When has the spreadsheet stopped working?

A spreadsheet has stopped working when it needs one person to keep it right, copies disagree, or mistakes reach customers first. That is the point where a small web app built around the spreadsheet's one job pays for itself.

A spreadsheet is a draft of a system. At some point the business outgrows the draft.

When has the spreadsheet stopped working?

The spreadsheet stopped working when the business started working around it.

A spreadsheet has stopped working when it needs one person to keep it right, when two copies disagree, or when a mistake in it reaches a customer before anyone notices. That is usually the point where a small web app, built around the one job the spreadsheet does, pays for itself. CodeFish has built that kind of system for Australian operators from Adelaide since 2010.

The spreadsheet is rarely the problem on day one. It was the fastest way to run a new process, and it worked. It stops working as volume grows: more rows, more people editing, more decisions made from a version that was right last Tuesday.

Signs the spreadsheet has outgrown the job

Any one of these is a nuisance. Two or more, every week, and the spreadsheet is costing more than a system would.

One person holds it together

Only one person knows which tabs matter and which formulas not to touch. When they are on leave, the process slows or stops.

Copies that disagree

Someone emails a version, someone else saves a local one, and the Monday meeting starts by working out which number is right.

Rekeying into other systems

Rows typed again into Xero, MYOB or the ERP. Every rekey is time spent and a fresh chance for an error.

Errors found by customers

A wrong price, a missed delivery, a double booking. The spreadsheet has no rules to stop it, so the customer finds the mistake first.

Nobody can see who changed what

No audit trail, no approval step, no way to tell an honest typo from a missing process. That matters the day an auditor or a client asks.

Growth means more admin

Each new customer, site or job adds the same block of manual work. Revenue grows, and so do the hours needed to service it.

Fix it, connect it, or replace it?

Fix it if the spreadsheet is still the right tool and the pain is a few broken formulas or a missing lock on the shared file. That is a day's work, and we will say so. Connect it if the data already lives in good systems and the spreadsheet exists only to move numbers between them; that is an integration job, covered on our systems integration page, and where the work is repetitive copying between screens, when RPA pays off sets out the test.

Replace it when the spreadsheet is doing the work of a system: holding the records, enforcing the rules and handing work from one person to the next. Whether to buy a product or build is its own question, answered in build or buy a web app. Either way you end up with one copy of the data, rules the software enforces, and hours handed back to the people who were keeping the spreadsheet alive.

How a spreadsheet becomes a system

Pull the spreadsheet apart (1–2 weeks)

Every formula, lookup and workaround traced and written down with the people who use it. Half the rules turn out to live in someone's head, and the scope captures them before any code is written.

A CodeFish developer reviewing a project's commit history on screen

Build the first job, not the whole thing

The workflow that costs the most hours goes live first, with real users, while the spreadsheet keeps running alongside it. You see the system working on your own data within weeks.

A designer reviewing a website layout in Figma on a large monitor

Move the data, then retire the file

Old rows are cleaned and migrated, and the spreadsheet is switched to read-only. Staff stop going back to it because the system is faster for the daily task.

Four CodeFish team members working through a problem around a table of laptops, CodeFish Studio sign on the wall

Spreadsheet questions we get asked

Sometimes, yes. If one person uses it, the data stays in one place and errors are rare, a tidy-up with locked cells and clear tabs is the cheaper fix. It stops being enough once several people edit it, other systems depend on it, or the business needs to grow without adding admin.

That suits a process that is simple, internal and close to a list. It struggles once the rules get specific to your business, data has to flow into Xero, MYOB or an ERP, or customers need their own view. Per-user licences also grow with the team, which a custom build does not have.

Because the new system made their daily task slower. Adoption is designed in by building around the job people do most, running both side by side for a short period, and fixing what slows them down in the first weeks after launch.

It starts with a $9,000 + GST Rapid Insight Workshop, which maps the process and the hours, and delivers a roadmap about three weeks later. The build is then scoped there and priced by the stream that fits its size. The cost section below shows what the spreadsheet is costing you now.

It moves across. Migration is planned in the scope because spreadsheet data is always messier than anyone expects: merged cells, free-text dates, duplicate customers. Cleaning it is part of the job, and the old file stays as a read-only record.

Yes. The code and the IP are yours, documented and handed over, with no per-user licence. You can extend it with us or take it to another team.

What's this costing you per year?

Two team members in a meeting room, one pointing at a whiteboard

Count the hours, then multiply by the true cost of the people.

Start with the hours. Add up the time each week spent updating, checking, reconciling and rekeying the spreadsheet, across everyone who touches it, then multiply by 48 working weeks. Price those hours at true employer cost, not salary: once super at 12%, payroll tax, workers compensation, leave loading and management overhead are added, a person costs 1.38 to 1.48 times their base salary. Then add the errors, such as a wrong invoice, a missed delivery or a write-off, which are harder to count and usually the larger number.

Compare that with the alternatives. Hiring a senior developer in Adelaide to build and keep a system running costs about $177,000 to $190,000 a year all in, plus a recruitment fee of 15 to 22% of salary and months to fill the role. An offshore assistant in the Philippines keeping the sheet up to date runs about $30,000 to $45,000 a year (VA Masters 2026), but the spreadsheet stays the same. An off-the-shelf platform at 20 users can run $12,000 to $20,000 a year in licences before support. A Rapid Insight Workshop at $9,000 + GST puts a real number on your spreadsheet in about three weeks, so the decision is made on your figures rather than ours.

Ready to build something great?

Local partners, global standards – full service design and development.