FAQs

Question: I am seeing a mismatch between avails in Gamera Signals and in GAM

Answer: Prebid may not load on all inventory, but Gamera sees everything on page

Question: I’m seeing a mismatch between CTR in GAM and in Gamera’s reporting

Answer: GAM only measures CTR on Direct Sales/AdX, and does not include Prebid. Gamera measures CTR on everything, including Prebid.

Gamera also counts clicks on the "X" button on outstream video ads as a click for CTR measurement, so if you have a lot of outstream ads it may skew your CTR rate as well.

Question: I'm seeing a mismatch between the pageviews in Gamera and in Google Analytics

Answer: The following checks may be helpful:

• Are timezones aligned? The Gamera UI is in UTC

• Is the Gamera tag configured to wait for consent (i.e. including the wait=1 parameter) in the GDPR region, and/or is your PV tracker CMP constrained in the same the way Gamera is? 

• Is there a difference in how the system you’re using for PV tracking is loaded to the page vs. Gamera?

• Do you have any customizations to how you are counting pageviews - for example, firing a custom event during infinite scroll pages - or just use the default PV count?

• Does Gamera load with your ad script bundle, or is it hardcoded on the page? If the former, Gamera won't load in ad blocked situations or on every single pageview

Question: I’m seeing Inventory Quality scores as 0 for some metrics - what does this mean? 

Answer: Gamera only highlights when there are issues, so 0 scores means you are doing well and there is nothing to flag

Question: What on-page troubleshooting tools can I use?

Answer: You can add ?gameraDiagnostics (case sensitive) to the end of any URL where our script is present to open the Diagnostic tool, or enter gamera.loadDiagnostics() in the dev console.

Question: Why does Gamera's Core Web Vitals data differ from PageSpeed Insights / Lighthouse?

Answer: Gamera data reflects what your real visitors experience — actual devices, networks, and browsers — reported at the 75th percentile, per Google's own Core Web Vitals standard. We go further than Google's own field data by measuring across all browsers, not just Chrome.

Lighthouse is different: it's a simulated test, not real visitors. It loads your page once under fixed, intentionally strict conditions (a mid-tier Android device on a throttled slow connection) to stress-test worst-case performance. That's why Lighthouse LCP scores can look dramatically worse than real-world data.

On PageSpeed Insights, real-user data appears under "Discover what your real users are experiencing"; Lighthouse appears further down under "Diagnose performance issues." Our numbers line up with the former.