Tool Guide / Data & Analytics

Data & Analytics: a complete beginner's guide

How to get useful answers out of your own numbers — what to measure, what to ignore, and the free tools that cover it.

Most small businesses have far more data than insight. There is a website analytics account nobody opens, a spreadsheet of orders, a payment processor full of transactions, and no answer to the question that actually matters, which is usually some version of "where do paying customers come from, and which ones are worth having".

The gap is not tooling. It is that analytics is usually approached backwards: install something, look at the dashboard, and hope a conclusion appears. Dashboards do not produce conclusions. Questions produce conclusions, and the data either answers them or does not. Starting from a question you would actually act on is the whole difference between useful analysis and a chart nobody reads.

This guide covers the sequence from question to decision, including the part everyone underestimates, which is cleaning the data. It is aimed at a business owner with a spreadsheet, not at an analyst, and every tool named is free.

How data & analytics works, stage by stage

1. Start with a question you would act on

Writing down the specific question, and what you would do differently depending on the answer.

What matters: If no answer would change your behaviour, the analysis is entertainment. "Which marketing channel produces customers who stay longest" is a real question because it changes where money goes. "How much traffic did we get" usually is not, because the answer rarely changes anything.

Common beginner mistake: Opening a dashboard and browsing for something interesting. This reliably surfaces whatever is easiest to measure rather than whatever matters, and it feels productive while producing nothing.

2. Find where the data already lives

Locating the systems that already hold the answer: payment processor, booking system, spreadsheet, website analytics, inbox.

What matters: The most valuable data in a small business is usually in the payment system, because that is where actual money is recorded, and it is the one people check least. Website analytics describes interest; the payment record describes revenue, and they frequently tell different stories.

Common beginner mistake: Building new tracking before using what already exists. Most first questions can be answered from records you already have, and setting up new collection delays the answer by months.

3. Clean it

Getting the data into a consistent shape: one row per thing, consistent dates, consistent spellings, duplicates removed.

What matters: This is the majority of the work, and skipping it produces confident wrong answers. Watch for the same customer entered three ways, dates stored as text, and totals that include cancelled orders. Fix the source of the mess where you can, because otherwise you will do this again every month.

Common beginner mistake: Trusting a total without checking what it includes. Refunds, cancellations, test orders and internal purchases quietly inflate almost every first analysis.

4. Look at it before you calculate anything

Sorting, filtering and charting the raw data to see its shape before computing averages.

What matters: Averages hide almost everything interesting. Sorting by value and looking at the top and bottom of a list tells you more in two minutes than a summary statistic will, particularly in small businesses where a handful of customers often account for most revenue.

Common beginner mistake: Jumping to an average. If a few large customers dominate, the average describes nobody, and decisions made from it will be wrong for both ends of your customer base.

5. Build the one view you will actually check

A single small dashboard with the few numbers that drive decisions, that you will look at weekly.

What matters: Small and habitual beats comprehensive and ignored. Five numbers you check every Monday are worth more than forty you check once. Include a comparison to the previous period, because a number alone is meaningless without direction.

Common beginner mistake: Building an impressive dashboard with thirty charts. It gets admired once and never opened again, and the effort would have been better spent answering one question properly.

6. Decide something

Making an actual change based on what you found, and writing down what you expect to happen.

What matters: Recording the prediction before the change is what turns analysis into learning. Without it, memory quietly rewrites what you expected to match whatever occurred, and you learn nothing from either outcome.

Common beginner mistake: Analysing continuously without deciding anything. Perpetual measurement is a comfortable substitute for acting on uncomfortable findings.

7. Check whether it worked

Returning to the same numbers after the change and comparing against what you predicted.

What matters: Leave enough time for the change to show up, and be alert to everything else that changed in the same period — seasonality especially. Most small-business "results" are seasonal patterns wearing a costume.

Common beginner mistake: Declaring success after a few days. Short windows are dominated by noise, and acting on noise means undoing good decisions as often as bad ones.

Choosing your tools

Do I even need analytics?

You need enough to answer where customers come from and which are profitable. Beyond that, most small businesses get more from asking every new customer how they found you than from any tool. Analytics earns its keep when you are spending money on marketing and need to know which spending works.

Spreadsheet or a real database?

Spreadsheet, for far longer than people expect. They handle tens of thousands of rows, everyone can read them, and every tool exports to them. Move to a database when several people need to edit simultaneously, when you need the same data in multiple places without copies, or when the file becomes slow. Most small businesses never hit any of those.

Which metrics actually matter?

For most small businesses: how many enquiries you get, what share become customers, what a customer is worth over their lifetime, and what it costs to acquire one. Those four support nearly every real decision. Traffic, followers and impressions are inputs to the first one, not goals.

Can I just ask AI to analyse my data?

For exploration and explanation, yes and it is genuinely useful: describing what a chart shows, suggesting angles you have not considered, writing spreadsheet formulas, explaining a statistical term. Be careful giving it customer data, since that may be personal information with obligations attached. And check its arithmetic, because it will produce a confident wrong number as readily as a right one.

Should I hire an analyst?

Not before you have specific recurring questions and clean data to answer them with. An analyst arriving to a mess spends months on cleanup, which you could have avoided by fixing the collection. Fix the inputs first; the analysis is the easy part.

A complete free starting setup

Tools covered in this category

Frequently asked questions

What is a vanity metric?

A number that reliably goes up and does not affect income: followers, page views, impressions, email opens. They are not worthless, but they are inputs at best. The test is simple — if it doubled, would you make more money? If the answer is not clearly yes, do not manage by it.

How much data do I need before analysis is meaningful?

Less than you think for large differences, more than you think for small ones. If one channel produces ten times another, thirty data points make that obvious. If you are comparing a five percent difference, small samples will mislead you repeatedly. Be suspicious of small differences in small numbers.

Why do my numbers never match between tools?

Because they measure different things and define them differently: a session, a user and a visit are not the same, ad blockers hide some traffic, and time zones and attribution windows differ between platforms. This is normal. Pick one tool as the source of truth for each question and stop reconciling.

Do I need a data warehouse?

Almost certainly not. Warehouses solve the problem of many large data sources needing to be combined reliably, which is a problem of scale most small businesses never reach. A spreadsheet and a free dashboard tool cover it.

What is the single most useful thing I can start tracking?

Where each new customer came from, recorded at the point of sale, in a spreadsheet. It takes seconds per customer, captures the referral and word-of-mouth routes no tracking tool can see, and answers the question most likely to change how you spend money.

Last reviewed: 2026-08-07.