Conversion Tracking

Hardial Singh 5 Reputation points
2026-08-07T12:47:22.65+00:00

Hi everyone,

I'm experiencing an unusual issue with one of my Microsoft Display campaigns and wanted to see if anyone has come across something similar.

Over the past 3–4 days, I've seen a sudden spike in conversions. The campaign has recorded around 20 leads in just 3 days, whereas it typically generates only 1–2 leads in an entire month. No significant changes have been made to the campaign, bidding strategy, targeting, or conversion setup.

The strange part is that GA4 has recorded only 4 actual form submission events, while Microsoft Ads is reporting 20 conversions.

I've already:

  • Verified that the GA4 form submission event is firing correctly.
  • Checked the Microsoft Ads conversion tracking, and everything appears to be configured properly.
  • Confirmed there haven't been any recent implementation changes.

One additional observation is that all of these reported conversions came from desktop devices.

Has anyone experienced a similar issue? Could this be related to invalid traffic, enhanced conversions, attribution, or another tracking discrepancy? Any suggestions on what else I should investigate would be greatly appreciated.

Microsoft Advertising | Reports | Conversion tracking
0 comments

2 answers

Sort by: Most helpful
Most helpful Newest Oldest
  1. MS Advertising - Dagmara 640 Reputation points Microsoft External Staff Moderator
    2026-08-07T15:11:38.3233333+00:00

    Hello Hardial Singh,

    Thank you for using our Microsoft Advertising Learn Q&A Platform! 

    A discrepancy between Microsoft Advertising and GA4 can happen even when the tracking implementation appears correct, because the two platforms use different attribution and reporting methodologies. Microsoft Advertising attributes conversions based on ad interactions tracked through UET, while GA4 uses its own attribution models and reporting logic. As a result, conversion counts will not always match exactly.

    Given the size of the difference you're seeing (20 conversions in Microsoft Advertising vs. 4 form submissions in GA4), I would recommend investigating a few areas:

    • Conversion goal configuration: Verify that only the intended conversion goal is firing and that there are no duplicate conversion goals matching the same user action. Microsoft notes that multiple goals can match a single event and increase reported conversion counts.
    • UET implementation: Check that the UET tag is not installed more than once on the conversion page. Duplicate UET implementations can lead to double counting.
    • Conversion counting settings: Review whether the goal is configured to count All conversions or Unique conversions, as this can significantly impact lead counts when the same user converts multiple times.
    • Conversion window and attribution: Microsoft Advertising allows conversion windows of up to 90 days. Conversions may be credited back to earlier ad clicks, and Microsoft Advertising reports conversions based on the date of the ad click, whereas other analytics platforms may report them differently.
    • Enhanced Conversions: If Enhanced Conversions are enabled, Microsoft Advertising can supplement conversion attribution using first-party data such as hashed email addresses or phone numbers, potentially identifying additional conversions that may not align perfectly with GA4 reporting.

    How does conversion tracking work?
    Why do I see a discrepancy between Microsoft Advertising conversion counts and those from Google Analytics and Google Ads (or my own logs)?
    https://learn.microsoft.com/en-us/answers/questions/5778112/ga4-and-microsoft-advertising-conversion-data-disc

    Regarding the fact that all reported conversions came from desktop devices, I would also suggest reviewing the conversion reports at the campaign, audience, and placement level to identify whether a specific source is driving the increase. Additionally, Microsoft recommends checking for test, duplicate, or potentially fraudulent conversions occurring on the website itself.

    If everything above looks correct, it may be worth comparing the individual conversion timestamps in Microsoft Advertising and GA4 to determine whether the discrepancy is related to attribution timing versus actual event collection.

    Our support teams are pleased to engage in more in-depth discussions about your account through phone, chat, or email to offer review assistance, please see our support page to reach out!

    I trust that the details shared here will partially address your inquiry. Should you have any further questions, feel free to contact our support. Additionally, I’ve sent you a private message, please don’t hesitate to reply if you require further assistance.

    Kind regards, 

    Dagmara | Microsoft Advertising Specialist | 800-518-5689

    Was this answer helpful?

    0 comments

  2. AI answer

    2026-08-07T12:47:45.8+00:00

    This pattern fits a reporting discrepancy more than a true lead surge.

    The main checks supported by Microsoft Advertising guidance are:

    1. Compare the exact same time frame and time zone
      A mismatch here can make two systems look inconsistent even when they are not. Daylight Saving differences can also cause confusion.
    2. Check how each system counts conversions
      Microsoft Advertising documentation notes that systems can count events differently. In the same way impression counts can differ between platforms, conversion totals can also diverge when attribution and event processing differ. Offline attribution in Microsoft Advertising is designed to more accurately apply the last-view/last-click model, including cases where the time between ad interaction and conversion is very short.
    3. Re-validate the conversion pixel implementation
      Microsoft Advertising specifically calls out conversion mismatches caused by conversion pixel configuration. Any modification can affect pixel functionality. If a tag manager is involved, confirm it is not altering the original pixel code.
    4. Check for duplicate or incorrect pixel setup on the advertiser page
      Microsoft guidance for suspicious traffic investigation explicitly recommends checking whether pixels were set up correctly and whether duplicate pixels exist on the page.
    5. Investigate possible suspicious traffic
      A sudden jump from 1–2 leads per month to 20 in 3 days, especially concentrated on desktop, is a strong reason to investigate traffic quality. Break reporting out by:
      • publisher, placement, or domain
      • creative or creative size
      • geographic area
      • day across the affected period
    6. Run a Site Domain Performance review
      If the spike is concentrated on a limited set of domains, inspect those domains for an explanation.
    7. Check whether the volume is statistically meaningful
      Microsoft Advertising notes that CTR or conversion-rate numbers should generally have at least 10,000 impressions on a given slice of inventory before being considered significant. If the spike comes from a small inventory slice, the rate may look extreme without being reliable.
    8. Escalate with evidence if the pattern remains isolated
      If the issue still looks abnormal after the checks above, send support the affected time range plus demand-side object data grouped by day, including clicks, CTR, conversion rate, and total conversions.

    Given the details provided, the highest-value next checks are:

    • confirm Microsoft Ads and GA4 are using the same date range and time zone
    • inspect the conversion pixel for duplication or modification
    • segment the campaign by placement/domain, geography, creative, and desktop device over the last 3–4 days
    • review whether the spike is coming from a narrow inventory slice that may indicate suspicious traffic
    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.