Technology · Digital agency · Mumbai and Bangalore
100+ hours back to HR, every review cycle.
The short version
SoCheers is a full-service digital agency across Mumbai and Bangalore, and it was growing fast enough that its performance review process had become a logistics exercise. The HR team was spending its cycle sharing draft KPI sheets, chasing managers by email, retrieving past reviews and collating ratings into spreadsheets. They had looked at HR platforms and found nothing that could be shaped to how they actually reviewed people. We built one on Zoho instead. It returns more than100hours to HR every cycle and another45across the team heads, and it has been running since 2019 at what is now twice the headcount.
The reviewing was never the problem
Ask an HR team what performance review costs them and they will describe a season rather than a task. Weeks of it, twice a year, absorbing the function almost entirely.
But when you list what those weeks actually contain, very little of it is review. It is emailing a template. Chasing the manager who has not filled it in. Finding what somebody was rated eighteen months ago because it came up in the conversation. Copying twenty-five ratings into a spreadsheet so they can be averaged. Sending the same reminder for the fourth time.
The judgement in a performance review takes a manager perhaps an hour per person. Everything around it takes the organisation weeks. Automating the judgement would be both difficult and undesirable. Automating the coordination is neither.
That distinction set the scope. Nothing we built forms an opinion about anybody. It moves paper, enforces who may touch what, calculates what was always going to be calculated, and remembers.
Why nothing off the shelf fitted
SoCheers had looked. The market for HR platforms is not short of performance modules, and the problem with all of them was the same: a review process is one of the most company-specific things an organisation has, and these products are built around somebody else’s.
How many levels approve. Whether a manager can edit what a reviewee submitted or only comment on it. How a rating is composed from KPIs. Who joins a cycle when they started in March. What a team head can see about somebody else’s team.
Each of those is configurable in an expensive enterprise product and fixed in an affordable one. And a review process bent to fit software is a review process people stop trusting — which matters more here than in most systems, because an untrusted review platform is worse than a spreadsheet. It has the appearance of rigour without the substance.
So we built on the Zoho ecosystem, where the data entry, the reporting and the dashboards could be shaped to their process, and where the same portal could later carry other HR tools rather than adding another login.
Who can see what, and who can change it
The permission model is the real design in a review system, and it is where most of them lose their users.
| Reviewee | Submits their own performance data for the cycle. Sees their own record and history. |
| Manager | Verifies or edits the submission, and adds feedback at KPI level or overall. |
| Manager’s manager | Approves the completed review. Nobody outside these three can change or approve it. |
| HR | Starts and stops the cycle, controls eligibility by joining and exit date, tracks status per employee and sends follow-ups from inside the platform. |
| Management | Sees the picture, filtered by review period, without the ability to reach into an individual review. |
Two details there do more work than they look like they should. HR controlling eligibility by joining and exit date means somebody who started in March is handled by the system rather than by a note in somebody’s calendar. And follow-ups being sent from within the platform is the single largest source of the hours saved — the chasing never leaves the place where the status lives, so nobody has to reconcile a mailbox against a tracker.
What it saved
| Per cycle | Previously spent on | |
|---|---|---|
| HR team | 100+ | Follow-ups, data consolidation, running and tracking the cycle |
| Team heads | 45+ | Consolidating team data, retrieving past reviews, maintaining spreadsheets, emailing feedback |
Since the platform went live the SoCheers team has roughly doubled. The saving is per cycle and scales with headcount, so it has grown — and the platform has not been rebuilt to accommodate any of it.
How to read this
These are hours per cycle, not per year. We have deliberately not annualised them, because the multiplier is the number of review cycles a company runs and that varies. Apply your own: at two cycles a year the figures above return roughly 290 hours to the business annually, at four they return nearly 600.
The figures are the client’s estimate. They came from the SoCheers team comparing the cycle before and after and were reported to us. Nobody ran a time-and-motion study, and we did not audit them.
Hours saved are not the same as value created. What HR did with the returned time is the part that matters and we have no visibility of it. A saving that gets absorbed rather than redeployed is a smaller result than it looks.
Nothing here improves review quality. The platform removes coordination cost. Whether the reviews themselves became better conversations depends on the managers having them, and that was outside this engagement.
Constraints that shaped the work
Building custom is the right call when a process is genuinely company-specific and the wrong one most of the time. We have told other clients not to — on a logistics engagement in the same period we recommended configuring rather than building, because their operation was unusual in scale rather than in shape. Review processes are unusual in shape, which is what tips it. The test is not whether an off-the-shelf product is imperfect. It is whether bending your process to fit it would cost you the trust of the people using it.
We also measured the wrong thing, or rather only the easy thing. Hours saved was straightforward to estimate and it is an input measure. Cycle completion time, the proportion of reviews finished without a chase, how long after the cycle closed before ratings were final — all of those would have said more about whether the process improved, and all were sitting in the system we built.
If your review season eats a month
Write down where the hours actually go before you buy anything. Almost none of them will be review. They will be template distribution, chasing, retrieval of old records and manual collation — four problems, all of them coordination, none of them requiring judgement.
Then be careful what you bend. A review process reshaped to suit a product is one people quietly stop believing in, and a performance system nobody trusts does more damage than the spreadsheet it replaced.
Read another
Every figure sourced from the client’s own systems, with the basis stated and the exclusions named.Bawarchi Atlanta
Smile Loft Dental
Sanpeggio’s Pizza
Measurement note. Scope, platform design and delivery are from our own project records. Hours saved per review cycle are the client’s own estimate, reported to us by the SoCheers team, and were not independently measured or audited by us. Team growth since 2019 is as described by the client.
