GA4 shows more form submissions than HubSpot shows new contacts. Marketing and sales call both numbers “leads,” so the campaign report starts with a disagreement about how many there were.
Before changing the tracking, check whether those submissions came from new people. An existing contact can submit another form without becoming another new contact.
The comparison gets harder when the report includes qualification and closed deals. Each step has its own record and timestamp. Start with the specific number you need to explain.
Why HubSpot and GA4 numbers don't match
When a visitor submits a form, GA4 can record the event if collection is configured and permitted. HubSpot may create a contact or update an existing one, depending on the form and identity rules.
A second submission does not automatically mean a second new person. An existing contact might request another resource, correct their details, or return with a different question.
A form-event report and a new-contact report therefore count different things. Labelling both columns “leads” hides a distinction the person reading the report needs to see.
Use the actual metric names: “recorded form submissions” and “new contacts.” Check repeat submissions and existing contacts before deciding that the difference is a tracking failure.
| Reporting question | Start with | Check before comparing |
|---|---|---|
| How many forms were submitted? | GA4 form events or the form submission log | Event collection, repeat submissions, and filters |
| How many new contacts were created? | HubSpot contacts | Creation date and contact matching rules |
| How many contacts qualified? | HubSpot lifecycle stages | Stage definition and the date it changed |
| How much pipeline closed? | HubSpot deals | Deal stage, amount field, and close date |
Should you report from GA4 or HubSpot?
Choose the source according to the decision. For landing-page performance, start with website activity. For lead qualification or pipeline, start with the CRM records your sales team maintains.
GA4 helps explain website behavior. HubSpot holds the contact and deal records used in the sales process. A campaign can bring more website activity without producing more qualified contacts.
Show those measures alongside each other, with their definitions intact. If submissions rose while qualified contacts fell, the difference is worth investigating. Combining them into one “lead” total would hide it.
When is HubSpot the source of truth?
Once the question becomes “Who qualified?” or “What entered the pipeline?”, work from the CRM records and the definitions your sales team uses.
HubSpot lifecycle stages describe where contacts and companies sit in your process. Those stages can be updated in different ways, so a stage total needs context: what moves a record into the stage, and when does that change happen?
HubSpot’s lifecycle-stage documentation is a useful reference when you check that setup. A report is only as clear as the process maintaining its records.
For a sales pipeline view, name the deal stage, amount field, and date field being used. If the team means booked revenue from closed-won deals, say that explicitly. It is a different question from cash collected or revenue recognized in the accounting system.
When is GA4 the source of truth?
Use GA4 for landing pages, sessions, engagement, and recorded events. These help you check what happened on the website before a contact entered the sales process.
If recorded form activity falls, check the page and event collection. If it stays steady but qualified contacts fall, look at form processing, contact matching, and qualification next.
Neither pattern proves the cause. Use it to narrow the investigation, then inspect the relevant records and settings.
How do you explain the gap before the meeting?
Choose a small set of records you are permitted to inspect. Confirm which event or CRM record exists in each system and which timestamp the report uses.
Then check the report settings together. A form submission date, contact creation date, and deal close date can belong to different periods. Attribution windows, traffic filters, and duplicate handling can introduce further differences.
Write down what you find. The next person building the report needs to know, for example, that an existing contact’s repeat submission should not increase the new-contact total.
- Name the metric and the exact event, CRM object, or property behind it.
- Confirm the date field, reporting period, and time zone.
- Record attribution settings where they affect the comparison.
- Document contact matching and exclusions for tests, spam, and internal traffic.
- Explain expected differences and flag anything still unresolved.
Keep unexplained gaps on the report until they are resolved. A difference caused by a known date rule needs a note; a difference you cannot account for needs an investigation.
The platform mismatch guide covers the wider pattern. For the next practical step, use the HubSpot cleanup checklist or the monthly reporting workflow. For credit across touches, see multi-touch attribution in HubSpot.
If you need help tracing the gap, book a tracking audit. Bring the disputed report and the definitions behind its numbers.
Frequently asked questions
Why don't HubSpot and GA4 ever show the same lead total?
HubSpot and GA4 count different things. GA4 starts with website events and attribution rules, while HubSpot starts with contacts, lifecycle stages, forms, deals, and CRM ownership.
Should HubSpot or GA4 be the source of truth?
Use GA4 for traffic and website behavior. Use HubSpot for contact, pipeline, lifecycle, and revenue reporting. Use a reconciliation layer when a single executive number needs both.
How much mismatch is normal between HubSpot and GA4?
Small mismatches are normal because attribution windows, sessions, contact deduplication, filters, and time zones differ. A large or sudden mismatch means the workflow needs review.
Can MetricMaven reconcile HubSpot and GA4?
Yes. We map the source fields, attribution windows, exclusions, and CRM stages so the team can explain what changed and why without rebuilding the report by hand.