Spreadsheets get unfairly maligned in conversations about business software.
They are flexible, universally understood, require no procurement cycle, and let a team model a process in an afternoon that a software project would take three months to deliver. Most operational processes should start in a spreadsheet. Many should stay there permanently.
The trouble is that spreadsheets fail quietly. There is no error message when a process outgrows one. The work just gets slower, more fragile, and more dependent on specific people, until one day the operation is visibly straining and nobody can point at when that started.
Where Spreadsheets Genuinely Win
Worth stating plainly, because the answer is often "keep the spreadsheet":
The process is still being figured out and changes week to week.
The volume is low enough that manual handling is not a burden.
One or two people touch it, and they sit near each other.
Being wrong is cheap and obvious.
The data does not need to reach any other system.
Under those conditions a spreadsheet beats custom software on every axis that matters, including speed of change. Replacing it would be a downgrade.
The Signals You Have Crossed the Line
1. Reconciliation Has Become Somebody's Job
The clearest signal. When a non-trivial part of someone's week is spent making two sources agree, the spreadsheet has stopped being a tool and become a system of record that nothing else knows about.
Reconciliation work is invisible on a budget and expensive in practice. It is also demoralizing, which shows up in retention long before it shows up in reporting.
2. Throughput Is Capped by Headcount
If handling twice the volume means hiring twice the people, the process does not scale. That is fine at low volume and fatal at growth.
A clean test is whether you could serve your best month if it arrived tomorrow. If the honest answer is "we would need to hire and train first," the spreadsheet is the constraint.
Class Action Settlement House hit exactly this ceiling. For every settlement, staff pulled each claim's details and typed each field into the settlement portal, one record at a time. As Lee at CASH described the before state:
The time spent loading the cases is 80% to 85% of the time we are spending.
Careful, repetitive work that scaled only by adding people, and it crowded out the work that grows the business. After automating the filing, roughly 13,500 hours of manual processing were eliminated and about $450,000 saved. The process itself did not change. What changed was that volume stopped requiring bodies.
3. One Person Understands It
Every organization has a spreadsheet with formulas nobody else can follow, maintained by someone who has been meaning to document it.
This is a continuity risk that scales with how important the process is. When the person who understands it is on holiday, work either stops or gets done wrong. Neither is acceptable for a process the operation depends on.
4. Copy-Paste Between Systems Is Routine
If data is being moved by hand between the spreadsheet and another system, you are paying a person to be an integration.
Human integrations are slow, error-prone in ways that surface late, and they do not scale. They are also one of the cheapest things to automate, because the inputs and outputs are already well defined.
5. Nobody Can Answer "What Changed and Who Changed It"
Spreadsheets have weak audit trails. For most processes that does not matter. For anything touching money, compliance, member records, or customer commitments, it matters a great deal, and it usually matters for the first time during an incident.
If the answer to "who changed this and when" is "let me check the version history and guess," the process has outgrown the tool.
6. The Workarounds Have Workarounds
The tell is a second spreadsheet that exists to manage the first one. A tracker for the tracker. A tab that reconciles the other tabs.
This is the same signal as the shadow-IT tooling that builds up around an ill-fitting SaaS product, described in Build vs Buy: When a Custom App Beats a SaaS Subscription. The organization has quietly built software. It is just built out of manual steps.
What to Do Before Building Anything
Crossing the line does not automatically mean commissioning custom software. Cheaper options come first:
Fix the process, not the tool. Some spreadsheet pain is process pain wearing a spreadsheet costume. If the rules cannot be written down, no software will fix that.
Check whether something off the shelf fits. Most operational needs are common. A tool that fits is cheaper than a system that is yours.
Tighten the spreadsheet. Validation, protected ranges, and a single owner solve more problems than people expect.
Automate one step, not the whole process. The highest-volume, most mechanical step usually carries most of the pain.
Only when those are exhausted, or clearly insufficient, is custom software the right answer.
What Replacement Actually Looks Like
The mistake at this point is scope. Organizations that have suffered under a spreadsheet for two years tend to specify everything they ever wanted, all at once.
The better pattern is to replace the single workflow that hurts most, keep the spreadsheet for everything else, and let the two coexist for a while. That is unglamorous and it works, largely because it keeps the first project small enough to finish.
Which workflow to pick first is its own decision, covered in How to Decide What to Automate First.
A Simple Rule of Thumb
If your spreadsheet has become a system of record that other systems do not know about, you have outgrown it.
Dependency is the tell. When other work waits on it, when people reconcile against it, and when one person maintains it, it is infrastructure, and should be built that way.
Final Thoughts
The problem is the unacknowledged spreadsheet. None of these signals is a crisis on its own. Together they mean the operation has quietly outgrown its tooling.
The organizations that handle this well notice the crossing point early, then replace one workflow at a time.
If you are trying to work out whether a process has outgrown its spreadsheet, and what replacing it would actually involve, we are happy to take a look with you. Book a short consult.