "I did good work?" You did, but you couldn't show that you did.
The fundamental truth of performance review: you're not being evaluated, your evidence is. Your manager believing in you isn't enough; in calibration, your manager has to defend your work to a committee that scores you. If you don't give them ammunition, they don't have it.
This article is a practical end-to-end guide to the performance review period. Specific to Türkiye's software sector - the "promotion package" logic of US companies doesn't operate the same way here, but the principles are similar.
According to getSalary 2026 data, the average annual nominal salary increase was 35.9%, but inflation came in at 29.53% - a real increase of about 5%. So those who stayed in their company saw modest losses on average; large raises went to those with strong performance reviews. What you need to make that difference is the topic of this article.
Anatomy of the Performance Cycle
Companies in Türkiye typically run 6 or 12 month performance cycles. Most follow this sequence:
- Goal setting (period start): you and your manager define 3-5 goals
- In-cycle check-ins: monthly or quarterly goal updates
- Self-review (period end): you submit your written performance
- Manager review: your manager writes their assessment of you
- Calibration: managers meet to "adjust" team scoring across teams
- Feedback and raise: final score, feedback, and salary adjustment
Critical step: calibration. Your manager may genuinely want to give you a "high performance" score, but in calibration, when compared with other teams, they have to defend you. If they don't have an answer to "what did this person do differently?", your score gets pulled down. Your job is to make sure your manager has that answer.
Brag Document: A Process, Not a Year-End Task
This single practice handles 50% of the performance review. The same practice also helps in Mid → Senior transition (see the promotion article).
Open a notes file (Notion, Obsidian, README.md - doesn't matter). Spend 5 minutes every weekend:
- PRs I merged this week: link, description, line count
- Reviews I did: which PR, what kind of comments, which bug I prevented
- Problems I solved: in production or about to be, what was the impact
- Help I asked/gave: who, why, outcome
- Important things I said in meetings: which decision did I influence
If this file is ready at year-end self-review, it takes 30 minutes. If not, you write the last three months and your subconscious labels "the other nine months ordinary."
STAR format: Use Situation, Task, Action, Result for each achievement. Dry: "I did performance optimization." STAR: "Homepage LCP dropped from 4.2s to 1.8s; bounce rate decreased 14%; 3 engineer-weeks of work; proven via A/B test." These two sentences describe the same work, but one can be used in calibration and the other can't.
Typical Traps in Goal Setting
How goals are written at the start of the period shapes how they're scored at the end. Three common traps:
Trap 1: Writing actions instead of outputs
Wrong: "I'll develop new features."
Right: "By end of Q3, [feature X] will be in production and [metric Y] will have improved by 5%."
When actions are written, "did/didn't do" is debated, but when outputs are written, "happened/didn't happen" is clear. You usually look better when responsible for outputs, because work reaching users is the only valid measure of engineering.
Trap 2: Out of your control
Wrong: "User acquisition will increase by 20%."
Right: "By year-end, two new variants A/B tested in the acquisition funnel will confirm a 1.5%+ uplift."
Acquiring users is marketing's job, but A/B testing and designing variants are yours. Write goals on the side you can influence.
Trap 3: Too few or too many goals
Common trap: 8-10 goals. None achieved seriously, end of period scored "you did half." The right number is 3-5 goals; each a clear output, each measurable.
Self-Review: Three Questions
Forms vary in companies in Türkiye, but at heart three questions get asked:
1. "What did you do this period?"
Pick 5-7 standout achievements from your brag document and write in STAR format. Highlight numerical measurement when present. Not "I did X" but "I did X in Y time, the result was Z."
2. "What did you learn?"
Write at least two technical developments: a new technology, an area you deepened, or a design problem you solved. This is the answer to "is this engineer growing?" and leaving it blank reads as "stalled" in calibration.
3. "What are your goals for next period?"
Here you can request promotion, scope expansion, put your mentorship interest in writing. When "I want to own the design of project X" is written, your manager can't forget it next quarter.
Bonus question (add yourself if not in form): "What didn't work this period, why?" Asking this yourself wins you two things: (1) demonstrates self-awareness (positive in calibration), and (2) prevents your manager from saying the same (you've acknowledged it instead of being defensive).
1:1 with Your Manager: Performance Isn't a Year-End Thing
Performance review isn't year-end; it's accumulation built in every 1:1. Once a month, ask these three:
"How is my performance currently perceived?" Follow "very good" with "based on what evidence?" If there's no data, your manager won't have data to defend you either.
"What's my next growth step?" Promotion, scope, or role change; keep it on the agenda monthly instead of springing it at year-end.
"What should I focus on this period?" This is where you find what your manager wants to hear from you and you note the answer, then close it out in the next 1:1 with "I did this."
These questions don't bother your manager; on the contrary, they're a natural part of management work. Developers who don't ask have to handle disappointment at year-end.
Calibration: The Meeting You Don't See
Calibration is where managers gather and compare team scores. You're not there but you're being talked about. Two critical truths:
Truth 1: Calibration is zero-sum: In many companies, the "high performance" rate isn't allowed to exceed a percentage of the team (e.g. 20%), and if your team has 5 people, only 1 gets the highest score. That person may or may not be you; the decision depends on your manager's capacity to defend you against other managers.
Truth 2: "Invisible work" loses value: A developer who quietly does work and solves problems before anyone notices has no story in calibration, and the question "what did this kid do?" hangs in the air. So sharing your brag document with your manager, talking about work in 1:1s, and doing short demos in team meetings all are calibration ammunition.
Practical advice: Two weeks before calibration, in your 1:1, honestly say "the three achievements I'm highlighting this period are..." because your manager will reflect them in their notes. This isn't shamelessness; it's professional communication.
Salary Raise Negotiation
Performance review results are out, you got "high performance," and raise is 20%. If sector inflation is 30%, you're losing in real terms, and here negotiation is openly necessary.
Two approaches exist:
Approach 1: Data-based request
"This period I demonstrated these concrete achievements. getSalary 2026 median is in the 132,500 TL band; I'm closer to the Senior side and currently below that median. My ask of 25% reflects this data."
This works for 15-25% additional asks and convinces manager and HR with data. The first salary negotiation article and counter-offer article cover the negotiation details.
Approach 2: Outside offer based
If internal negotiation fails, the job hopping article's data shows the way: those moving to new companies get 40-60% raises. Don't use this approach unless you're genuinely ready to leave, because bluffing isn't received well.
If You Got a Low Performance Score
Sometimes we end up where we don't want to. If "doesn't meet expectations" came:
- Don't panic, but don't rush either. Ask your manager for detailed feedback: "On what specific topic did I fall short?"
- Get written feedback: "If I achieve goals X, Y, Z in the next 90 days, will my situation normalize?"
- PIP (Performance Improvement Plan) typically means "polite breakup conversation," and PIPs rarely end successfully; during this process, also start looking outside.
- Keep building evidence: The way to pass a PIP is also keeping a brag document; this time to convince your manager.
Three Things to Do This Week
- Open a brag document file (Notion, Markdown, doesn't matter). Backfill the last 3 months.
- Add 1:1 to your manager calendar: monthly, 30 minutes. If you're not doing it, start.
- Mark the next performance period on your calendar. Learn the date; the countdown starts.
Performance review shouldn't be a surprise. If at period end you struggle when your manager asks "what did you do this period?", the same thing happens in calibration.
In one sentence: "Performance is the measure of your documentation, not you; without documentation, there is no performance."