Analytics and Tracking

Analytics

Also called digital analytics, marketing analytics

The practice of collecting behavioural data and interpreting it well enough to change what you decide to do next.

Quick facts: Analytics

Category
Analytics and Tracking
Also called
digital analytics, marketing analytics
Level
Beginner
Affects
Budget allocation, campaign decisions, site changes, reporting
Where to see it
GA4, Google Search Console, Looker Studio, ad platform reporting
In this article4
  1. How analytics works
  2. Why analytics matters
  3. Common mistakes with analytics
  4. How to act on it

How analytics works

Analytics has three parts that people routinely blur together. Something records what happened — a tag on a website, an ad platform logging clicks, a till counting orders. Something stores those records and groups them into countable things like sessions, enquiries or sales. Then a person reads the result and decides whether anything should change. Only the third part is analytics in the sense that matters; the first two are collection.

The records are not neutral. They contain only what someone chose to capture, they carry the labels that person applied, and they stop at the boundary of the system doing the recording. An ad platform sees its own clicks and its own conversions. A website tool sees visits and on-site behaviour. Your accounts see money received. None of them sees the whole journey, which is why analytics is a reading exercise rather than a lookup.

Why analytics matters

Marketing budget that is never measured gets allocated by whoever argues best in the meeting. Measurement replaces that with evidence: which channels bring people who actually buy, which pages lose them, which campaign is quietly carrying the others. For a small business the biggest gain is usually in stopping things rather than starting them — cutting spend that has never produced an enquiry costs nothing and frees money immediately.

It also protects you from your own memory. A month that felt busy and a month that felt slow can turn out almost identical once counted, and a seasonal dip gets blamed on whatever was changed last rather than on the season. Records settle that kind of argument in minutes.

Common mistakes with analytics

The most common is collecting far more than anyone reads. A dashboard with dozens of charts and no stated question becomes wallpaper: glanced at, never acted on. A handful of numbers, each attached to a decision someone is willing to make, beats a complete picture nobody uses.

The second is expecting exactness. Ad blockers, consent choices, people moving between phone and laptop, and platforms counting conversions by their own rules all mean the totals will never agree perfectly. Treating that gap as a bug to be eliminated burns weeks. Treating it as a known margin, and comparing like with like inside one tool, gets the decision made.

The third is judging on too little evidence. Short periods and small conversion counts swing hard for no reason at all, and reacting to that swing usually leaves you worse off than doing nothing.

How to act on it

Start from the decision, not the data. Write down the questions the business genuinely asks each month — where enquiries come from, which pages convert, what a lead costs, whether the trend is up or down — and build only the reporting that answers them. Everything else can wait until somebody asks for it, and a focused reporting dashboard is far easier to keep honest than a sprawling one.

Then make the reading a habit rather than an event. A short monthly review that always looks at the same things, with a written note of what changed and why, compounds faster than an occasional deep dive. If the underlying numbers are not trusted yet, fix collection first: a small, clearly defined set of events is worth more than a rich set nobody believes. A measurement plan is the normal way to agree that scope before anyone builds anything.

Do and do not

Do

  • Decide the question before you open a report
  • Compare periods of equal length and similar season
  • Name the decision each number is meant to inform

Do not

  • Report numbers nobody is willing to act on
  • Expect two different tools to agree exactly
  • Read a short, noisy period as a trend

Questions people ask about this

What is the difference between analytics and reporting?

Reporting is the presentation of what happened: a table, a chart, a monthly summary. Analytics is the step after it, where someone asks why the numbers moved and what ought to change as a result. A business can have excellent reporting and no analytics at all, and that is usually why nothing improves from one month to the next.

Do I need paid analytics tools?

Not to begin with. A correctly configured free web analytics tool, the search data already sitting in Google Search Console, and the reporting inside your own ad accounts will answer most of the questions a small business has. Paid tools earn their place later, usually for competitor research, rank tracking, or joining data from several systems into one view.

Why do my analytics numbers never match my ad platform?

Because they count different things under different rules. An ad platform credits a conversion back to the click that led to it, sometimes days earlier, while a website tool credits the last source it saw during that visit. Add consent choices and blocked scripts and the totals will always differ. Compare trends inside one tool rather than chasing an exact match.

Related terms

Found this useful?

Share it, or ask an AI to summarise it

Back to the glossary

Knowing the term is the easy part

Applying it to your own site and budget is the work. Book a call and I will tell you what actually applies to you.