The short version
- Run your first startup performance review at least six months after incorporation or after an employee has completed two meaningful quarters, not because a calendar says so.
- Judge people against written outcomes, not hours online, founder loyalty, or who speaks most confidently in meetings.
- Use a simple four-point rating scale, separate performance from salary, and document one clear improvement plan for every weak review.
- Your first review is mainly a test of your management system; if expectations were unclear, the failure is partly yours.
Your first startup performance review in India should be a structured conversation about outcomes, behaviour, and the next six months. It should not be a surprise appraisal, a disguised salary negotiation, or a founder’s memory contest. Set role-specific goals, collect evidence, ask the employee for a self-review, rate performance on a four-point scale, and finish with written actions. Do this before your team reaches 15-20 people. After that, informal founder feedback stops being reliable.
Most founders delay the first review because the team is small. That is precisely why they should not. At 10 people, one weak hire can consume 10% of your operating capacity. One excellent employee can carry a product launch. You cannot manage either outcome through corridor conversations and WhatsApp messages.
Why your first performance review is not an HR exercise
At a seed-stage company, performance reviews are often treated as corporate theatre. Founders fear creating bureaucracy. Employees fear a vague rating will affect their salary. Managers fear difficult conversations.
These fears produce the worst possible system: no written expectations, random feedback, and a salary discussion every few months that pretends to be a review.
Your first review has three jobs:
- Tell the employee what they did well and what they did poorly.
- Tell you whether the role, goals, and reporting structure are working.
- Create a written operating record before memory and politics take over.
This matters in India because early teams often mix full-time employees, consultants, interns, ESOP holders, and remote staff. People doing similar work may have different titles and different pay. A review gives you a common language for contribution.
Do not copy a large company’s annual appraisal process. You do not have the HR team, manager layers, or budget to support it. Build a small system that produces honest decisions.
When should a startup run its first performance review?
Run the first formal review when the employee has completed two meaningful quarters, or roughly six months in the role. For a senior hire, review at 90 days and again at six months. For an intern or short-term contract role, use a project-end review instead.
Do not run a review after three weeks simply because the probation period ended. You will measure onboarding quality and first impressions, not sustained performance.
Also do not wait until the end of the financial year. Indian companies inherit annual appraisal habits from large employers, but your business does not operate on the same rhythm. A seed-stage startup can change its product, pricing, customer segment, and priorities within one quarter.
Use this review cadence
| Company situation | Recommended cadence | What to measure |
|---|---|---|
| 1-10 employees | Monthly one-to-ones and a written review every six months | Role clarity, output, ownership, learning speed |
| 11-30 employees | Monthly one-to-ones and quarterly written check-ins | Goals, execution quality, cross-team behaviour |
| 31-75 employees | Quarterly reviews with manager calibration | Business outcomes, management quality, role scope |
| 75+ employees | Quarterly goals with a formal half-year and annual cycle | Results, leadership behaviours, succession, compensation |
At every stage, the one-to-one is more important than the formal review. A review should never reveal a problem for the first time. If you are telling an employee in December that their work has been below standard since July, you failed as a manager.
Investor rule: A performance review must contain no surprise. Surprises belong in product launches, not people management.
Start with the job, not the rating
The biggest mistake founders make is asking, “How would I rate this person?” before asking, “What was this person hired to achieve?”
Write a one-page role scorecard before you review anyone. If the role has no scorecard, your review will become a personality judgement.
The scorecard should contain:
- Mission: Why this role exists.
- Three to five outcomes: What must be true after six months.
- Measures: The numbers or evidence that show progress.
- Behaviours: How the person is expected to work.
- Decision rights: What they can decide without founder approval.
For a first sales hire, “build the sales function” is not a scorecard. It is a slogan. A better version might include:
- Build a qualified pipeline of 150 target accounts in a defined segment.
- Close ₹25 lakh of collected revenue, not merely signed orders.
- Document the sales process, objections, pricing exceptions, and handoff to implementation.
- Keep customer acquisition cost within the agreed range.
For an early product engineer:
- Ship the agreed onboarding flow by a specified date.
- Reduce failed payment or signup errors from the baseline to the target level.
- Own production incidents and publish post-mortems within 48 hours.
- Improve test coverage or release reliability without slowing urgent fixes.
For an operations manager in a marketplace or B2B business, outcomes may include fulfilment time, gross margin, rejection rates, collections, and vendor quality. Do not rate them on “being proactive” unless you can describe the action and its business effect.
Separate outcomes from effort and activity
Indian startup founders often overvalue visible effort. The employee who stays online until midnight feels committed. The employee who quietly fixes a broken process looks less heroic.
That is a dangerous bias. Long hours are not a business metric. Neither are the number of meetings attended, messages sent, or tasks moved to “done”.
