Keep it or replace it
Spreadsheet vs Custom Application: The Specific Moment the Sheet Stops Working
A spreadsheet is a very good business application, and nobody had to sit through a demo to get it. Yours probably works. This page names the seven moments when it stops working, and what to try before anyone quotes you for an app.
November 12, 2026 · 9 min read
The short answer
Keep the spreadsheet while one person edits it and a mistake is cheap. Move to a custom application when a second person overwrites the first, you need it on a phone in the field, customers need to see part of it, a row needs a history of who changed what, or an error has cost you real money. Try repairing the sheet first. DataCram's web apps start at $575, depending on complexity.
On this page
In defense of the thing you are embarrassed about
Owners apologize for their spreadsheets the way people apologize for a messy kitchen before you have even sat down. There is no need. A sheet costs almost nothing, opens anywhere and bends to whatever you discovered about your business this morning. You can add a column in four seconds with no change request and no developer. Google Workspace Business Starter, listed at $7 per user per month as of November 2026, comes with one, along with mail and a calendar. Confirm the term at checkout, since the price depends on it.
Flexibility is the whole reason the sheet is good, and also the reason it eventually hurts. Nothing stops a thumb from typing over a formula. Nothing stops a column called Status from containing "done," "Done," "DONE," "paid I think" and a phone number. A spreadsheet has no opinion about your data. An application has nothing but opinions, and that is the entire difference. You are deciding when you need the opinions more than the freedom.
If nothing below has happened to you, stop reading and go do something billable. We mean it. A sheet that one person understands, and that has never cost you a job, should not be replaced because a salesperson looked at it with pity. Whether you need custom software at all is a separate test, and it usually comes out on the side of doing nothing.
Seven moments the sheet has stopped doing its job
These are the moments that count. Each one is an event you can remember, not a feeling. A vague sense that you should be more modern is not on the list. One of these is a nuisance. Three, recurring, are a pattern. Write down the date of the last time each one happened, since that note will do more to help a future quote than any requirements document.
- The second editor. Two people change it the same afternoon, and you learn on Thursday that one of them undid the other. Shared sheets have version history, but nobody looks until something is already gone.
- The cold thumb. You need to update it from a truck, on a phone, with gloves. A grid of 400 rows at that size is hard to read and easy to get wrong, so the update waits until tonight, and then it does not happen.
- The overwritten formula. A pasted value lands on a formula. Totals look fine for three weeks. Then a quote goes out low, and nobody can say since when.
- The customer who needs to see one row. You cannot share a row, so you read it to them over the phone, repeatedly. That call is where a customer portal begins, or a simpler private link.
- The question of who changed it. A price moved, or a job got marked paid, and you need to know who and when. Cell history is a poor tool for answering that when a customer is on the line.
- The nightly copy. You retype the same facts from the sheet into another system. That chore is rarely a sheet problem. It is a connection problem, and much cheaper to fix.
- The only person who understands it. The sheet has 11 tabs, and three of them hold the logic. You went to the hospital for two days and work stopped. That is a risk you can price, and it is worth pricing.
Try repairing the sheet before you replace it
Most of those moments have a fix that costs an evening. If the fix fails, our San Antonio guide shows how to read the spreadsheet as a specification for whatever replaces it. Do this first. If it works, you have spent nothing. If it does not, the failures you hit will tell you exactly what an application needs to do. Either way you will write a much better brief, and a brief is most of what you would be paying someone to work out.
- 1.Lock the formula cells. Protect those ranges so only you can edit them. Most overwritten formulas stop right there.
- 2.Force the choices. Use a dropdown on any column with a fixed set of answers. "Paid I think" stops being an option.
- 3.Enter through a form. A shared form writes rows for you and works on a phone. Helpers never touch the grid, so they cannot break it.
- 4.One row, one thing. One job per row, with a date column, and no merged cells. That habit alone also makes a later import much cheaper.
- 5.Name a keeper. One person changes the structure, everyone else adds rows, and a dated copy gets saved monthly.
- 6.Connect it with a rented tool. Zapier's Free plan has 100 tasks a month as of November 2026, enough to see whether moving rows automatically solves the nightly copy.

What actually changes when the grid becomes an app
The word "app" makes this sound bigger than it needs to be. An application is a spreadsheet that has been given rules and a door. The data still sits in rows, in a database. What you add is the rules about what those rows may contain, the screens that show one job at a time instead of 400, and the door that decides who sees what. Here is the same job in each.
| Shared spreadsheet | Small custom web app | |
|---|---|---|
| Adding a new field | Seconds, by anyone | A change request, and maybe a bill |
| Bad data | Accepted without comment | Refused at the door |
| On a phone | Tiny grid, pinch and hope | One job, big buttons, your words |
| Who changed what | Cell history, if you can find it | A log, if you asked for one |
| Showing a customer one record | Not really possible | Possible, with real access rules |
| Cost to start | Often nothing beyond what you already pay | From $575, depending on complexity |
| When it breaks | You fix it at the kitchen table | You call someone, and you need to know whom |
Read the first row again, because people forget it. Moving to an app costs you the very thing that makes your sheet good. Next spring, when a new service needs a new column, you will have to ask. That is why it pays to replace the sheet only for the parts that have stabilized. A sensible first version often leaves half of the old sheet alone, and an internal business app covers how to draw that line.
