When should a company buy performance management software? A company should buy performance management software once it has a review process it can describe on one page and has run at least once, and the work of running that process by hand has started to cost more than a tool would. In practice that point arrives when three things are true together: several layers of managers have to be compared fairly in calibration, HR can no longer find the evidence behind a rating quickly, and the cycle is moving from once a year to quarterly or six-monthly. Headcount alone is a weak signal. A 120-person company with quarterly reviews and four departments may need software sooner than a 300-person company that reviews once a year in one office. Buying before the process exists usually means paying to automate confusion.
Key takeaways
- Process first, tool second. If you cannot write down the rating scale, the cycle dates and who decides, software will not decide it for you.
- Watch for triggers, not a headcount number. Multi-team calibration, a faster cycle, scattered evidence, data requests and a merge that takes days are the signals that matter.
- Manual reviews are expensive in time. CEB research reported by SHRM (2015) put manager time at about 210 hours a year on performance management activities, and about 40 hours for each employee.
- Spreadsheets fail quietly. Panko’s review of spreadsheet audits (1998) found errors in a few per cent of cells, so in a large review workbook the question is how many errors there are, not whether there are any.
- Data rights do not wait for your tooling. Under UK GDPR, the ICO (2026) expects a reply to a subject access request within one month, after a reasonable and proportionate search. Review notes spread across files make that hard.
Seven triggers that it is time to buy
Most People teams do not decide to buy software in a planning meeting. They decide in the week before calibration, when the master spreadsheet breaks. It is better to check for the triggers ahead of time. Two or more “yes” answers in the table below is a reasonable point to start a structured evaluation; one is usually a process fix.
| Trigger | What you see | Can a process fix wait it out? |
|---|---|---|
| Calibration across several teams | Department heads compare ratings from different managers, and nobody trusts that a “4” means the same thing everywhere | Briefly. A written rating scale helps, but comparing distributions by hand gets slower with every team you add. |
| A faster cycle | You are moving from annual to six-monthly or quarterly reviews, so the admin repeats two to four times a year | Rarely. Each extra cycle multiplies the copying, chasing and merging. |
| Evidence you cannot find | A rating is challenged and the 1:1 notes, goals and peer comments are in different places, or in a manager’s notebook | No. This is a records problem, and it gets worse as managers change roles or leave. |
| Data requests and access control | An employee asks for their data, or you realise half the company can open the review workbook | Not safely. Locking down shared files is possible but fragile. |
| The merge takes days | Someone in HR spends more than a day stitching manager files together before calibration | For one cycle, perhaps. That time is the clearest cost you can measure. |
| Goals drift from reviews | Goals are set in January in one tool and scored in December from memory in another | Partly. A shared goals sheet helps; linking goals to the review form by hand rarely survives a reorganisation. |
| Managers stop completing on time | Completion rates fall and chasing becomes a weekly task for HR | Sometimes. Check whether the form is too long before blaming the tooling. |
These map closely to the signs in the guide to switching from spreadsheets to performance management software, which also covers the data you need before migrating and a 10-week plan. This post is about the earlier question: whether to start at all.
What the evidence says about running reviews by hand
The time cost is large and mostly invisible. In 2015 SHRM reported CEB survey findings that managers spent an average of 210 hours a year on performance management activities and that each of their employees spent about 40 hours (SHRM, 2015). That was a US survey, and software does not remove all of those hours. But it shows where the cost sits: in managers’ and employees’ time, not in the licence. The same year, Deloitte’s Marcus Buckingham and Ashley Goodall wrote in Harvard Business Review (2015) that their firm had counted close to 2 million hours a year going into performance management. If you want a figure for your own company, the performance review cost calculator works it out from your headcount, manager count and cycle length.
Spreadsheets carry a built-in error rate. Raymond Panko’s paper What We Know About Spreadsheet Errors (Journal of End User Computing, 1998) reviewed field audits and experiments and concluded that errors seem to occur in a few per cent of all cells. A review workbook with hundreds of employees, several rating columns and lookup formulas is exactly the kind of file where a wrong row or a broken formula goes unnoticed until someone is paid or promoted on it.
