Data fundamentals

Why it matters

Owners who can see what is happening in their business make faster, calmer decisions. They know which services make money, which customers pay late, which months are quiet and which marketing actually brought people through the door. Without that view, decisions default to the loudest recent memory, which is usually a complaint or a lucky win. Getting the fundamentals right also prepares you for everything else in this section: dashboards, KPIs and eventually forecasting are all built on records you can trust.

What counts as data in a business

Data is simply a recorded fact. The date and amount on an invoice, the postcode on a delivery docket, the time a staff member clocked on, the number of people who opened your newsletter. None of it is useful on its own; it becomes useful when it is counted, compared and connected to other facts.

Some data is structured: it sits in neat rows and columns with a consistent meaning, such as sales in your accounting software or appointments in a booking system. Some is unstructured: emails, call notes, photos, reviews and PDFs. Structured data is far easier to analyse, which is why one of the most valuable habits in a small business is turning unstructured events into structured records, for example logging every complaint in a form rather than leaving it in an inbox.

From data to information to action

The lead figure shows the journey. "212 jobs completed this month" is data. "212 jobs, up from 180 last month and 195 in the same month last year" is information, because it has context. "The growth came almost entirely from one suburb and one referral partner" is insight, because it explains why. "Call the partner, thank them and ask what else they need" is action, and it is the only step that changes anything.

Most businesses get stuck at the first layer. They can produce a sales total but cannot easily compare it, break it down or explain it. The fix is rarely more data; it is better questions. Start with the decision you need to make, work out what information would make it obvious, and only then look for the data.

Where your data already lives

A typical small business keeps data in its accounting software, a point-of-sale or invoicing tool, a booking or job management system, a customer list of some kind, website and social media analytics, the bank, and a collection of spreadsheets and inboxes. Each one answers a few questions well and the rest not at all.

The real problem is not the number of systems but that they do not agree on who a customer is. The accounting system has "J Smith Pty Ltd", the job system has "Smith, John" and the email list has his personal address. Until a customer, product or job has one consistent identifier across systems, you cannot ask simple questions like "which customers are most profitable" without hours of manual matching.

• Accounting software: revenue, costs, debtors, GST and BAS figures.
• Point of sale, invoicing or e-commerce platform: individual sales and products.
• Job, booking or project system: what work was done, by whom, how long it took.
• Customer list or CRM: contacts, history, source of enquiry.
• Website and marketing analytics: traffic, enquiries, campaign results.
• Spreadsheets and inboxes: everything that did not fit anywhere else.

What makes data trustworthy

Data is only worth analysing if people believe it. Four qualities matter most: accuracy (the record matches reality), completeness (the fields you need are filled in), consistency (the same thing is recorded the same way every time) and timeliness (it is entered soon enough to act on). A sales figure that is accurate but three weeks late, or a customer list where half the postcodes are blank, will quietly undermine every report built on it.

The cheapest way to earn trust is to agree on a single source of truth for each kind of data. Revenue comes from the accounting system, not the sales team's spreadsheet. Stock counts come from the inventory system, not memory. When two sources disagree, you fix the source rather than arguing about the report.

How to take stock of your business data

1. List your systems Write down every place a business record is kept, including spreadsheets and shared inboxes.
2. Map what each one holds For each system, note the kinds of data (transactional, customer, operational, financial, behavioural) and who owns it.
3. Write down your top five questions The decisions you make most often or that cost the most when wrong.
4. Check which questions you can answer today Try to answer each one in under ten minutes. Note where you get stuck.
5. Pick one identifier per thing Decide how a customer, product and job will be named across every system, and start using it.
6. Fix one capture habit a month Make a required field, add a dropdown, retire a spreadsheet. Small, permanent improvements beat a big clean-up.

Real-world example

A Ballarat plumbing business finds out where its profit comes from

Dale runs a plumbing business with six vans. His job management system records every job, and his accounting software records every invoice and bill, but he has never joined the two. He believes bathroom renovations are his best work because the invoices are large. For the last financial year he exports 2,340 jobs worth $1.87 million and groups them by job type, matching labour hours from the job system against invoice value.

Emergency call-outs are 28 per cent of jobs but 41 per cent of gross profit, because they are priced for urgency and finished quickly. Renovations are 9 per cent of jobs and 22 per cent of revenue, but once labour overruns are counted they contribute only 6 per cent of gross profit. The information was always there; it simply lived in two systems that did not talk. Dale sets a minimum margin for renovation quotes, launches an emergency maintenance membership for landlords, and asks his office manager to record labour hours against every job as it is closed so the comparison can be rerun each quarter.

Common mistakes

• Collecting everything, using nothing Capturing dozens of fields nobody looks at wastes staff time and buries the few numbers that matter. Collect what answers a real question.
• Trusting the spreadsheet over the system Side spreadsheets drift away from the source. Decide on one source of truth for each kind of data and report from it.
• Inconsistent naming The same customer or product recorded three ways makes every analysis a manual matching exercise. Use codes or dropdowns instead of free text.
• Looking only at totals A healthy total hides an unprofitable segment. Always break revenue and margin down by customer, product or job type.
• Waiting for the perfect system A tidy export from what you already have will teach you more this month than a new platform will next year.

KPIs to track

• Questions answerable in ten minutes Of your top five business questions, how many can you answer quickly from existing records? Aim for all five.
• Record completeness The share of key records (customers, jobs, products) with every required field filled. Above 95 per cent is a realistic target.
• Duplicate rate Duplicate customers or products as a share of the list. Under 2 per cent keeps reports trustworthy.
• Days to close the month How long after month end your figures are final. Under ten working days lets you act while it still matters.

Frequently asked questions

Do I need a database or special software to work with data?

Not to begin with. Most small businesses can answer their important questions by exporting from existing systems into a spreadsheet. Dedicated tools become worthwhile when you are repeating the same export every week or joining several systems.

What is the difference between data and information?

Data is a raw fact, such as a sales total. Information is that fact placed in context: compared with last month, broken down by product or measured against a target. Information is what you actually use to decide.

How much history do I need before data is useful?

Twelve months lets you see seasonality, but even one clean quarter will show you your best customers and most profitable work. Start with what you have and keep capturing consistently.

Is customer data covered by privacy law?

Personal information about customers may be covered by the Privacy Act and the Australian Privacy Principles, depending on your turnover and the kind of business. Check the OAIC guidance for whether it applies to you, and see the Data privacy article in Risk & Compliance 101.

Key takeaways

• You already have more data than you think; the gap is turning it into information.
• Start with the decision, then find the three numbers that would settle it.
• Agree on one source of truth and one identifier for each customer, product and job.
• Accuracy, completeness, consistency and timeliness are what make data trustworthy.
• Fix one capture habit a month rather than waiting for a perfect system.