Moving the rows without moving the mess
Old data is where inexpensive projects become expensive ones. Somebody has to decide what "paid I think" meant in 2023. Our advice surprises some owners: do not bring it all. Import what is open and what you actually look up, and keep the rest as a read-only file, dated and stored somewhere safe. You almost certainly do not need seven years of closed jobs on your phone.
- 1
Freeze a copy
Save a dated, untouched copy before anyone starts. It is your safety net and your proof of what the numbers were.
- 2
List columns you actually use
Ask of each one: when did this last answer a question? Half of the columns in a long-lived sheet tend to be fossils.
- 3
Normalize the choices
Filter each column, count the spellings and collapse them. An hour of this by the person who knows the business saves a developer guessing.
- 4
Import open work only
Active customers, open jobs, unpaid invoices. Fresh records are entered through the new screens, which will refuse bad data from here on.
- 5
Run both for two weeks
Enter work in the app and check it against the sheet. When the two agree and nobody opens the sheet, set it to view only. Do not delete it.
The one export you must keep is the way back. Insist that the app can give you all of your records in a plain file, on demand, without asking anyone. A sheet is never held hostage, and your software should not be either. Our page on choosing a technology partner has more on who holds what, and the same logic runs through owning your code and accounts.
One more option: your spreadsheet might be a business
Once in a while a sheet has been refined by someone who knows a trade cold, and other people in that trade would pay for it. If you have ever had a competitor ask for a copy, read from spreadsheet to SaaS. Keep in mind it is a different business, with customers, support and a monthly bill. It does not make the case for replacing the sheet you use yourself.
An application is a spreadsheet that has been given rules and a door.
If you have repaired the sheet and the moments keep coming, the next step is small. Our custom software overview shows what a first tool looks like, and the web app service and workflow automation cover the build. If what you wanted from the sheet was really a Monday summary, a one-page dashboard is the smaller purchase. Published prices are here, the savings calculator gives you a first figure on phone and follow-up time, and the section for very small businesses has everything else. Send us the sheet with names removed, or call (210) 346-0848, and we will tell you if it should stay.
Questions owner-operators ask
Row count is rarely what breaks first. A sheet with thousands of tidy rows and one editor can run happily for years, while one with 200 rows and four editors can be a disaster within a month. Watch for overwritten formulas, conflicting edits and people avoiding the file. Those signs matter far more than its size.
It can be reasonably safe, if you control who has access, avoid public links and remove people when they leave. But sharing tends to be all or nothing, and copies wander. Do not keep card numbers in a sheet, ever. If you handle medical, legal or financial records, ask your own attorney what your obligations are, since we are not lawyers.
For a while, yes, and a two-week overlap is wise. Long term, pick one home for each kind of record, or you will have two answers to every question. Many small companies keep a sheet for planning and experiments, and let the app hold live jobs and customers. That split works as long as nobody copies live data back by hand.
DataCram's web apps start at $575, depending on complexity. Price depends less on the sheet's size and more on who signs in, whether the tool changes data and how much old data needs cleaning. Importing only open work keeps the cost down. Running costs such as hosting apply too, so ask for those line by line.
It will stop many of them, because it can refuse bad values, require fields and offer fixed choices. It will not stop someone typing the wrong address correctly. Expect fewer mistakes of the careless kind and the same number of honest ones. Reliable data also depends on making the correct way the fastest way, so test the screens on a real phone.
Keep reading
- Small, and yoursCustom Software for Microbusinesses: Small Tools That Retire the SpreadsheetDashboards, portals and integrations from $575, for owners who run the business from a phone and a spreadsheet, and the rented tools to try first.
- The office nobody seesAn Internal Business Management App: The Owner's Back Office, Minus the Owner's MemoryA small internal business app: three screens, who can touch what, how crews get it without an app store, and what to keep renting instead of building.
- Your internal tool, for saleFrom Spreadsheet to SaaS: Who Else Would Pay for the Tool You Built at Your Kitchen TableTurning a spreadsheet into a subscription business: who else would pay, five tests before a login, how tabs become screens, and what the jump changes.
- One owner, one phoneWeb Design for One-Person Businesses: What You Need and What to RefuseWebsite design for one-person businesses: the short page list, what to refuse to pay for, and fixed prices from $1,500, or when a builder plan is better.
- The build sideSmall Business Website Development: Who Holds the Keys When the Builder LeavesWebsite development for independent contractors: who owns the code, domain and accounts, what custom means at $1,500, and what you must maintain yourself.
- Probably not. Sometimes yes.iOS App Development for Small Businesses: When an iPhone App Earns Its KeepWhen an iPhone app makes sense for a very small company, what from $3,750 buys, Apple's fees and rules as of 2026, and what to try before you build.
- Everything in MicrobusinessSee the section
From the Insights desk
- Your Website Is Telling You It's Broken. Here Is the Translation.500, 502, mixed content, white screen, robots.txt blocking Google. Your site talks in a dialect built for engineers. Here is what each message actually means about your business — and what it costs while you look it up.
- 7 Signs Your Website Is Quietly Losing You CustomersA website rarely fails loudly — it just quietly lets customers slip away, and you never see the ones who left. Seven signs yours is leaking, from slow mobile loads to stale listings to no conversion tracking, and the fix for each.
Send us the sheet you apologize for
Tell us which of the seven moments keeps happening. A person replies with the fix we would try in the sheet itself, or what a small app would take and what it would cost.
Which of these sounds like you? (pick any)
Rather talk now? Call (210) 346-0848.