Spreadsheet to database: when a UK small business should switch


Spreadsheet to database: when a UK small business should switch
The spreadsheet that runs the business
Most small businesses have one: a spreadsheet that started as a quick list of customers, jobs or stock and slowly became the place where the business actually lives. Several people update it, a few formulas hold it together, and everyone is a little nervous about breaking it.
There is nothing wrong with that at the start. Spreadsheets are cheap, flexible and everybody knows how to use them. The question is when the same flexibility starts costing you more than it saves, and what to move to when it does. This guide gives you the warning signs, the honest case for staying put, the options from smallest step to biggest, and what a move typically involves and costs for a UK small business.
If your data is messy but you are not sure it needs a new home yet, start with my 30-day data starter kit. It shows how far a tidy spreadsheet can take you before you change anything.
Spreadsheet or database: what is the real difference?
A spreadsheet is a grid where data, calculations and presentation all sit together. That is why it is so quick to start with, and also why it gets fragile. Anyone can type anything into any cell, sorting one column can scramble another, and the rules about what is allowed only exist in people's heads.
A database separates those jobs. Data lives in tables with fixed rules: a customer must have an email address, an order must belong to a customer that exists, a date must be a date. People enter and edit data through forms or an app, several of them at once, and reports read from the same single source. You lose a little of the freedom and gain a lot of trust in the numbers.
Seven signs you have outgrown your spreadsheet
One of these on its own is not a reason to change. Three or more usually is.
- More than one person edits it at the same time. Files get locked, copies called "final v3" appear, and someone spends Monday morning working out which version is right.
- The same data is typed in several places. A customer's details live in the quotes sheet, the jobs sheet and the invoicing sheet, and they never quite match.
- Every report means copying and pasting. If producing the monthly numbers is a half-day job, the data is in there but you cannot reach it. My post on the 5 metrics that actually matter covers which numbers deserve that effort.
- The rules live in one person's head. Only Sarah knows that column F must be filled in before the status changes. Nothing in the file enforces it, and that knowledge leaves when she does.
- The file is slow, huge or close to its limits. Excel has a hard limit of 1,048,576 rows per worksheet (Microsoft's Excel specifications) and Google Sheets allows up to 20 million cells per spreadsheet (Google Drive file limits). Most small businesses never get close, but long before that the file becomes slow and fragile.
- It holds personal data and anyone can see or change all of it. Under UK GDPR you must protect personal data with "appropriate technical and organisational measures" and be able to restore it after an incident, as the ICO's guide to data security explains. A shared spreadsheet with no permissions and no history makes that hard to show.
- You keep history by saving copies. Monthly files with dates in the name, and nobody can say what changed between two of them or who changed it.
Why this matters more than it looks
Spreadsheet errors are rarely dramatic until they are. In October 2020, Public Health England failed to report 15,841 positive Covid cases because data was being passed through an old Excel file format that ran out of rows, as the BBC reported at the time. That was a national organisation, but the mechanism is exactly the one small businesses face: a file quietly doing a job it was never designed for.
Research backs this up. Raymond Panko's review of the evidence, What We Don't Know About Spreadsheet Errors Today, concludes that errors are rare per cell but that a large spreadsheet is very likely to contain at least one wrong bottom-line figure, that errors are very hard to spot, and that people are overconfident about how accurate their spreadsheets are. The EuSpRIG horror stories collection has plenty of real examples if you want to lose some sleep.
When you should keep the spreadsheet
I would rather tell you when you do not need a database. Stay with a spreadsheet when:
- One person owns the file and is the only one who edits it.
- The data is small and stable, such as a supplier list or a simple budget.
- Nobody outside the business, such as a customer, feels it when something goes wrong.
- It is a one-off analysis rather than a process you run every week.
In those cases, tidy the file instead: one tab per purpose, data validation on the key columns, protected ranges for formulas, and a proper backup. The NCSC's guidance on backing up your data is a sensible checklist for any small business, whatever tool you use.
Your options, from smallest step to biggest
Moving off a spreadsheet does not have to mean an expensive software project. There is a ladder, and most small businesses only need to climb one or two rungs.
1. A tidier, locked-down spreadsheet
Validation, drop-down lists, protected formulas and a single master file shared from one place. It costs almost nothing and fixes a surprising amount, but it does not solve several people editing the same records or the lack of history.
2. A no-code database tool
Tools such as Airtable, Microsoft Lists or similar products look like a spreadsheet but behave like a database underneath: proper field types, permissions, forms, change history and, in most cases, links between tables. They suit one or two processes run by a small team, and you can often set one up yourself. Watch the per-user subscription costs as the team grows, and check where the data is stored if it includes personal data.
3. A proper relational database with simple tools on top
A database such as PostgreSQL, hosted in the cloud, with data entry forms or a small internal app in front of it and a reporting tool such as Looker Studio or Power BI reading from it. This is the setup I most often recommend for SMEs. It handles several connected processes, scales comfortably for years and lets you join your data to other systems such as your accounts, CRM or online shop.
