
Fixing a tracking problem can leave you with fewer reported enquiries, even when the business is receiving just as many.
That can be an uncomfortable result to explain. A website update goes live, the tracking is corrected and the next report shows fewer sessions or conversions. The obvious concern is whether something has gone wrong.
It is a reasonable concern. A falling number deserves investigation, particularly when it influences where a business spends its marketing budget.
The explanation might be that an inflated count has been corrected. It might be that consent changes have altered what analytics can observe. There could also be a genuine decline in demand, traffic or enquiries.
Key principle
A fall in reported performance needs an explanation. That explanation may sit in the business, the measurement, or both.
My starting point is to establish what the number represents and find evidence outside that single report. These are the checks I would make before deciding what needs to change next.
What Was the Tracking Actually Counting?
An enquiry can mean different things in different reports. Clicking “Get a quote”, attempting to submit a form, successfully sending it and becoming a qualified lead are separate actions.
If a report calls all of them conversions, it becomes difficult to know whether the business is attracting customers or simply recording activity.
GA4 uses key events for actions designated as important to the business. Marking an event as important does not establish that its trigger is correct, that the enquiry arrived or that it was a suitable lead.
I would therefore trace the event back to the action that produces it. Does it fire on a click, after validation or after a successful submission? Can one enquiry produce two records because both a website integration and Google Tag Manager send the event?
The key-event counting method also matters. However, limiting a count to once per session can conceal duplicate firing. The implementation still needs checking, including whether someone can legitimately make more than one enquiry.
Internal activity is another possible contributor. Excluding staff visits or test events changes what is collected. Google explains that active exclusion filters affect incoming data; they do not clean up historical figures retrospectively. A lower total after such a change needs that context.
Why Consent Changes Can Alter the Picture
Consent introduces a different measurement issue. The question becomes how much of the visitor’s activity the implementation can observe.
If a consent change means fewer visitors are measured in the same way, reported sessions or enquiries may fall even when their underlying activity has not changed. That is different from removing records that were counted twice.
The implementation matters. With basic Consent Mode, Google tags are blocked until consent is granted. An advanced implementation can send measurements without cookies while consent is denied. Google’s Consent Mode documentation explains these different behaviours.
GA4 may also use behavioural modelling to estimate missing behaviour when a property is eligible. Enabling Consent Mode does not guarantee that modelling is available or that every missing visit and enquiry will be reconstructed.
A smaller observed audience can therefore be an expected consequence of a change. It can also be a sign of faulty implementation, such as consent updates failing to reach a tag.
The useful distinction comes from testing the behaviour. The fact that a drop happened after a consent change gives me a starting hypothesis. I still need to check it.
How I Investigate a Change in the Numbers
I want to connect the reporting change with something I can explain or reproduce. In practice, that means checking the timing, the business records, the traffic sources and the customer journey.
Establish When the Change Began
I would identify the first meaningful change and compare it with website releases, tracking edits, consent updates, filters and changes to event definitions. Campaign budgets, promotions and changes to the service offered belong on the same timeline.
A sudden step down on a release date is a useful clue. A gradual decline may need a broader investigation. Neither pattern proves its own explanation.
Use complete reporting periods where possible. Google notes that reports can change while data is being processed, so an incomplete recent day is a weak basis for a performance verdict.
Check What the Business Actually Received
For enquiry tracking, I would compare the relevant events with successful website enquiries in the form system or CRM. For bookings or sales, I would use the corresponding operational records.
The scope needs to match. A website form event should not be compared with a total that also includes telephone enquiries, repeat customers and manually added leads. Check timestamps, duplicates and test submissions too.
Those records provide another source of evidence, but they need interpretation. Stable enquiry volume can conceal declining lead quality or a weaker channel being offset by another. A CRM’s source labels may also depend on the tracking being investigated.
Compare the Relevant Traffic Sources
For Google organic traffic, compare Search Console clicks with GA4 sessions attributed to Google organic search. Match the pages, date ranges, countries and devices as closely as possible, and account for time-zone differences.
Google explains why clicks and sessions differ. I would look for a change in their usual relationship rather than expect equal totals.
If clicks remain steady while recorded sessions fall, collection or attribution deserves investigation. That does not prove every visitor arrived successfully or that the website continued converting them.
Test the Customer Journey
Test a successful enquiry, an unsuccessful attempt and a repeat visit to the confirmation page. Check what fires, where it is sent and whether the business receives the enquiry. A validation error should not produce the event intended to represent success.
Test consent acceptance, rejection and a visit with no banner interaction. Where choices can be changed, test those updates too. Google’s consent verification guidance explains how to inspect defaults, updates and tag behaviour.