Review four dimensions:
| Dimension | Question to ask | Evidence |
|---|---|---|
| Results | Did the person deliver the agreed business outcome? | Revenue, retention, releases, margin, cycle time, quality |
| Quality | Did the work hold up without repeated rescue? | Defects, rework, customer complaints, incident history |
| Ownership | Did the person identify problems and close them? | Decisions made, risks raised, follow-through |
| Team behaviour | Did the person make others faster or slower? | Handoffs, documentation, conflict, feedback from collaborators |
Do not turn this into a fake mathematical exercise. A scorecard is a forcing device, not an excuse to pretend that human performance can be calculated to two decimal places.
Use evidence from the company’s actual operating system:
- CRM and collections data for sales.
- Product analytics and incident logs for engineering.
- Customer tickets and renewal data for support.
- Gross margin, fulfilment time, and rejection rates for operations.
- Hiring funnel, joining rate, and time-to-fill for recruiting.
If your systems cannot produce evidence, fix that before blaming the employee. A salesperson cannot be judged on conversion if leads are not tagged. An engineer cannot be judged on delivery if priorities change every week. A customer success manager cannot own churn when the product is unusable.
Use a four-point rating scale
Do not use a five-point scale. It invites everyone to choose three, which means “I do not want to have this conversation.” A four-point scale forces a decision.
| Rating | Meaning | Management action |
|---|---|---|
| 1, Below expectations | Important outcomes were missed, or behaviour created material risk. | Written improvement plan with a 30-60 day checkpoint. |
| 2, Partly meets expectations | Some outcomes landed, but performance was inconsistent or required heavy supervision. | Specific correction plan and closer weekly follow-up. |
| 3, Meets expectations | The person delivered the role’s agreed outcomes at the expected level. | Continue scope; set harder or broader goals. |
| 4, Exceeds expectations | The person delivered materially beyond the role and improved the company’s capacity. | Recognise clearly; consider more scope, pay, or ESOP review. |
“Exceeds expectations” should be rare. If half your team receives a 4, your expectations are too low or your ratings are too generous. Investors care about this because inflated reviews create inflated salary expectations. Then your burn rises without a matching increase in output.
Do not force a distribution. A seed company with eight employees does not need one person in every rating bucket. Forced ranking is especially stupid when roles are different and sample sizes are tiny.
Prepare the review in five working days
A good review is prepared, not improvised. Give yourself and the employee enough time to bring evidence.
- Send the role scorecard and review period. State the dates, goals, and format. If goals changed, list the changes and why.
- Ask for a self-review. Request three achievements, two misses, evidence for each, and the support needed next quarter.
- Collect manager evidence. Use metrics, work samples, customer feedback, and examples of decisions. Do not collect gossip.
- Ask for focused peer input. Speak to two or three people who worked closely with the employee. Ask what the person should continue, stop, and start doing.
- Write your assessment before the meeting. Include the rating, reasons, and next-period goals. Do not decide in the room because the employee is likeable.
Keep peer feedback private where appropriate, but do not hide behind anonymous comments. If feedback is serious enough to affect a rating, you must be able to explain the behaviour and its effect.
For a small team, the founder should review managers and senior individual contributors. Once you cross roughly 15-20 people, add a calibration meeting. Managers should compare standards and catch obvious bias. The founder still owns the final call on senior roles.
Run the conversation without turning it into a trial
Block 60-90 minutes. Meet privately. Put the phone away. Start with the purpose, not the rating.
A practical structure:
- Set the frame. “We are reviewing your work against the role goals for the last six months. This is a two-way discussion. We will finish with your next goals and support required.”
- Ask for the employee’s view first. “What are you proud of? What did not go well? What blocked you?”
- Discuss results. Start with what happened, then explain why it mattered.
- Discuss working behaviour. Use specific examples. “You missed three handoffs without warning” is useful. “You need to be more professional” is not.
- Share the rating and reasoning. Do not bury it after 45 minutes of compliments.
- Listen to disagreement. Ask for evidence. Change your view if the evidence changes. Do not change it merely because the employee is upset.
- Agree on the next period. Set three to five goals, measures, deadlines, and decision rights.
- Write the record. Send a short summary within 24 hours. Ask the employee to acknowledge receipt, not necessarily agreement.
Use the SBI format for difficult feedback: situation, behaviour, impact.
“In the March enterprise launch, you changed the pricing logic without informing finance. The invoice batch had to be reversed and collections were delayed by nine days. From now on, pricing changes need written approval from finance and product.”
That is uncomfortable. It is also fair.
Do not mix performance, salary, and ESOPs into one confused meeting
Compensation belongs near the review, but it should not control the review.
If you announce a salary increase before discussing performance, the employee will remember only the number. If you discuss performance and salary in the same meeting, every piece of feedback will sound like a negotiation tactic.
Run the performance conversation first. Make compensation decisions separately, ideally within two to four weeks. Explain the factors:
- Performance against the role.
- Change in role scope.
- Internal pay consistency.
- Cash runway and the company’s financial position.
- Market data for the role and location.
Do not promise a raise because an employee performed well if the company cannot afford it. Say the truth: “Your performance is strong. We cannot make the full cash adjustment this month. Here is what we can do, and here is when we will review it.”
