Cost Examples, TCO and Selection Checklists
Payback on switching payroll: from spreadsheets or a bureau to software, worked honestly
How to calculate payback when moving from manual or outsourced payroll to software in India: one-time costs, net yearly saving, two worked examples and what not to count.
Switching payroll pays back when the net yearly saving (what you pay today minus the new software cost minus any extra in-house time) covers the one-time switching cost: setup, migration, a parallel run, training and devices. In the illustrative examples below, a 60-person business leaving a bureau recovers ₹80,000 of switching cost in about eight months, and a 40-person business leaving spreadsheets recovers ₹40,000 in about eight months too. The method matters more than the result: count only real savings, include the parallel-run months and do not count penalties you have never paid.
This guide belongs to the best-fit selection method. Use the switching payback calculator with your own figures.
The formula
- Net one-time cost = setup + migration + parallel run effort + training + devices − written waivers.
- Net yearly saving = current yearly cost − new software cost per year − extra in-house time per year.
- Payback (years) = net one-time cost ÷ net yearly saving (the calculator also grows the saving each year if you enter a growth rate).
- Net saving over the horizon = sum of yearly savings − net one-time cost.
Example 1: from an outsourced bureau to software
A 60-employee business pays a bureau ₹400 per employee per month (illustrative).
| Line | Calculation | Amount |
|---|---|---|
| Current bureau cost | 60 × ₹400 × 12 | ₹2,88,000 a year |
| New software | 60 × ₹150 × 12 | ₹1,08,000 a year |
| Extra in-house time | 10 h × ₹500 × 12 | ₹60,000 a year |
| Net yearly saving | ₹2,88,000 − ₹1,08,000 − ₹60,000 | ₹1,20,000 |
| One-time cost | Setup ₹30,000 + migration ₹20,000 + parallel run ₹20,000 + training ₹10,000 | ₹80,000 |
| Payback | ₹80,000 ÷ ₹1,20,000 | 0.67 years (about 8 months) |
Over three years with the saving growing 5% a year (more employees, fee inflation avoided): ₹1,20,000 + ₹1,26,000 + ₹1,32,300 = ₹3,78,300; minus ₹80,000 = ₹2,98,300 net saving. These are the calculator's default figures.
Example 2: from spreadsheets to software
A 40-employee business runs payroll in spreadsheets. The accountant spends about 30 hours a month on attendance, calculations, challans and returns, at a loaded cost of ₹400 an hour.
| Line | Calculation | Amount |
|---|---|---|
| Current time cost | 30 h × ₹400 × 12 | ₹1,44,000 a year |
| Time with software | 8 h × ₹400 × 12 | ₹38,400 a year |
| Time saved | ₹1,05,600 a year | |
| New software | 40 × ₹100 × 12 | ₹48,000 a year |
| Net yearly saving | ₹1,05,600 − ₹48,000 | ₹57,600 |
| One-time cost | Setup ₹20,000 + cleaning masters ₹10,000 + parallel run ₹10,000 | ₹40,000 |
| Payback | ₹40,000 ÷ ₹57,600 | 0.69 years (about 8 months) |
The time saving is real only if the accountant's freed hours go to other work or remove overtime or outside help. If the hours simply disappear into the day, the cash saving is zero, though the risk reduction may still justify the switch.
How sensitive is the answer?
Change one assumption at a time in Example 1 to see which matters most (all illustrative):
| Change | Net yearly saving | Payback |
|---|---|---|
| Base case | ₹1,20,000 | 0.67 years |
| In-house time 20 h a month instead of 10 | ₹60,000 | 1.33 years |
| Software at ₹200 PEPM instead of ₹150 | ₹84,000 | 0.95 years |
| One-time cost ₹1,20,000 instead of ₹80,000 | ₹1,20,000 | 1.00 year |
| Bureau fee ₹300 PEPM instead of ₹400 | ₹48,000 | 1.67 years |
The in-house time and the current bureau fee move the result most. Measure both for two or three months before deciding; do not rely on a guess.
What to count and what not to
| Count | Do not count (or count separately) |
|---|---|
| Bureau or CA fees you will stop paying | Penalties and interest you have not actually been paying |
| Overtime or temporary help you will no longer need | "Productivity" without a named use for freed time |
| Hours redeployed to named work | Savings from features you will not switch on |
| Fewer bank and journal corrections, if you have tracked them | Vendor ROI claims |
If you have paid interest or damages for late PF, ESI or TDS in the past year, you can show that figure separately as risk reduction. Keep it out of the headline payback.
One-time costs people forget
- The parallel run: at least one month of doing payroll twice.
- Cleaning employee masters: UAN, ESI numbers, PAN, bank details, joining dates.
- Restructuring salaries for the 50% wage rule, if not done.
- Year-to-date data from the previous system or bureau, which some providers charge to export.
- Biometric devices and their integration.
- Employee communication and app onboarding.
See the migration guide.
When a switch does not pay back
- Very small payrolls (under about 10 employees) where software minimums exceed what you pay a CA.
- Businesses without anyone to own payroll in-house; the software will be underused.
- Moving to software that fails your statutory tests; corrections will eat the savings.
In these cases compare the alternatives in outsourcing versus software versus in-house.
Frequently asked questions
What is a good payback period?
Under a year is common for switches from bureaus or heavy spreadsheet work at MSME scale, but your figures decide. Over two years suggests the case rests on risk reduction rather than cost.
Should we include GST?
If you can claim input tax credit, compare before GST. If not, include GST on both the old and new costs.
Does the saving grow over time?
Often, as headcount grows and per-head software prices fall relative to bureau fees. Enter a modest growth rate, or zero to be cautious.
Are these examples real businesses?
No. They are illustrations to show the arithmetic.
Sources
- EPFO — interest and damages for late PF deposits
- ESIC — ESI deposit rules
- Income Tax Department — TDS deposit and return obligations
- CBIC GST — GST on software and services
