Start with what the website is meant to help people do
A conversion is an action that indicates a visitor has made progress toward a goal. On a website without a checkout, that might be submitting an inquiry, registering for an event, downloading an application, or finding the information needed to visit a physical location. The useful question is not “What can we count?” but “What does the visitor need to accomplish, and how does that support our work?”
Write down two or three main purposes of the site before choosing metrics. A local service provider may want suitable project inquiries. A community group may want people to attend activities. A public information site may want visitors to find clear answers without needing to contact staff. These purposes call for different conversions, even if the sites have similar pages.
Separate strong signals from convenient clicks
An action is a stronger conversion when it shows clear intent. A completed contact form usually says more than a visit to the contact page. An event registration says more than a click on the events listing. Neither action guarantees that the visitor will become a client or attend, but each is a reasonable sign of progress.
Some useful actions are harder to interpret. A tap on a phone number may lead to a call, but it may also be accidental. A resource download may mean someone plans to use the document, or simply wants to inspect it. Treat these as supporting signals rather than proof of an outcome. Page views and time on page can help diagnose a journey, but they rarely make good primary conversions on their own.
It helps to name each conversion precisely. “Contact form submitted” is clearer than “lead.” “Directions link opened” is more accurate than “store visit.” Plain names keep reports honest and make it easier for colleagues to understand what the numbers do—and do not—show.
Choose a small set of actions you can act on
For each proposed conversion, ask three questions: Does it represent a useful step for the visitor? Can we measure it reliably? Would a change in the result prompt us to do something? If the answer to the last question is no, the action may not deserve a place in a short monthly report.
A practical starting set might include one primary conversion and two supporting actions. For a nonprofit with a volunteer program, the primary conversion could be a completed volunteer application. Supporting actions might be visits to the eligibility page and clicks to start the application. Reviewing these together can suggest where to look if completed applications fall: Are fewer people reaching the information, or are people starting but not finishing?
Keep accessibility in mind when defining an action. If visitors can complete the same task through a form, an email link, or a phone call, a form-only count will miss part of the picture. You may not be able to attribute every outcome to a website visit, but you can note the measurement gap rather than treating one interaction path as the only valid one.
Measure without collecting more than you need
Record the event at the point where the action is confirmed, when possible. A form confirmation page or successful submission message is generally a better indicator than a click on the submit button, which may occur even when the form has errors. Test the action on a phone and a desktop browser, and check that a refresh does not produce misleading duplicate counts.
Avoid putting names, email addresses, message contents, or other personal details into analytics events. An event label such as “inquiry form completed” is usually enough to understand the trend. Make sure your measurement approach fits the privacy information and settings you provide to visitors. If a tool cannot distinguish a completed action without exposing sensitive details, consider a simpler count from the form system instead.
Review conversions alongside real-world feedback
Choose a review rhythm your team can maintain, such as a short monthly check. Compare conversions with the paths people take to reach them, and note any changes you made to pages or forms. Look for patterns over time rather than treating a small change between two periods as a verdict. For low-traffic sites, individual conversations with visitors or staff may explain more than a chart.
Finally, check whether the measured actions are leading to the outcomes you care about. Are inquiries relevant? Do registrants attend? Are staff receiving fewer repeat questions after an information page improves? These checks may be informal, but they protect you from optimizing for easy clicks instead of useful experiences. A good conversion definition is one your team can explain, measure consistently, and revise when it stops reflecting the work the website is meant to do.
