How to measure website enquiries that actually become work
Set up reliable form and purchase tracking, follow accepted enquiries into your CRM, and measure which pages bring in qualified leads and jobs.
By Jan Stevens
A contact button can attract hundreds of clicks while the inbox stays quiet. A form can deliver plenty of enquiries while the sales team spends all its time replying to people outside the service area. Website reporting gets useful when it follows the enquiry far enough to tell those situations apart.
Pick one outcome to measure first. For a service business that might be a qualified quote request. For a shop it might be a paid order. Agree what counts before you start changing the dashboard.
In this article
Write down what counts before adding tags
Pick one customer journey and mark the places where the business learns something new. A visitor opening the contact form is interested. A successful submission gives the business something to handle. A qualified lead is one the team has accepted as a viable opportunity. A won job is a commercial outcome. Give each of those a definition everyone can apply, especially if marketing and sales normally use different meanings of the word conversion.
For example, a property maintenance company might define a qualified enquiry as a message about a service it provides, for a town it covers, from a person the team can contact. A spam submission wouldn’t qualify. Neither would a genuine visitor asking about repairs in a region the firm doesn’t serve. Keep both as reasons for closing a record. Over time those reasons show whether the form needs better protection or the page is attracting the wrong audience.
| Stage | Record it when | Where to check |
|---|---|---|
| Started | Someone begins using the quote form. | Website analytics, where collected. |
| Accepted | The site has saved or otherwise accepted a submission. | The form system and a lead event in analytics where collected. |
| Qualified | The team confirms the request fits the service and area. | CRM or booking records. |
| Won | An agreed job is recorded as won. | CRM, booking system or order records. |
Separate interest from a successful submission
Track a click on the contact button if you want to understand navigation. Record an enquiry only once the system has accepted the submission. A phone link click can show intent, but it doesn’t tell you whether the call connected.
Choose a consistent deduplication rule. Test by submitting twice, refreshing a confirmation page and repeating the form after a validation error. Avoid sending names, emails or phone numbers to web analytics.
Test the trigger before interpreting the numbers
For a form, attach the lead event to the success signal from the submission rather than the submit button. Open the analytics debugger and send one labelled test. Check that the event appears once and the matching test enquiry reaches the place your team will use it. Then force a validation error and refresh the confirmation page. Neither should produce another lead. Repeat with the normal production consent settings so you understand when the event is available.
For a shop, test purchase tracking with an order ID that stays the same if the receipt loads again. Test a declined payment, a retry and a paid order. Confirm how your store handles deposits, refunds and cancelled orders before comparing its figures with another system. If the shop takes a deposit now and bills the balance later, decide whether the analytics purchase value means the amount charged or the full order value. Use the same definition in reports and label it clearly.
The website can have a genuine accepted submission even when no analytics event is available. Some visits can’t be measured or attributed. Use the operational record for the number of leads the team actually has to answer and treat analytics as a view of the measurable journeys. Check the data being sent to analytics and keep out contact information and free-text messages.
Give the sales team a simple way to qualify
Agree on a few statuses the team will reliably update. New, qualified and won may be enough. An enquiry that falls outside your service area or is clearly spam can be closed without counting it as an opportunity.
Use the accepted submission reference to match customer records and your internal reporting where appropriate. Review a handful of leads manually first. That catches mistakes such as a booking appearing under a different record because the customer used another email address.
Compare the rate that answers the question
Suppose one service page generates 60 accepted enquiries and 12 are qualified. Another generates 30 and 15 are qualified. The first brings twice as many submissions. The second has a 50% qualification rate compared with 20%, and it brings three more viable opportunities. Look at the reason codes and the service mix before shifting the budget. If the first page is attracting people outside your area, improving the coverage wording may be a better first move than rewriting the form.
Use counts with the rates. Three jobs from four qualified leads looks impressive, but four leads gives you limited evidence. Record the date range and where the lead came from when you know it. Compare like with like, and give open leads time to move through the pipeline. For work that takes several weeks to approve, reporting this month’s closed jobs against this month’s new enquiries can make a healthy process look broken.
Set a clear denominator. Enquiries divided by measurable visits helps diagnose the website journey. Qualified leads divided by accepted enquiries shows the quality of the incoming work. Wins divided by qualified leads tells you about the later sales process once those leads have had time to mature. Write the calculation in the dashboard so it’s clear what changed when one of the numbers moves.
When you compare campaigns, use neutral labels for data the business cannot connect. A booking team can record “asked customer,” “form start page available” or “unknown” without turning every unattributed call into organic search. In a dashboard, always pair the attribution breakdown with the operational total. If a report shows 22 measured leads while the office has 30 accepted submissions, ask how the other eight entered and whether the difference is consistent with your collection setup. Treat that as a reconciliation exercise before you act on the channel mix.
Plan for tests and refunds too. Mark test records in the system that owns the order or enquiry and exclude them from the business figures. If someone cancels an online order or a job falls through, keep the history and update its status. Then the same record can explain why a campaign appears to generate sales while the actual paid or completed work is lower.
Build a report people can act on
Start with enquiries received, qualified opportunities and jobs won. Add a breakdown by the pages or services you’re deciding whether to improve. Keep an unknown category for leads you can’t attribute confidently.
Once a month, pick one step where you can reduce friction. If lots of prospects start the form and few finish, try submitting it on a phone. If enquiries arrive and rarely qualify, make the service area, eligibility or starting price clearer near the form.
Build a monthly review around one decision
Set aside 30 minutes with whoever handles new enquiries. Bring the numbers, a small sample of records and a note of any changes to the website or campaigns. Pick the part of the journey that seems to be causing the most waste. If several likely customers start a quote and give up, watch someone fill it in on a phone. If the forms work but requests take too long to reach a person, the next improvement may belong in the handoff.
Agree which result you want to see after the change and when you’ll look again. A shorter form might bring more accepted messages and a lower qualification rate. A clearer service page might bring fewer messages while giving the team more viable work. The monthly review puts both effects in view so you can decide whether the change helped.

