Custom Software or Excel: When to Make the Leap

Excel isn't the enemy

Before we talk about software, a word in Excel's defense. It's flexible, everybody knows how to use it, it adapts to any process in minutes and it doesn't require a vendor. For one person tracking their quotes, for a one-off analysis or for a process that's still changing every month, it's the right tool.

The problem isn't the spreadsheet; it's using it as if it were a system: database, access control, workflow and reporting, all in one file that travels by email or WhatsApp. That's when files named "inventory_final_v3_for_real.xlsx" show up, along with meetings to decide which version is the right one.

In La Laguna (the Torreón metro area in northern Mexico) we see it all the time at distributors, repair shops and professional firms that grew from five to twenty people without changing tools: the file that worked for the owner and an assistant is now touched by ten people. The right question isn't "Excel or software?" but "What is sticking with this file costing my business?" If the answer is very little, don't change a thing. If it's a lot, keep reading.

Seven signs your business has outgrown Excel

1. Several people edit the same file and someone overwrites someone else's work.
2. Different versions are circulating and nobody knows which one is current.
3. A formula error has already cost you money: a miscalculated price, an overpaid commission, an invoice with the wrong amount.
4. You don't know who changed what. If a piece of data disappears, there's no way to trace it.
5. Your team enters the same thing two or three times: in the file, in the invoicing system and in a WhatsApp message.
6. You need to work from a phone, on a job site, on a route or on the sales floor, and the file is useless there.
7. The file is slow to open or is nearing the tool's limits. Excel caps out at just over a million rows per sheet and Google Sheets at 10 million cells per file; well before you get there, performance is already a problem.

If three or more sound familiar, the spreadsheet has stopped saving you time. You can measure it with the automation calculator: enter how many times a month the task repeats and how long it takes, and it tells you how many hours go into it.

The steps between Excel and your own software

It isn't a jump from zero to sixty. There are in-between options that solve many cases with less investment:

For inventory, for example, a well-designed Google Sheet holds up longer than people think. We explain it step by step in the guide to inventory in Google Sheets. And if what you need is customer follow-up, evaluate a CRM first: it's covered in our guide to CRMs for small businesses in Mexico.

When custom software really makes sense

Your own software is justified when at least one of these conditions applies:

• Your process is your edge. The way you quote, schedule production or deliver service is what sets you apart. Forcing it into generic software makes you work like everyone else.
• No product handles it well. You've already tried two or three SaaS tools and they all need Excel workarounds.
• You need to connect several pieces: CFDI invoicing (Mexico's mandatory e-invoice), WhatsApp, point of sale, bank, online store.
• The volume pays for it. If the process repeats hundreds of times a month, a few minutes saved per transaction add up to weeks a year.
• You want to own your data and not depend on a vendor raising prices or shutting down.

In our software development service, the starting point is always the current process, Excel included. Often the file already contains the business logic, and it just needs to move into a system with permissions, validations and backups.

An example: from a collections spreadsheet to a system

The case of the municipal water utility in Durango is a good example of when Excel can't take any more. Its registry of more than 18,000 service connections lived in one Excel file per district, with formulas only one person understood. Every two months, three data-entry clerks spent six days cross-referencing meter readings, applying rates and putting together bills, and the daily cash reconciliation took two hours.

The custom system centralized the registry, calculates rates with the utility's actual rules, issues bills in one day and produces the cash reconciliation in about 15 minutes. The important part: it wasn't designed from scratch. The rules were already in the Excel formulas; the work was to understand them, clean them up and turn them into a system that doesn't break when someone deletes a cell.

Something similar happened with Hanabi Ramen & Sushi in Gómez Palacio: attendance on signed paper sheets and closing out the register by adding up tickets were replaced by an in-house time clock and point of sale, with reports that still export to a spreadsheet for the accountant.

How to figure out whether it's worth it

You don't need a financial study, just four ballpark numbers:

1. Hours per month your team spends on the process (data entry, review, corrections, hunting for information).
2. Cost of errors: discounts given by mistake, lost inventory, customers who never came back.
3. Opportunities you pass up because operations don't scale: another location, more orders, a big client.
4. Total cost of the software over two or three years: development, hosting and maintenance.

If two years of hours and errors clearly exceed the total cost, the project is justified. The digital transformation assessment helps you spot which process would have the biggest impact if you digitize it first.

Common mistakes when leaving Excel

Changing tools without changing how you work is the fastest way to throw money away. These are the mistakes we see most:

• Copying the Excel file as is. If the new system replicates all 40 columns of the file, including the ones nobody uses, you inherit the mess. This is the moment to ask which data helps you decide something and which is only entered out of habit.
• Not involving the person who does the data entry. Whoever runs the file knows where the exceptions are: the customer who pays in two installments, the product sold both by the kilo and by the piece. If they don't take part, the system won't account for them.
• Buying a SaaS because of the demo. Demos show the ideal case. Ask to test it with your real data for two weeks before signing an annual contract.
• Forgetting about permissions. A system where anyone can delete anything carries the same risk as the shared Excel file.
• Not planning backups. At least Excel lived on several computers. A system on a single server with no copies is a bigger risk.

None of these mistakes is technical; all of them are avoided with a good discovery phase before anyone writes code.

How to migrate without stopping operations

• Start with one process, not the whole company. The one that hurts the most and repeats the most.
• Clean the data before migrating. Duplicates, customers listed under three different names, products with no SKU. This is often half the work.
• Run both systems in parallel for two to four weeks and compare results.
• Keep the Excel export. Your accountant and your team will still want to analyze data in a spreadsheet; the system should allow it.
• Train with real cases, not manuals. Ten minutes with the morning shift are worth more than a 40-page PDF.
• Decide who owns the system inside the company: who adds users, who requests changes.

You can see examples of systems we've already built in business software. And if the process you want to fix is mostly about moving data between tools, you may not need software at all, but automation: see our guide to n8n vs. Zapier vs. Make.