The following article breaks down the key factors influencing the gap between SA360 Floodlight and Google Analytics data
‘Why don’t our numbers match? ‘Why does the same conversion action report differently across platforms?’ These are some of the most common questions we receive in the Search AdTech team here at Incubeta, especially when it comes to comparing Search Ads 360 (SA360) and Google Analytics (GA) conversion data. The truth is, a 100% match is rarely, if ever, attainable. SA360 and GA are fundamentally distinct platforms, measuring conversions with different methodologies.
However, understanding why these discrepancies occur is crucial. It’s not just about acknowledging them, but about gaining clarity, building trust in your data and making better decisions. For example, you’ll better understand the metrics driving your return on ad spend and fully comprehend your marketing reports.
This article will break down the key factors in influencing the gap between SA360 Floodlight and GA data and is not just written for PPC practitioners. Chief Marketing Officers and budget decision makers will also benefit from understanding the differences.
We’ll explore critical areas like:
- Where in GA the data is pulled from
- Differences between attribution models
- Click-time vs conversion-time reporting in SA360
But first, we’ll need to ensure your tag setup is sound. If the SA360 and GA tags aren’t configured with consistent triggers, that’s a good starting point.
Are your tags triggering correctly?
Before we dig into where to look in which platform and other settings, it’s important to establish that the tags you are comparing across SA360 and GA are firing in the same way. If one tag fires on a “click button” trigger and the other only on a specific confirmation page, you’ve found a major source of discrepancy right off the bat. Aligning these triggers is paramount before you investigate further.
Once your triggers are consistent, we can move onto the next layer of complexity. To truly understand why your numbers may differ, we first need to clarify where in GA you are looking at your conversion data.
GA session-based vs event-based data: Are you looking in the right place?
It’s important to establish where in GA you are pulling data from as GA can report on data in three different ways: user-based, session-based and event-based. The location where you’re pulling data in GA directly impacts how conversions are counted, which then affects the discrepancies with SA360.
Let's look at the two most common areas:
- Traffic Acquisition Reports (Session-based): If you're viewing "Traffic acquisition" within the "Reports" section of GA, you're looking at session-based data. A session begins when a user lands on your website and typically times out after 30 minutes* of inactivity.
Here's an example: A user clicks an email link, browses your site, then clicks an affiliate link for a discount, landing back on your site, and finally makes a purchase. In a session-based report, this purchase would be attributed to "email" because that was the very first touchpoint that initiated the session.
- Advertising Section Reports (Event-based): In contrast, if you're looking in the "Advertising" section of GA (e.g., "Search Ads 360" or "Attribution models"), you're viewing event-based data. Assuming the attribution settings within the property are set to last click, in this view, that same purchase from our example above would show up as "affiliate" because it was the last touchpoint directly before the purchase event.
Understanding these differences is vital. Floodlights in SA360 only report on what GA considers event-based last-click google-paid only data. This means if you’re comparing GA’s session-based data to SA360’s Floodlight data, you’ll see discrepancies by default. It’s tough to put an exact number on this discrepancy, as it varies widely by advertiser and industry. Our best advice? Keep a close eye on your data from different sources over time to understand what a “normal” range looks like for your specific accounts.
*This timeout period can be adjusted, 30 minutes is the default
Attribution models: Are GA and SA360 speaking the same language?
Attribution models play a huge role in how conversions are credited across different touchpoints. Getting these aligned between platforms is key to minimizing discrepancies.
- Session-based GA Data: This data does not adhere to the attribution model settings in GA. It simply reports on the session source that originated the session. Even if a user visits the site across multiple sessions, only the last session before the event will receive credit; the event isn't attributed across multiple touchpoints.
- Event-based GA Data: This does adhere to the attribution model settings you define in GA. Most reports in the "Advertising" section of GA, along with event-based data you pull via "Explore," use these settings. You can check your GA property's attribution model by navigating to Admin > Data Settings > Attribution settings.
Within the "Advertising" > "Attribution models" section, you can even compare event-based data using different models like "last click paid and organic," "last click paid only," and "data-driven paid and organic."
- SA360 Floodlights: By default, Floodlights in SA360 are usually set to "last click." However, this can be changed. It's worth checking your conversion settings in SA360 to align them as closely as possible with your GA attribution model, especially if you're using a "last click" model in GA.
Click-time vs conversion-time data in SA360: Timing is everything
Finally, understanding if you are looking at click-time or conversion-time data in SA360 is important, particularly if your account experiences a significant delay between a click and a conversion (otherwise known as ‘conversion delay’).
- By default, the Advertising section in GA reports on ‘event time’ aligned data, which marries up to ‘conversion-time’ data in SA360.
- You can change this to ‘ad interaction’ which better marries up to ‘click-time’ data in SA360.
In conclusion: Managing the ‘normal’ discrepancy
To reiterate, your GA data and your SA360 Floodlight data will never 100% match up. This is a fundamental reality of working with two distinct measurement platforms.
However, you can significantly reduce discrepancies and improve your understanding of your performance by taking these steps:
- Ensure consistent tag triggers between GA and SA360.
- Compare GA's event-based data (from the Advertising section) with your SA360 Floodlight data.
- Align attribution models as closely as possible, especially if you're using a last-click model.
- Be mindful of click-time vs. conversion-time reporting settings.
By consistently applying these checks and building a repository of different data sources, you'll start to recognise what a "normal" level of discrepancy looks like for your accounts. This understanding is key to trusting your data and making more informed marketing decisions.
What’s the biggest discrepancy challenge you face with GA & SA360 data? Let us know, and see how Incubeta can help you navigate the multitude of complexities that advertisers now face with regards to paid search measurement!
This article was authored by Merel van Steen | Senior Search Consultant at Incubeta, UK