ESOPs need even more care. A performance review is not a substitute for a grant. Explain vesting, cliff, exercise price, dilution, and the board or shareholder approvals that apply to your company. Review your cap table before making ESOP promises. Casual verbal grants create legal and trust problems later.
Do not use ESOPs to avoid paying a fair salary to a critical employee. Most early options are illiquid and may never produce value. Treat them as upside, not cash.
What to do when the rating is poor
A low rating is useful only if it leads to a decision. Do not place someone on an indefinite “watch”. That is cowardice disguised as patience.
Write a performance improvement plan with:
- Two or three specific gaps.
- The required standard.
- Evidence that will prove improvement.
- Support the company will provide.
- A 30-day or 60-day review date.
- The consequence if the standard is not met.
Example:
“Customer escalations were not acknowledged within the agreed four-hour window in seven of the last 20 cases. For the next 30 days, acknowledge every escalation within four working hours, maintain a daily queue review, and reduce unresolved cases older than 48 hours to fewer than five. We will provide a support dashboard and weekly coaching. If this standard is not met, we will reassess your fit for the role.”
Do not create an impossible plan after you have already decided to terminate someone. That is unfair and may create employment-law risk. If the role is not working, follow your employment agreement, maintain proper documentation, and obtain qualified Indian legal advice before taking action.
Also check whether the problem is really performance. Common founder failures include:
- Hiring someone for sales and giving them no qualified leads.
- Hiring an engineer and changing priorities every three days.
- Promoting the best individual contributor into management without training.
- Giving two people overlapping authority and blaming both for conflict.
- Setting a target that ignores cash collection, gross margin, or regulatory limits.
A review is not a ritual for assigning blame. It is a diagnostic of the entire operating system.
What your first review reveals about you
If employees cannot state their goals, you did not communicate them.
If every goal changed without a written reset, your planning process is weak.
If one manager rates everyone highly and another rates everyone poorly, you have a calibration problem.
If the same employee is praised for “ownership” but punished whenever they make a decision without approval, you have a founder-control problem.
If strong employees receive no increase in scope, recognition, or reward, they will eventually leave. In India, they can find another startup role quickly, especially in engineering, product, growth, and B2B sales. Your brand will not retain them forever.
Track review quality with a few operating measures:
- Percentage of employees with written goals.
- Percentage of reviews completed on time.
- Number of goals changed during the period, with reasons documented.
- Regretted attrition among high performers.
- Completion rate of improvement plans.
- Employee clarity score: can people state their top three priorities?
Do not run an employee satisfaction survey before you can answer these basic questions. You will collect feelings while ignoring the operating failures that caused them.
Your first-review checklist for this week
You do not need HR software. A shared document, a calendar, and disciplined follow-through are enough for your first 10-15 employees.
- List every employee and the date they started their current role.
- Write a one-page scorecard for each role.
- Ask each employee for a self-review by a fixed date.
- Collect evidence from the systems you already use.
- Hold a 60-90 minute review meeting.
- Record a four-point rating and three next-period goals.
- Separate salary and ESOP discussions from the performance conversation.
- Schedule the next one-to-one and review checkpoint before closing the process.
For guidance on choosing the right support around your company, see our practical guide on finding a startup mentor in India. A good mentor should challenge your management decisions, not merely approve your founder instincts.
Bottom line
Your first performance review is not mainly about judging employees. It is about proving that your company can turn strategy into clear work, measure the work fairly, and make hard decisions without hiding behind founder mood. If the review produces vague praise, arbitrary ratings, or a salary bargain, you have built a ceremony. Build the system first. Then judge the person.
FAQs
When should a startup conduct its first performance review in India?
Conduct it after roughly six months in the role, once the employee has completed two meaningful quarters. Review senior hires at 90 days and six months. Use project-end reviews for interns and short-term contractors.
Should startup performance reviews be annual or quarterly?
Use monthly one-to-ones and a written review every six months for a team of up to 10 people. Move to quarterly written check-ins as the team grows. Annual-only reviews are too slow for a seed-stage company.
What should be included in a startup performance review?
Include the role mission, agreed outcomes, evidence of results, work quality, ownership, team behaviour, a four-point rating, next-period goals, support required, and a review date. Keep salary and ESOP decisions separate.
What rating scale should an early-stage startup use?
Use four ratings: below expectations, partly meets expectations, meets expectations, and exceeds expectations. Four points prevent the safe middle rating that makes reviews meaningless.
Can a founder put an employee on a performance improvement plan?
Yes, but the plan must identify specific gaps, measurable standards, support, a 30-60 day timeline, and the consequence of missing the standard. Check the employment agreement and take qualified Indian legal advice before termination.
Should performance reviews decide salary increases?
Performance should inform compensation, but do not turn the review into a negotiation. Discuss performance first and compensation separately, considering results, role scope, internal fairness, market pay, and runway.

