How to Prepare Your Business Data Before Building a New System
Moving to a new system? Here's how to clean, organise, and prepare your business data so your new software actually works from day one.
If you've ever moved house, you know how it goes. You open a drawer, pull out a tangle of cables, three remote controls for TVs you no longer own, and a battery you're fairly sure is dead. It all gets shoved into a box, moved to the new place, and eventually stuffed into a different drawer — still useless, just relocated.
Getting a new business system works the same way. If you don't sort your data before you move it, you're just carrying your mess somewhere more expensive. This is the problem nobody warns UK small business owners about when they decide to "get a proper system in place." The technology is usually the easy bit. The data is where things go sideways.
Here's how to prepare your business data before building a new system — so you actually get the outcome you paid for.
What "business data" actually means
When we talk about data, founders often think of rows in a spreadsheet. But your business data is anything you need to run the operation: customer records, job histories, product lists, staff rotas, supplier contacts, pricing, notes about ongoing projects. It lives in spreadsheets, yes — but also in email threads, WhatsApp messages, Post-it notes, someone's memory, and the third tab of a Google Sheet nobody has opened since 2022.
Before you build anything new, you need to know what you have and where it lives.
Step 1: Audit what you actually hold
Don't assume you know. Spend an hour asking the people who do the day-to-day work where they keep information. You'll almost certainly find that:
- Different team members track the same thing in different places
- Some data only exists in one person's head (or inbox)
- Historical records are incomplete, duplicated, or inconsistent
Write a simple list: what data do we have, where does it live, and how current is it? That list is your starting point.
Step 2: Decide what you're migrating — and what you're leaving behind
Not everything needs to come with you. A common mistake is trying to migrate every historical record from the past ten years when you only actually need the last 18 months to run the business day-to-day.
Ask yourself: if we never had access to this data again, would it affect how we operate tomorrow? If the answer is no, it probably doesn't need to be in the new system. Archived records can stay archived. You're building for the present and future, not preserving the past.
Being ruthless here saves you enormous amounts of cleaning time — and a bloated new system that's just as messy as the old one.
Step 3: Clean before you move
This is the part most people skip, and the part that causes the most pain later. Importing dirty data into a clean system means your new system is immediately dirty.
Some practical cleaning tasks:
- Merge duplicates. Do you have "J Smith", "John Smith", and "john.smith@gmail.com" as three separate customer records? Pick one, consolidate.
- Fill obvious gaps. If you're migrating customer records and half of them are missing phone numbers, is it worth a quick chase before migration?
- Standardise formats. Phone numbers stored as 07700900000, 07700 900 000, and +447700900000 are three different-looking things that mean the same thing. Pick one format.
- Remove test entries. Almost every spreadsheet has a "TEST" row or a "DO NOT DELETE" entry that no one can explain. Delete them.
You don't need perfection. You need consistency.
Step 4: Agree on naming conventions before you go live
One of the most overlooked preparation steps is agreeing on how things will be named going forward. If your new system has a "status" field for jobs, what are the allowed values? Who decides when something moves from "In Progress" to "Awaiting Sign-Off" to "Complete"?
Without this, you'll have a clean system on day one and a chaotic one by week three, because different team members will invent their own labels.
Write a simple one-page document — even a bulleted list — that defines your categories, statuses, and conventions. Laminate it if you have to.
Step 5: Nominate a data owner
Every system eventually drifts unless someone is responsible for keeping it tidy. Before you go live, decide who that person is. This doesn't need to be a full-time role — it might just mean someone checks for duplicates once a month and makes sure new starters are trained on the conventions.
Systems don't maintain themselves. A data owner turns a one-off cleanup into an ongoing habit.
The real cost of skipping this
We've seen it more times than we'd like to count: a business invests in a new system, gets it built beautifully, and then spends the first three months frustrated that the data "just doesn't feel right." Reports are wrong because the underlying records are inconsistent. Searches return duplicates. Staff lose trust in the system and go back to their spreadsheets.
The technology wasn't the problem. The preparation was.
Getting your data ready before you build isn't glamorous. But it's the difference between a new system that transforms how you work and one that just gives your existing mess a glossier interface.
If you're thinking about building a new system and not sure where to start with your data, we can help you map it out before a line of code is written. Book a free discovery call and we'll work through it together.