12 Signs Your Business Has Outgrown Excel

Excel is not the enemy. It is the most flexible business tool most Nigerian companies own, and plenty of well-run firms use it daily by choice. The problem is that spreadsheets fail gradually and invisibly: no error message, no downtime, just slowly worsening numbers, duplicated effort and one person who has become irreplaceable because only they understand the workbook.

This article helps you tell the difference between "Excel is fine here" and "Excel is now costing us money", explains what technically breaks and why, and sets out the sensible next steps in order of cost.

What Excel is genuinely good at

Before the signs, it is worth being fair to the tool, because businesses often replace it for the wrong reasons and then rebuild spreadsheets inside their new system.

Excel is excellent for:

• Analysis and modelling: pricing scenarios, cash-flow projections, budget variance, quick what-if questions.
• One-off work: a supplier comparison, a stock count, an event budget.
• Prototyping a process before committing to software, which is a legitimate and cheap way to learn what you need.
• Reporting on top of other systems, where data is exported and analysed.
• Small volumes with one owner, where the person who builds it is the person who uses it.

Excel is poor as a system of record for data that many people create, update and depend on. That distinction — analysis tool versus system of record — is the whole argument.

Twelve signs you have outgrown Excel

Count how many apply to you. Each one has a specific consequence, not just an inconvenience.

1. Two or more people need the file at the same time. Consequence: someone works on an outdated copy, and a day of entries vanishes when their version is saved over the other.
2. Files have names like "stock_final_v3_new.xlsx". Consequence: nobody knows which version is authoritative, and decisions get made from the wrong one.
3. You cannot tell who changed a number, or when. Consequence: disputes cannot be resolved, and errors cannot be traced to their source.
4. One person is the only one who understands the workbook. Consequence: a key-person risk. When they travel or resign, part of the business stops.
5. The same data is typed into more than one file. Consequence: customer details in a sales sheet, a delivery sheet and an invoice template, all slightly different, all wrong in different ways.
6. Reports are rebuilt by hand every week or month. Consequence: hours of skilled time spent on copying and formatting rather than on decisions.
7. The file has become slow, or has crashed with unsaved work. Consequence: lost entries, and staff who avoid the file until the end of the day, which is when errors happen.
8. Formulas break and nobody notices for weeks. Consequence: a dragged formula that missed the last thirty rows understates stock or overstates revenue, silently.
9. Everyone can see everything. Consequence: salaries, margins, supplier prices and customer lists are visible to anyone with the file, and portable on a flash drive.
10. You cannot answer a simple question quickly. "How much does this customer owe, and since when?" should take seconds. If it requires opening three files and a phone call, the spreadsheet has stopped serving you.
11. Staff maintain their own private copies. Consequence: the real data is now fragmented across laptops and WhatsApp, and the official file is a fiction.
12. The business would be seriously damaged if the file were lost. Consequence: you have made a laptop, a flash drive or one cloud account the foundation of the company.

Interpretation. One or two signs: manageable, fix the specific problem. Three to five: you are paying a real weekly cost; move at least this process to a proper system. Six or more: the spreadsheet is now a business risk, not just an inefficiency.

What actually breaks, and why

Understanding the mechanism helps you avoid rebuilding the same problems elsewhere.

No concurrency control. Desktop spreadsheets assume one user. Cloud versions allow simultaneous editing but still have no concept of a transaction, so two people can create conflicting entries that both look valid.

No data validation by default. A customer can be "Chinedu Okeke", "C. Okeke" and "Okeke Chinedu" in the same column. Databases enforce one record per customer; spreadsheets happily hold three, and your customer count, balances and reports are then wrong.

No audit trail. Spreadsheets do not record who changed what. For stock, payments, payroll or anything involving money, that absence is a control weakness, not merely an inconvenience.

Formulas are fragile. Inserting a row, sorting a range or dragging a formula one cell short produces results that look plausible. Errors of this kind are widespread and rarely detected quickly, because nothing announces them.

No role-based access. Access is all or nothing at file level. You cannot let a sales officer add orders while preventing them from reading margins.

No relationships between data. A proper system links a customer to their orders, invoices and payments. In a spreadsheet those live in separate sheets connected by typed references, which drift apart over time.