Data protection sets a deadline. The ICO’s guide to subject access (updated 16 July 2026 for the Data (Use and Access) Act 2025) says organisations must respond without undue delay and within one month, must make a reasonable and proportionate search, and should have information management systems that let them find and retrieve information efficiently. In Ireland, the Data Protection Commission’s page on the right of access explains that Article 15 GDPR gives employees a copy of the personal data held about them, including the retention period. Review ratings, comments and 1:1 notes are personal data. If they live in a dozen spreadsheets and inboxes, every request becomes a search exercise.
Software does not fix a review people do not value. Gallup reported in 2019 that only 14% of employees strongly agree their performance reviews inspire them to improve (Gallup, 2019). Moving a weak process into a tool produces the same weak reviews, faster. That is why the decision starts with the process, not the vendor list.
Why the timing is hardest at 100 to 1,500 people
Below about 100 people, one People lead can often hold the whole cycle in their head. Above a few thousand, the HRIS and an enterprise performance module are usually already in place. Companies in between tend to hit the triggers all at once:
- The second layer of management arrives. Once managers report to managers, the founder can no longer sanity-check every rating, and performance calibration becomes a real meeting rather than a chat.
- The People team is small. One or two people run reviews alongside hiring, payroll queries and policy. A spreadsheet merge that takes three days is three days of nothing else.
- The first data request or dispute lands. Usually after a difficult exit, and usually when the notes are hardest to find.
- The board asks for numbers. Goal completion, rating spread by department and regretted attrition are hard to produce from files that were never designed to be added up.
- The tool question gets tangled with the HRIS question. If you are also reviewing your HR system, the HRIS switch checker helps separate the two decisions, and the year-end performance and HRIS guide covers the timing around year-end.
When you should not buy yet
Being honest about this saves money. Wait if any of these apply:
- You have never run a full cycle. Run one on a template first. The review cycle checklist and a shared folder are enough to learn what your process actually needs.
- The rating scale and the decision owner are still disputed. Settle them on paper. A tool will faithfully reproduce the argument.
- You are a few dozen people with one review a year. The switching guide makes the same point: spreadsheets can still work when one person can see the whole cycle and access is tight.
- No one will own the system. Someone needs time to configure it, train managers and keep the data clean. Without an owner, adoption fades within two cycles.
If you are unsure which side you are on, the two-minute performance health check scores standards, goal alignment, bias and admin burden, and the admin score is a good proxy for whether tooling is the bottleneck.
Our perspective
The companies I speak to rarely buy too early. Far more often they buy a cycle too late, in October, with calibration six weeks away and a spreadsheet nobody trusts. That is the worst time to choose: there is no room to pilot, managers meet the new tool for the first time in the middle of year-end, and the People team ends up running two systems at once.
The question I ask is simple: in your last cycle, where did the hours go? If the answer is writing and talking, that is the work reviews are for, and software will only shave the edges. If the answer is copying, chasing, merging and searching, the process is ready and the tooling is what is failing it. That is the moment to buy, and ideally at least one quarter before the next cycle opens.
— Ash Taherian, Product Sales Executive, Revolut People
Where Revolut People fits, and where it doesn’t
I work at Revolut People, so read this section with that in mind. The Performance plan covers goals, scorecard reviews, calibration, 360 feedback, skills and values scorecards, 1:1s and growth plans on one employee record, so the evidence behind a rating sits next to the rating. Public list prices on the pricing estimator start at £10 per employee per month for Performance, billed annually, with no standard implementation fees. The product tour shows reviews, calibration and 1:1s in short films of under 40 seconds each, and the security and access one-pager covers the permission groups that keep ratings and salary data apart.
The limits matter for the timing question. There is a minimum annual spend of £5,000 in the UK and EEA, so at the £10 list price a company below roughly 42 employees pays for more seats than it has; for those teams, a well-run template is often the better choice for now. Software will not write your rating scale, run the calibration conversation or make a reluctant manager give honest feedback. And if you already have an HRIS whose performance module you are happy with, adding a separate tool may not be worth the extra integration. The case is strongest when the triggers above are stacking up and you want goals, reviews and calibration in one place before the next cycle.