4. A bespoke application
A fully custom system built around your workflow. Occasionally this is the right answer, but usually only for the one process that is genuinely unique to you. I set out how to make that call in Buy the standard parts, build the part that is yours.
Option Best for Effort
--------------------------------------------------------------------------
Tidier spreadsheet One owner, small stable data Hours
No-code database tool One or two team processes Days
Relational database + tools Several connected processes Weeks
Bespoke application A truly unique core process Months
--------------------------------------------------------------------------
Rough guide only. The right rung depends on your data, not your size.What a move actually involves
The businesses that switch successfully nearly all follow the same pattern.
- Pick one process. Quotes, jobs or stock, not all three. Start with the one that causes the most pain or the most errors.
- Use the spreadsheet as the specification. Your existing file shows what you really track, which columns nobody uses any more and where the exceptions are. It is a better brief than any meeting.
- Clean before you copy. Remove duplicates, fix inconsistent names and decide which of two conflicting records is right. Moving messy data just gives you a faster mess.
- Run both side by side for a few weeks. Keep the old file read-only for reference and enter new work only in the new system. This is where the surprises come out.
- Set a cut-off date and stick to it. If the old spreadsheet stays editable, people will keep using it. You can still export to Excel for analysis, but the database becomes the single source of truth.
What it costs
The cost depends far more on how messy the data is and how many processes are involved than on the size of your business. If you bring in outside help, the median UK contract day rate for a data analyst was £475 in the six months to 16 November 2026, according to IT Jobs Watch. As a rough planning guide rather than a quote:
- Setting up a no-code tool for one process, with your data moved across and a simple report: two to five days of help, so roughly £950 to £2,375 at that rate.
- A relational database for one or two connected processes, with entry forms and a dashboard: often 10 to 20 days of work, so roughly £4,750 to £9,500.
- Ongoing support afterwards: usually a day or two a month, or nothing at all if the system is simple and well documented.
For more detail on how freelance pricing works, read How much does a freelance data analyst cost in the UK?, and if you expect ongoing work, my fractional data analyst vs full-time hire guide compares the options.
Compare it with what the spreadsheet already costs you
The cost of staying put is real but hidden. As an illustration, if two people each spend three hours a week reconciling copies, re-typing data and building reports by hand, that is 312 hours a year. At £20 an hour, that is about £6,240 a year, before counting the cost of a single wrong invoice or a missed order. It is the same reasoning I used in The Smarter Way to Cut Costs: find where time and money leak out quietly before you cut anything else.
What you get once the data has a proper home
The database itself is not the goal. What it unlocks is. Once your data lives in one trusted place, reports refresh themselves instead of being rebuilt each month, and you can start answering forward-looking questions. A reliable sales history, for example, is the starting point for the method in my guide to sales forecasting for small businesses. For the bigger picture of what analytics can do for a business your size, see my guide to data analytics for SMEs.
Frequently asked questions
Is Microsoft Access still a good option for a small business?
It can be, if your team works on Windows in one place and the system is small. For teams that work remotely or across devices, a cloud database or a no-code tool is usually easier to share, secure and back up.
Can we keep using Excel after moving to a database?
Yes. Many businesses keep Excel or Google Sheets for analysis and ad hoc reports, pulling data from the database rather than typing it in. The rule is simply that the database is where data is entered and corrected.
How long does it take to move from a spreadsheet to a database?
For a single process with reasonably clean data, a no-code tool can be running within a week or two. A relational database with forms and reporting for a few connected processes typically takes a few weeks, including the side-by-side period.
Will we lose data in the move?
Not if you plan it properly. Keep the original spreadsheet as a read-only archive, check record counts and totals after the import, and run both systems in parallel until the numbers match.
Do we need a developer to build a database?
Not always. No-code tools are designed for business users. You need technical help when several processes connect, when data has to flow in from other systems automatically, or when personal data needs careful access controls.
Not sure whether your spreadsheet needs replacing?
I build databases, data pipelines and simple internal tools for small and medium-sized businesses in Leeds, across Yorkshire and remotely across the UK. If you have a spreadsheet that the business depends on, I am happy to look at it and tell you honestly whether it needs a database, a no-code tool or just a tidy-up. Have a look at my services and get in touch.
Related Posts

Fractional data analyst vs hiring full-time: a UK cost guide
Fractional data analyst or full-time hire? A UK cost comparison for SMEs with 2026 salaries, day rates, National Insurance, pensions and a break-even point.

Freelance data analyst rates in the UK: what you will actually pay
Freelance data analyst rates in the UK in 2026 typically run from £400 to £550 a day. Covers project budgets, freelance versus a permanent hire, and IR35.

Sales forecasting for small businesses: a practical guide
A five-step sales forecasting method for UK small businesses, built in Excel or Google Sheets from data you already have. Baselines, seasonality, ranges and how to measure accuracy.