Website Measurement Beyond Pageviews: Events, Enquiries and Outcomes

Most website reports open with traffic: visits, pageviews, bounce rate, time on page, each with an arrow showing whether it went up or down. Then the meeting moves on, because nobody can say what to do differently. Pageviews climb when a post gets shared and dip in the holidays. They rarely tell a business whether the website is bringing in the work it wants.

Useful measurement starts somewhere else: with the decisions the business makes about its website, and the evidence each one needs. Reports should explain business actions. A figure that leaves every decision exactly where it was can move off the first page of the report.

#What a website report should help you decide

A website report should help you decide where effort goes next: which pages to improve, which channels to fund, which services to put forward, and whether the enquiries arriving are the ones the business wants. Pageviews answer none of those alone, because they count attention, not intent.

So write the decisions down before choosing a single metric. For a service business the list is usually short: keep or cut a channel, rewrite or leave a page, change the enquiry form, adjust what sales follows up first. Each decision then names the evidence it needs, and that list becomes the measurement plan.

#The five steps from a first visit to a business outcome

The path from a first visit to a business result has five steps worth recording: the visit, the content read, the enquiry, its qualification, and the outcome. Most analytics setups see the first two clearly, the third partly, and the last two not at all, which is why their reports stop at traffic.

Step	What to record	Where it usually lives
Visit	Source, campaign, landing page	Analytics
Content	Key pages read, such as a service page or a case study	Analytics
Enquiry	Form, call or booking, with its source carried through	Analytics and the enquiry record
Qualification	A fit, or the reason it was not	CRM
Outcome	Proposal sent, won or lost	CRM

The break usually sits between the third and fourth steps. Analytics can name the campaign behind a visit, only the sales record can say whether that visitor ever signed anything, and in most setups the two are never joined.

#Choose events that mark a real business action

A meaningful event marks a step a buyer takes toward working with you: submitting an enquiry form, tapping a phone number, booking a meeting, downloading a specification, or reaching the quote page. Scroll depth and time on page describe reading. They help when diagnosing one page, but they are not outcomes.

Keep the list short and define each event once. "Enquiry submitted" should fire when the form actually succeeds, not when the button is pressed, and it should mean the same thing on every form. Write each definition into a measurement plan the whole team can read, with the decision it serves beside it. Events that exist only because a tool switched them on by default tend to fill reports that nobody acts on.

Then test them as a visitor would. Send the form, tap the phone link on a mobile, and check that each action is recorded once, with its source attached.

#Attribution is an estimate, so report it as one

Attribution assigns credit for an enquiry to the channels a visitor touched before it, and every model does that by rule rather than by observation. Last click credits the final touch, first click the first, and multi touch models spread the credit by formula. None of them can see the conversation at an event, the colleague who recommended you, or the visit on a work laptop before the enquiry from a phone.

Use attribution to compare channels against each other over time, not to settle what caused a single deal. Two habits keep it honest. Ask every enquirer how they heard of you, in a free text field, and set that answer beside the tracked source. And report the share of enquiries with no known source, rather than letting a model quietly hand them to a channel. When that unknown share grows, it is a finding in itself.

#Count qualified enquiries, not form submissions

Enquiry quality is the share of enquiries that fit the work the business wants, and it is a better target than the number of forms. A page that doubles its submissions with spam, job applications and supplier pitches has made the report look better and the sales team's week worse.

Agree a small set of qualification outcomes with the people who answer the enquiries: a fit, not a fit with the reason, spam, or not a sales enquiry at all. Record one on every enquiry. After a few months the reasons say more about the website than the volume does. A service page may be drawing the wrong size of client, a form may invite requests the business does not take, or a channel may send plenty of people who never qualify. Handling each enquiry well is its own workflow, set out in our post on the lead follow-up workflow.

#Connect website enquiries to CRM outcomes

Connecting the website to the CRM means each enquiry arrives carrying its source, campaign and landing page, so the eventual outcome can be traced back to where it started. Without that link, marketing reports forms, sales reports revenue, and the two never meet.

The connection can be simple. Hidden fields on the form pass the source and landing page into the enquiry record, and the CRM keeps them alongside the qualification and the outcome. A monthly view can then show enquiries, qualified enquiries and won work by channel and by page, which is the view that justifies moving budget.

Two cautions. Sales cycles in many service businesses are long, so this quarter's wins reflect enquiries from months ago: compare like periods. And the CRM view is only as good as its records. An enquiry nobody qualified or closed is a hole in the report.

#Privacy and consent in website analytics

Privacy is a design input for website measurement, not a check made after the tags go live. Depending on where your visitors are, analytics and advertising cookies may need consent before they load, and visitors who decline, or who block scripts, do not appear in the numbers at all.

Plan for a partial view instead of treating the totals as complete. Collect the least that answers the decisions on your list. Keep personal data such as names, email addresses and form contents out of analytics events and page addresses; it belongs in the CRM, under its own access and retention rules. Make sure the privacy notice describes what the site actually loads. One record is not affected by any of this: the enquiry itself, which the visitor chose to send, still exists when the cookie banner is declined, even if its source is unknown.

#Start each report with the decision it supports

A useful website report opens with the decisions it supports and then shows the evidence for each. Case study pages produce a large share of the qualified enquiries, so the next quarter adds more of them. One paid channel brings plenty of forms but few that qualify, so its budget moves. The traffic charts can sit behind that, for anyone diagnosing a page.

A report built this way is often shorter than the traffic deck it replaces, and it gets argued with, which is the point. If your reports describe traffic but cannot say which enquiries became work, review your website measurement with us: choosing the events, connecting enquiries to outcomes and building reports the business acts on is part of our growth marketing work.