If you are weighing up whether this cycle or the next is the right time, book a 15-minute review. Bring your cycle dates and your current review template, and I will tell you honestly if it is too early.
Step plan: a 30-day buy-or-wait decision
- Week 1: write the process on one page. Cycle dates, rating scale, who reviews whom, how calibration works and who signs off. If you cannot, stop here and fix that first.
- Week 1: count the hours. Ask HR and three managers how long the last cycle took, split into writing and talking versus copying, chasing and merging. Put the numbers into the cost calculator.
- Week 2: score the seven triggers. Use the table above. Two or more is a case for evaluation; one is a process fix.
- Week 2: check your stack and data. List the HRIS, single sign-on and goal tools any new system must connect to. The stack-fit checker shows which have a documented Revolut People integration; ask every vendor the same question.
- Week 3: build a short scorecard. Use the performance management software buyer’s guide to weight process fit, calibration, evidence, integrations and reporting, and check published prices on the performance management software cost guide.
- Week 4: decide, and set a date. If you buy, work back from the day your next cycle opens with the go-live planner so there is time for a pilot. If you wait, write down which trigger would change your mind and review it next quarter.
Once you have decided, the performance management rollout checklist covers owners, manager adoption and the pitfalls of the first cycle.
Bottom line
Buy performance management software when the process is clear and has been run at least once, and when the hand-work of running it is the thing slowing you down. Multi-team calibration, a faster cycle, evidence you cannot find, data requests and a merge that takes days are the triggers to watch; headcount on its own is not. Do not buy to settle an argument about the rating scale, and do not buy in the six weeks before calibration. Decide a quarter ahead, pilot with two teams, and give the system a named owner.
FAQ
What features should performance management software have?
At minimum: goals that link to the review, a configurable review form and rating scale, calibration views that compare ratings across managers and teams, a record of 1:1 notes and feedback, role-based access that keeps performance and salary data separate, and reporting you can export. For UK and Irish teams, also check data retention controls and where data is hosted. The buyer’s guide turns these into a weighted scorecard.
How do you choose performance management software?
Start from your written process, not from vendor sales presentations. Shortlist two or three tools that match it, score them on the same criteria, check integrations with your HRIS and single sign-on, and run a pilot with two job families before a full rollout. Ask each vendor for a written quote at your real headcount, including any minimum spend and implementation fees.
How much does performance management software cost?
Most tools are priced per employee per month, often billed annually, and some add a minimum annual spend or implementation fees. Published prices vary widely and several vendors only quote on request. The 2026 pricing guide lists dated public prices and notes where a vendor does not publish them.
Can you run performance reviews in spreadsheets?
Yes, for a small company with one cycle a year, a handful of managers and tight file access. The risks grow with size: formula and copy errors, uncontrolled access to ratings, version confusion before calibration and slow responses to data requests. The switching guide lists the signs that a spreadsheet has stopped working.
Sources
- SHRM, Dana Wilkie: Is the Annual Performance Review Dead? (19 August 2015) - CEB survey: managers about 210 hours a year on performance management activities; employees about 40 hours
- Marcus Buckingham and Ashley Goodall, Harvard Business Review: Reinventing Performance Management (April 2015) - Deloitte counted close to 2 million hours a year on performance management
- Raymond R. Panko, Journal of End User Computing 10(2): What We Know About Spreadsheet Errors (1998) - errors in a few per cent of all cells across audits and experiments
- Information Commissioner’s Office: A guide to subject access (last updated 16 July 2026) - one-month response, reasonable and proportionate search, information management systems
- Data Protection Commission (Ireland): The Right of Access (Article 15 GDPR, 2016/679; page accessed 18 November 2026) - what an access request covers, including retention period
- Robert Sutton and Ben Wigert, Gallup: More Harm Than Good: The Truth About Performance Reviews (6 May 2019) - 14% of employees strongly agree their reviews inspire them to improve