Spreadsheet, Provider or HR System: Three Ways to Run PeopleOps

Spreadsheet, Provider or HR System: Three Ways to Run PeopleOps

Every growing company reaches the same awkward moment. The founder’s spreadsheet of staff details has become unreliable, somebody has asked who approved a raise, and a leaver turns out to still have a working login. At that point you need a working method for PeopleOps that does not hinge on what a single person happens to remember. There are three realistic options, and this piece compares them without hype.

What you are actually choosing between

PeopleOps here means running the employee lifecycle, from hiring through onboarding, role changes, pay, performance and exit, as an operation with owners and written service levels. The reference page behind this comparison, the Atlas page on this discipline, boils it down to three working parts: a ledger of effective-dated records, a set of owned state changes, and a service desk with a catalogue of what people can ask for.

Any option you pick has to supply those three parts. The comparison below checks each option against them.

Option one: shared spreadsheets and email

This is where nearly everyone starts, and for a team of a dozen it can be fine. It costs nothing, everyone understands it, and changes are instant.

Where it holds up: a handful of staff, one location, one person who cares. A tab per topic and a habit of dating entries covers the basics.

Where it breaks: there is no history you can trust, because anyone can overwrite a cell. There is no event trail, so a resignation lives in an email thread. Two copies drift apart. Privacy is also hard to control, since a pay tab shared widely exposes everyone. And the moment a second person starts editing, you lose the single trustworthy version.

The Atlas treats the effective-dated record as the foundation, and a spreadsheet can imitate it only through discipline, not through design.

Option two: a payroll provider or employer-of-record service

The second route is to hand the heavy lifting to an outside firm that runs payroll, contracts and statutory filings. The Atlas notes that very small companies, roughly under fifty people, typically lean on a payroll provider, a professional employer organisation or an employer-of-record arrangement alongside one internal owner.

Where it holds up: compliance in unfamiliar places, accurate pay runs and less admin on your side. You are buying expertise you do not have in-house.

Where it breaks: the provider holds the pay data but not your day-to-day events. Someone still has to tell it about moves, leavers and new hires on time, and someone still has to make sure your own tools hear about them. Outsourcing the machinery does not outsource the accountability. The Atlas is explicit that a small single-country company needs one accountable owner plus a good provider, not a team.

Option three: a people system with automated events

The third route is a dedicated human-resources information system. Large commercial suites dominate the corporate end, while open-source choices such as OrangeHRM and Frappe HR, both released under the GPL-3.0 licence, suit smaller budgets and teams comfortable running their own software.

Where it holds up: a single record with dates, workflows that fire when someone is hired or leaves, and reporting you can reconcile. Those are the traits the Atlas associates with the ops mindset.

Where it breaks: cost, set-up effort and the temptation to buy features nobody will configure. The Atlas also flags a trade-off from one documented large-company case: when terminations automatically start downstream tasks, the timely entry of that termination becomes critical. If the manager is late, everything is late. Automation amplifies a good process and a bad one alike.

Hidden costs people forget

Whichever route you take, three costs rarely appear on the first quote. The first is migration: moving old records into a new place takes a careful reconciliation, and the Atlas recommends running payroll in parallel for a cycle before you trust the new setup. The second is integration, since a feed between your people record and your account system built on flat files can fail silently. The third is training managers, who must enter changes promptly for any of this to work.

Side-by-side summary

  • Cost: spreadsheets are cheapest at first, providers charge per head or per service, systems cost most to set up.
  • History and audit: weak in spreadsheets, decent in providers for pay, strong in systems.
  • Handling leavers fast: manual in spreadsheets, partly manual with providers, automatable with systems.
  • Effort to maintain: rises with the discipline you must supply yourself in option one and falls as you buy structure.
  • Privacy control: poor by default in spreadsheets, good with role-based access in the other two.

A decision rule that fits real companies

Use the headcount as a rough guide, not a law. The Atlas offers size bands as editorial heuristics, not benchmarks, so treat them as starting points. Below about fifty people, an owner and a provider are usually enough, with the spreadsheet cleaned up and locked down. Between fifty and a few hundred, the first dedicated generalist and a people system start to pay for themselves. Above that, a service catalogue with written response promises becomes the major lever.

Switch earlier if any of these is true: you hire in several countries, a leaver has already kept access, or you cannot answer “who was employed on date X” in under five minutes.

Whatever you choose, do these four things

  1. Give each person a permanent identifier that is not their email address.
  2. Name an owner and a deadline for each lifecycle step, especially exits.
  3. Set up a daily check of the people list against the live accounts, and make any difference raise an alarm.
  4. Keep sensitive employee-relations matters out of any shared view.

Notice that none of these depends on which option you pick. Tools change; the habits carry over.

More to read here

Should you be weighing automation more broadly, our note on planning business automation for a smaller company“>planning business automation for a smaller company covers the budgeting side. And for anyone whose career runs through operations work, the primer on learning operations skills at a calm pace“>learning operations skills at a calm pace is a good companion.

The honest conclusion is dull but useful. Most companies should begin with a tidier spreadsheet and a named owner, add a provider when compliance gets complicated, and adopt a people system when the volume of events outgrows human attention. Pick the cheapest option that still lets you prove who was employed, when, and who owned each change.