Start with the job each page needs to do
A pageview records an arrival, not a successful visit. Someone might open a page, find exactly what they need in seconds, and leave. Another person might spend several minutes searching without finding an answer. If you judge both visits by pageviews alone, you miss the difference that matters: whether the content did its job.
Before choosing a metric, write down the purpose of the page in one sentence. A service page might help visitors decide whether to make an enquiry. A help article might answer a recurring question so visitors can complete a task on their own. An event page might help people find a location and prepare to attend. For each page, name one useful next step a visitor could take and one question the page should answer. This gives you a standard against which to assess the content, even when traffic is modest.
Choose signals that fit the purpose
Useful signals depend on the page. For an enquiry page, you might count visits to the contact page or completed form submissions. For an instructional page, you might look at clicks to the next step, downloads of a relevant resource, or searches made immediately afterward. A search for the same answer after reading an article may suggest that the article needs work, although it is not proof. Record a small set of signals for each important page rather than applying one score to every page on the site.
Direct feedback can fill gaps that analytics cannot. Questions received by email, comments from staff who speak with visitors, and a short optional prompt such as “Was this page helpful?” can reveal confusion. Keep feedback easy to provide, and avoid treating a handful of responses as a verdict on everyone’s experience. Together, behavior and feedback provide a better starting point than either one alone.
Give numbers enough context to be useful
A low-traffic page can still be valuable if it answers a specific, important question. A highly visited page may be underperforming if visitors cannot find the action they came to take. When reviewing a number, ask what changed around it. Did you update the page title, navigation or call to action? Is the page meant for a narrow audience? Are visitors arriving from a link that promises something different from what the page provides?
Compare a page primarily with its own earlier results and with pages that serve a similar purpose. Note the time period and any substantial changes when you record the figures. On a small site, a few visits or actions can noticeably change a percentage, so look for repeated patterns rather than reacting to a single spike or dip. If a measurement tool cannot reliably capture an action, label it as an estimate or leave it out instead of presenting it as a precise result.
Measure responsibly and keep the setup manageable
You do not need to track every movement a visitor makes. Start with information you can use to make a decision: page visits, paths to important tasks, completed actions where they can be measured, and recurring questions. Avoid collecting personal details simply because a tool makes it possible. Check that your measurement approach matches the privacy information you give visitors and the settings available in your tools.
Create a short measurement note for each priority page. List its purpose, the signals you will check, where those signals come from, and any limitations. For example, a click on an email link shows interest but does not show whether an email was sent. A form submission shows that a form was completed but not whether the visitor ultimately received the help they needed. Writing down such limits helps everyone interpret the results consistently.
Turn a regular review into one small improvement
Set aside a recurring review period that suits your capacity. Pick a few important pages, examine their signals and read any related feedback. Then choose one change you can explain: make a heading clearer, move a key answer higher on the page, add a missing next step, or revise a link label so visitors know where it leads. Record what you changed and why, then check again after the page has had time to receive visits.
The aim is not to make every number rise. A clearer help page might lead visitors to finish their task without opening several more pages. A more specific page title might attract fewer irrelevant visits while better serving the people it is for. Measure content performance by connecting observable behavior to the page’s purpose, acknowledging what the data cannot tell you, and using what you learn to make the next useful edit.
