Machina

Machina — Marketing that moves numbers.

Reach out today

Industries

One growth engine.

Design, advertising, development, and marketing built for the sectors driving what comes next — AI, fintech, health tech, and beyond.

All industries

Machina — Marketing that moves numbers.

Reach out today
Technology

Why your website and sales team count different numbers of leads

Your website reports a lead, but sales cannot find the person. Or sales has more inquiries than the dashboard. Before changing spend, trace one request through the saved record, delivery to sales, analytics report and later outcome. Definitions, dates and privacy choices can explain a gap. A failed handoff or duplicate needs repair.

Francisco Contreras

Francisco Contreras · Founder, Machina

4 min read

A vermilion plate cut with parallel grooves, one groove traced end to end by a cream line and the next holding a solid point beside a faint duplicate, on a black background.

First decide what a lead means in each report

The website says leads arrived. Sales cannot find every person. Before buying more traffic or blaming the sales team, find out whether those reports count the same action. A form interaction, an accepted request and a person ready for a proposal are three different stages. We write the definition beside each number before we call a report wrong.

For a business that sells after a conversation, we follow a request through acceptance, the saved record, delivery to sales, review and final outcome. A phone call needs its own record and status. We count an accepted request as an inquiry, then record qualification and won work when those decisions happen.

Use accepted inquiries to judge whether the website delivers contact requests. Use reviewed, qualified records to judge lead quality. Use won work and its value to judge business return. Keep the earlier stages visible so you can see where the counts begin to differ.

Follow one inquiry from submit to sales status

We send a clearly labeled test request without using a real customer’s details. We submit the public form once and note the time, page and form. Did the site save one record with an internal ID? Did the inbox or sales system receive one usable request, with the service requested and a person responsible for it? We look for the matching count in analytics, then confirm sales can update the status without losing the original request.

We keep a small comparison sheet with the internal ID, submission time, page, saved record, confirmation that sales received it, analytics count, owner and status. Personal contact details stay in the sales system, out of the sheet and analytics report. The internal ID connects the business records. The known test time and action name help us find the analytics entry when the systems cannot share that ID.

Our published measurement plan counts a lead after the request is saved. Earlier form activity remains a separate signal. We saw why that distinction matters in an anonymized pest-control audit. The company’s analytics counted page visits, but messages meant to report saved inquiries went to the wrong address. The audit recorded a 404 response. None of those saved-inquiry messages reached analytics. Traffic in a dashboard could not prove that inquiries were being counted, so we test delivery at each step.

Where did the counts part company?

If the site saved the test but sales never received it, check where the notification went and whether the connection to the sales system failed. If sales received it but analytics did not, check whether the website sent the inquiry to analytics and whether the visitor’s privacy choice allowed it to be counted. If analytics counted something but the site saved nothing, it may have counted a form attempt instead of an accepted request.

Then look for duplicates. Compare internal IDs and times. Did a retry create a second sales record? Did the website and analytics each count the same form as a lead? Google Analytics can automatically record form starts and submissions. That form submission is not proof that the business saved an inquiry. We name each recorded action for what it means and count each saved inquiry once as a lead. Adding every form action together inflates the lead count.

A timing gap can also look like a missing lead. In a hypothetical example, a request is saved Monday and qualified by sales Tuesday. Monday’s inquiry count includes it. Tuesday’s newly qualified count includes it. Compare submission dates to see whether sales received the requests. Compare status-change dates to see how many were qualified that day. When the same record appears in both places and the dates explain the difference, label the counts clearly. For remaining gaps, align the time zone, included forms and status rules before hunting for lost records.

What should pass before you trust the report?

For our release check, one test request should save one inquiry and reach one person responsible for it in sales. The business records should share an ID. If the visitor’s privacy choice allows measurement and the website has the information needed to connect the visit, analytics should count that accepted inquiry once. We also test a form rejected before it saves. It should show an error, create no inquiry and add no accepted-inquiry count. Finally, we check that sales can update the status on the same record.

Google Analytics lets you mark an action important to your business as a key event. Marking a button click does not make it a reviewed lead. We choose the action that reflects the agreed outcome and keep clicks and form starts for diagnosis. After a form or sales-system change, we run the test again.

A won job belongs to a later decision than an inquiry. We give it a separate status and value. In our offline-revenue chapter, the won-job upload is still a plan pending account access. We build the lead record and status history first, because an upload is only as sound as the record behind it. Then we decide whether connecting won work to advertising is worth the extra work. Our marketing analytics and conversion tracking work covers the measurement plan, connection tests and reconciled counts across those tools.

Decide whether to relabel or repair

When definitions, dates or privacy choices explain the difference, label the reports and keep the underlying inquiry visible. When a request truly fails to reach sales or analytics, or gets counted twice, give the broken handoff an owner and test it again after the fix. Do this before changing spend based on either count.

When the handoff crosses your website, analytics and sales tools, we trace a real request from its sales status back to its source. We agree on what counts, test each connection and reconcile the reports. Reach out to Machina with the two reports that disagree and one inquiry your team can safely trace.

Next steps

Let's talk about your project

Tell us what you're working on. We'll tell you if we can help — and exactly what we'd do.

We use cookies and analytics to understand how visitors use our site and improve your experience. Privacy Policy · Do Not Sell or Share My Personal Information