A tag firing is one part of that evidence. Where collection is expected, verify the event at its destination and account for test-traffic filters. If the journey crosses into a booking platform, test through to completion.
This is the practical work behind my GA4 and tracking audits: connecting the configuration with the action the business needs to understand.
What Different Patterns Would Lead Me to Check
The following patterns help choose the next investigation. Each needs supporting evidence before it becomes a diagnosis.
| What the report shows | What I would check next |
|---|---|
| GA4 sessions fall while comparable Search Console clicks remain steady. | Check collection, consent and the affected pages or visitor segments. |
| Recorded enquiry events fall while received website enquiries remain steady. | Check duplicate removal, event definitions, counting settings and exclusions. |
| Organic sessions fall while direct sessions rise. | Check source attribution and the journey before concluding that traffic moved between channels. |
| Relevant enquiries, search visibility and recorded traffic all weaken. | Investigate demand, acquisition and website performance alongside tracking. |
I would also check whether a headline total hides different patterns underneath it. A mobile form problem can matter commercially even if desktop enquiries keep the overall number steady.
A Worked Example: One Enquiry Counted Twice
This is a simplified illustration of duplicate tracking. Imagine a website where a form integration and a Google Tag Manager tag both send the same enquiry event after one successful submission.
The prospect submits once. The business receives one enquiry. With both sends counted, analytics records two events.
The duplicate send is removed. A controlled test then produces one event and one received enquiry. Subsequent reports may show fewer events even if enquiry volume remains steady.
To establish the cause, I would want evidence at each stage: the two sends before the correction, the single send afterwards and the corresponding enquiry record. I would also test unsuccessful submissions and confirmation-page refreshes to check for other unwanted triggers.
That finding supports a specific conclusion: the tested journey was producing duplicate records, and the corrected journey now behaves as intended.
It does not tell us how many historical enquiries were duplicated across every form or device. Simply dividing all earlier totals by two would assume the same fault affected everything, consistently, for the entire period.
The practical response is to document the change, explain the affected comparison and monitor the corrected event alongside received enquiries. The improvement is in how reliably the report represents that action. Any claim about lead growth still needs its own evidence.
When I Would Investigate a Genuine Performance Decline
Finding a measurement issue should not end the investigation when there is also evidence of lost business. I would look more closely at performance if:
- Independently recorded website enquiries, bookings or orders are falling.
- Relevant search clicks or demand have weakened.
- A form or booking journey fails for visitors.
- The decline is concentrated on particular pages, devices or campaigns.
- Enquiry quality has deteriorated even though the volume looks steady.
A website release can affect both a tracking tag and the form it measures. Consent changes can coincide with a quieter trading period. More than one explanation can be true.
The next action should follow the strongest evidence. A reproducible form failure needs attention. A fall concentrated in particular searches calls for a different investigation. An unexplained reporting discrepancy needs further testing before it can support a confident performance claim.
How I Compare Performance After a Tracking Change
Once measurement changes, the old baseline may no longer be directly comparable. That deserves a clear explanation in the report.
I would record the date, what changed and the metrics affected. For example, removing duplicate enquiry events affects that event’s count and potentially any rate or cost calculation based on it. Other measures may remain useful.
Before you compare
Record what changed, when it changed and which measures remain comparable. A new baseline should make the next report easier to interpret.
Keep the historical record and explain the break in comparability. A tracking repair cannot recreate visits or events that were never collected. Historical figures should only be reconstructed where the underlying records support it, with estimates clearly labelled.
Ratios need attention too. If measured sessions fall but a reported enquiry count changes differently, the apparent conversion rate can improve without the website becoming more effective. Check what both parts of the calculation include before celebrating the percentage.
For the next comparison, use a stable event definition and a period that reflects the business’s volume, weekday patterns and seasonality. A low-volume business may need longer to reveal a useful pattern than a busy online shop.
A useful report note should tell the reader what changed, how it affects interpretation and which evidence will be used to judge performance from that point onwards.
What I Want a Marketing Report to Help Me Decide
I want a report to help someone decide where to investigate, what to improve and whether the evidence supports that decision. That requires understanding how the figures were produced.
This is part of how I think about digital growth: individual metrics become useful when they help explain the commercial outcome.
The useful question is what the evidence allows us to decide with confidence.
After a tracking change, being able to explain a smaller number can be more valuable than continuing to rely on an inflated one. Equally, understanding a measurement limitation helps you recognise when another source of evidence is needed.
If your traffic or enquiry reports have changed and you are unsure what to trust, my analytics consultancy starts by checking what is being measured and how it relates to the enquiries or sales your business receives.