Almost every company runs its first performance reviews in spreadsheets and documents. It is fast, free and flexible. Somewhere between 100 and 300 employees, it usually stops working: version conflicts, broken formulas, manual merges before calibration, and nobody able to see one person’s goals, feedback and review history in one place. This guide explains how to tell when it is time to switch, what to fix before you buy anything, how to choose software, and a practical migration plan.

Before reading further, the HRIS switch checker gives a structured read-out of where your current setup holds up and where it creates friction, and the review cost calculator shows how many hours your current cycle takes.

Signs spreadsheets have stopped working

  • Calibration needs a merge. Someone spends a day combining managers’ sheets before every calibration meeting, and errors creep in.
  • Nobody trusts the version. There are several copies of the ratings file, and people ask which one is final.
  • Evidence lives elsewhere. Goals are in slides, feedback in chat, 1:1 notes in personal documents; the spreadsheet holds only the rating.
  • History is hard to find. Pulling someone’s last three cycles takes an afternoon.
  • Access control is a worry. Sensitive ratings sit in files that are easy to share by mistake.
  • Reporting is manual. Every leadership report is built by hand from exports.

McKinsey’s 2024 article “In the spotlight: Performance management that puts people first” lists fragmentation and informal “shadow” systems among the common reasons leaders lack confidence in performance management. Spreadsheet-based reviews are often both.

Fix the process before you buy the tool

Software makes a good process faster. It does not make an unclear process fair. Before you evaluate vendors, make sure three things exist on paper:

  1. A written standard. A talent bar by seniority and scorecards for at least your largest job families.
  2. A defined cycle. Dates, steps and owners, as in the review-cycle checklist.
  3. A calibration method. Who attends, what they check and how changes are logged, as in the calibration meeting guide.

Run at least one cycle with these in your spreadsheets first if you can. You will learn what you actually need from software, and you will be configuring a process you trust rather than inventing one inside a vendor’s defaults.

What to require from performance management software

RequirementWhy it matters when leaving spreadsheetsHow to test it
Your own scorecards and rating scaleYou should not have to change your standard to fit the toolBuild one of your real scorecards during the trial
Goals, feedback, 1:1 notes and reviews in one placeRemoves the evidence hunt that makes reviews slowAsk a manager to prepare a review using only the tool
Calibration viewsReplaces the pre-calibration mergeRun a mock calibration with sample, anonymised ratings
Change log and approvalsEvery rating change recorded with a reasonChange a rating and check what is recorded
PermissionsSensitive data visible only to the right peopleLog in as a manager and check what they can see
Import and exportBring history in; never be locked inImport a sample sheet; export a full cycle
ReportingLeadership reports without manual workRebuild your current leadership report

The performance management software buyer’s guide expands these into a full evaluation scorecard.

When spreadsheets are still fine

Switching is not always the right answer. Spreadsheets can work well when the company is small enough that one person can hold the whole cycle in view, typically a few dozen people with a handful of managers; when reviews happen once or twice a year and calibration is a single meeting; and when the People team has the time and discipline to keep one master file with tight access. If that describes you, invest in the process first (a written bar, scorecards, a clear cycle) and revisit software when the merge before calibration starts taking more than a day, or when you plan to move to quarterly reviews.

Data checklist before migration

  • One current list of employees with a unique ID, manager, level, job family, department and start date.
  • Manager relationships checked against the org chart, including dotted lines that should give review input.
  • Final grades for the cycles you plan to import, in one consistent scale.
  • Current goals for every employee, or a decision to set new goals in the system.
  • A list of who should see what: People team, department heads, managers, employees.
  • A named owner for data quality during and after the switch.

A 10-week migration plan

An illustrative plan for a company of 150 to 400 people, timed to finish before the next review cycle opens:

WeeksWorkstreamOutput
1–2Confirm requirements, shortlist and run trials with your real scorecardsChosen vendor and signed plan
3–4Clean data: employee list, managers, levels, job families; decide what history to importClean import files
4–5Configure scorecards, rating scale, review cycle and permissionsConfigured system
5–6Import data; test with the People team and two pilot managersIssues list resolved
7Set goals for the coming period in the new systemGoals live for everyone
8Manager training (30–45 minutes) and employee announcementManagers ready
9–10Buffer, then first feedback and 1:1 notes captured in the systemEvidence building before the cycle

Durations are illustrative; ask each vendor for a dated plan based on your headcount and data. The key principle is to go live with goals and feedback before the first review, so the first cycle on the new system already has evidence in it.

What to do with historical spreadsheets

You rarely need every old file. Import the last one or two cycles’ final grades and promotion history so managers can see trends. Archive the rest in a restricted folder with a retention date. Do not import drafts or unofficial copies; they create confusion about which rating was final.

Getting managers to actually use it

  • Make it the only place. If spreadsheets remain an option, some managers will keep using them. Set a date after which reviews are accepted only in the new system.
  • Show the time saving. Show managers how a review draws on goals and feedback already in the system. Published results such as Ebury’s calibration falling from 4 hours to 1 are useful illustrations.
  • Start with 1:1s and feedback. Weekly use builds the habit before review season.
  • Support the first cycle. Hold drop-in sessions during the review window.

Worked example

An illustrative 220-person company runs reviews in one spreadsheet per department, merged by the People lead before calibration. Its last cycle took seven weeks, two of them spent merging and fixing sheets. Using the cost calculator, the People lead estimates 1,320 manager hours a year on writing alone (220 reviews × 3 hours × 2 cycles). The team writes scorecards for its four largest job families, runs one more cycle in spreadsheets with them, then migrates over ten weeks between cycles. The first cycle in software runs in about four weeks, because calibration views replace the merge and managers draw on goals and notes already in the system.

Common mistakes

  • Recreating the spreadsheet in software. Use the switch to simplify, not to digitise every old column.
  • Migrating mid-cycle. Switch between cycles.
  • Skipping permissions testing. Check what each role can see before anyone else logs in.
  • Buying before the standard exists. The tool cannot tell managers what good looks like.

For budgeting, the pricing estimator shows indicative Revolut People list pricing (Ash works for Revolut People; compare with other vendors directly), and the full method is in the performance management framework.

Sources: McKinsey, In the spotlight: Performance management that puts people first (2024); CIPD, Performance management factsheet. Migration plan and figures are illustrative.

Use with this guide