Limited automation. Reminders, approvals, notifications and scheduled reports either do not exist or depend on macros only one person can maintain.

Scale limits. Long before the row limit, files slow down, formulas recalculate sluggishly and crashes become frequent — usually on the machine of the person doing the most work.

A quick self-assessment

Score each question 0 (never), 1 (sometimes) or 2 (often).

• 0–4: Excel is serving you. Tighten backups and access.
• 5–9: fix structure first — move to cloud spreadsheets with permissions, and standardise the important files.
• 10–16: move the affected processes into a proper system. The weekly cost already exceeds the price of the alternative.

What changes for Nigerian businesses

Devices and data loss

Laptops are shared, borrowed, stolen and repaired. A workbook that exists only on one machine is exposed in a way that is easy to underestimate. Cloud storage with accounts controlled by the business — not by an individual staff member's personal email — is the minimum sensible arrangement.

Power interruptions

Unsaved work disappears when power drops and a laptop battery is weak. This pushes staff into habits like entering data once at the end of the day from paper notes, which introduces exactly the errors the spreadsheet was meant to prevent.

Sharing through WhatsApp

Sending the stock file as a WhatsApp attachment is normal practice in many Nigerian businesses, and it is how version control dies. Once a file is downloaded and edited on three phones, no one can say which copy is current.

Staff turnover and handover

When the person who built the workbook leaves, the formulas, the conventions and the undocumented rules leave too. Systems with defined fields and permissions survive handover far better than a personal spreadsheet ever will.

Data protection obligations

Customer names, phone numbers, addresses and identity details are personal data under the Nigeria Data Protection Act 2023. A file circulating on personal devices and in chat groups is difficult to reconcile with any reasonable interpretation of your obligations. Verify what applies to your organisation with the Nigeria Data Protection Commission's guidance or a qualified adviser.

Connectivity

Cloud tools need data. For businesses in areas with weak coverage, choose tools that work acceptably on mobile data and consider whether offline capability is genuinely required before paying for it.

Where to move next, in order of cost

Moving off Excel does not mean commissioning software. Work up this ladder and stop at the cheapest step that solves your problem.

Most Nigerian SMEs should look hardest at step 3. Accounting, inventory and customer records are standard problems with mature products, and our guide to CRM versus spreadsheets covers the customer side of that move specifically. Step 5 is justified when your rules genuinely cannot be expressed in a product — the test set out in our article on when to build custom software.

Example (hypothetical): a Lagos building materials retailer

Example (hypothetical): a retailer in Lagos sells cement, tiles and plumbing fittings from a shop and a yard, with eleven staff. Stock is tracked in one Excel file, credit customers in another, and daily sales are entered each evening by the shop supervisor from handwritten receipts.

Symptoms: the stock file says 240 bags of cement when the yard holds 195, because deliveries out of the yard are recorded only when the supervisor remembers. Two customers dispute their balances. The owner cannot see daily sales without ringing the shop. When the supervisor travelled for two weeks, no entries were made at all.

Applying the twelve signs, the business scores eight, and the self-assessment lands at 12. The recommendation is step 3, not step 5: a ready-made inventory and sales product that runs on a phone or tablet in the shop, records sales at the point of sale, tracks credit customers with balances, and separates yard releases from shop sales.

Custom software would only be justified if the retailer later adds branches with stock transfers, customer-specific pricing tiers and delivery scheduling — at which point the earlier product has already cleaned the data and proved the process. Figures and details here are illustrative.

How to migrate off Excel without losing data

1. Choose one process to move first. Stock, or receivables, or customer records. Moving everything at once is the usual reason migrations stall.
2. Clean the data in Excel before moving it. Deduplicate customers, standardise product names, fix dates, remove blank rows. This is easier in the spreadsheet than after import.
3. Agree the master record. Decide which file is authoritative and freeze the others. Every stray copy is a future dispute.
4. Map fields to the new system before importing, including what happens to columns the new system has no home for.
5. Import a small sample first, check it thoroughly, then import the rest.
6. Enter opening balances so the new system is correct from day one rather than only for new transactions.
7. Run in parallel for two to four weeks for anything involving money, and reconcile the difference every week.
8. Set permissions so that people see what they need and no more.
9. Archive the old files read-only and stop new entries in them on a fixed date.
10. Train on the exceptions, not just the happy path. Returns, part payments, damaged stock and cancellations are where staff revert to spreadsheets.

