Website Analytics Metrics That Actually Matter

Open any analytics tool and you'll find dozens of reports. Most of them are interesting. Few of them change what you do on Monday morning. The useful metrics are the ones connected to a decision.

A good test for any number: if it went up or down by 20% next week, would you do something different? If the answer is no, it doesn't need to be on your main report.

The core: is the site reaching people?

Visitors and sessions

How many people came, and how many times. Use them to see trends, not exact counts. Compare with the previous period and with the same period last year if your business is seasonal.

Sources

Where visits come from: organic search, paid, direct, referral, social, email. This is often more useful than the total, because it tells you which effort is working. See How to Understand Your Website Traffic Sources.

Entry pages

The first page of each visit. They show what attracts people: an article, a product page, the homepage. If a new page starts appearing on this list, something you did is working.

The result: are people doing what you want?

Conversions

The actions that matter to your business: purchases, contact forms, sign-ups, quote requests, calls. Define them clearly and count each one once.

Conversion rate

Conversions divided by sessions (or visitors). Look at it by source and by entry page. An average for the whole site hides most of the useful information: paid visitors may convert well while social visitors don't, or the opposite.

Value per conversion

When possible, assign a value: average order value, or an estimate of what a lead is worth. It turns a traffic discussion into a business discussion.

Context: what happens in between?

Engagement

Different tools measure this differently: pages per session, time on page, or "engaged sessions" (sessions that lasted a while or had some interaction). Use these to compare pages and sources with each other, not as goals on their own.

Returning visitors

People who come back. Useful for content sites and products used regularly. Remember that the same person on different devices may count as new.

Devices and countries

Mainly diagnostic. If conversions on mobile are far below desktop, look at the mobile experience. If a country sends a lot of visits but no conversions, check whether you actually serve that country.

Metrics that are easy to misread

Bounce rate. A visitor who reads your entire article and leaves satisfied is still a "bounce" in many tools. A high bounce rate on a blog post isn't necessarily bad; on a checkout page, it is.

Average time on site. Averages are pulled by a few very long sessions, and some tools can't measure time on the last page viewed. Use it to compare, not as an absolute truth.

Total pageviews. They can grow because the site got more confusing and people need more clicks to find things.

Followers, likes and impressions on other platforms. They describe activity there, not results on your site.

Building a simple report

A one-page report that fits most small and medium websites:

1. Sessions this period vs. the previous one.
2. Sessions by source, with the change for each.
3. Top 10 entry pages.
4. Conversions and conversion rate by source.
5. A list of what was done in the period: campaigns, content, technical changes.

The fifth item is not a metric, but it's what makes the other four understandable. In WEBPilot, it comes from the strategy timeline, recorded per website, and the traffic analytics view is being built to show these core numbers together.

Summary

Focus on reach (visitors, sources, entry pages), results (conversions, rate, value) and a short list of supporting numbers. Treat the rest as tools for diagnosis. And always read numbers alongside what you changed. For that part of the method, see How to Measure the Impact of Website Changes.