What to keep in Excel even after you move

A system does not replace analysis. Keep spreadsheets for:

• Budgeting, forecasting and scenario modelling.
• Ad hoc analysis of exported data, where flexibility is the point.
• One-off exercises such as a supplier tender comparison.
• Board or management packs assembled from system reports.
• Prototyping a new process before asking a developer to build anything.

The rule that keeps you out of trouble: spreadsheets may consume data from your systems, but they should not be the place where operational data is created.

Mistakes to avoid

• Replacing Excel with a bigger Excel. Adding macros and complex formulas to defer the decision increases fragility and deepens the dependency on one person.
• Buying software before cleaning data. Importing duplicates and inconsistent names moves the mess rather than removing it.
• Migrating everything at once. One process, proven, then the next.
• Letting the old file stay open for entries. Two systems mean neither is trusted.
• Ignoring opening balances. A receivables system that starts at zero hides the debt that actually matters.
• Choosing a product no one tested with real data. Demos use clean data; your data is not clean.
• Jumping to custom software. Build only if a product genuinely cannot express your rules, not because the spreadsheet is painful.
• Forgetting the phone. If shop or field staff cannot enter data on an Android phone, they will write on paper and someone will retype it.

Conclusion

Excel stops being appropriate the moment a spreadsheet becomes the system of record for data that several people create and depend on. The practical signals are concurrent editing, version confusion, untraceable changes, key-person dependency and reports rebuilt by hand. Score yourself against the twelve signs, then take the cheapest step that fixes the specific failure: restructure and control the file, move to cloud sheets, or adopt a ready-made product. Reserve custom software for the case where your rules genuinely cannot be expressed in a product — and whichever route you take, clean the data first and run in parallel before switching.

If you are weighing a ready-made product against a custom system to replace a critical spreadsheet, Linestech can review the process and the data and advise on the least expensive route that will actually hold.

Frequently asked questions

Is Google Sheets enough to fix our spreadsheet problems?

It solves version chaos, simultaneous editing and file-level access control, which addresses several of the twelve signs cheaply. It does not add proper validation, audit trails, relationships between records or workflow. If your problems are collaboration and versioning, cloud sheets may be all you need. If your problems are data quality, control and reporting, they are a temporary improvement.

How many rows can Excel handle before it becomes a problem?

Practical trouble arrives long before any technical limit. Files with tens of thousands of rows, heavy formulas and multiple linked sheets become slow and crash-prone on ordinary office laptops. Treat performance complaints as a signal about structure and usage, not as a question about row counts.

We are a small business. Is software not overkill?

Size is less important than the number of people touching the data and how costly an error is. A three-person business handling credit customers and stock may need a proper system sooner than a ten-person consultancy with five invoices a month. Judge by the signs, not the headcount.

Can we automate our spreadsheets instead of replacing them?

Sometimes, and it can be good value. Scripts and connectors can pull data automatically, produce a report on a schedule and send it out, which removes the manual rebuild. What automation cannot fix is the absence of validation, permissions and audit trails. If the problem is repetitive work, automate. If the problem is trust in the data, replace.

What is the safest way to keep spreadsheet backups?

Store the master file in a business-controlled cloud account with version history, not on a personal account or a single laptop. Keep a second copy elsewhere, restrict who can delete files, and test occasionally that an old version can actually be restored. A backup nobody has ever restored is an assumption rather than a backup.

Our accountant prefers Excel. Does that change the decision?

No, but it shapes the choice. Accountants work in Excel for analysis, which is appropriate. What matters is that your operational records — sales, stock, invoices, payments — live in a system that exports cleanly to Excel. Choose products with good export, and your accountant keeps the flexibility without the business keeping the risk.

How long does it take to move off spreadsheets?

For one process with clean-ish data, expect two to six weeks including cleaning, import, parallel running and training. Messy data, multiple branches or several processes at once extend this considerably. The data cleaning almost always takes longer than the software setup, which is why it should start before you choose a product.

Sources and further reading

Figures, platform rules and regulations change. These are the primary references behind this article and the places to check before you act on it.

• Nigeria Data Protection Commission (NDPC)
• Core Web Vitals — web.dev
• PostgreSQL documentation