Reshaping our UX score and how RUMvision is aligning with FID changes

Summary: As FID is being removed from Google interfaces, endpoints and libraries, we will need to align accordingly. The calculation of our UX score will be reshaped as well.

  • by Erwin Hofman
  • Published
  • Reading time ± 2 minutes
  • RUMvision FID
Reshaping our UX score and how RUMvision is aligning with FID changes

The Core Web Vitals INP swap is happening today. Hello INP and goodbye FID. If you're a RUMvision user and are looking to use RUMvision in your pagespeed journeys, you might want to know what changes are occurring within RUMvision.

FID being removed

Last week, we covered how Google is deprecating FID within its tools. Here's how RUMvision is aligning with this change.

Tracking library

RUMvision is built on the web-vitals library. With Google planning to deprecate FID in the next web-vitals.js release, we have ceased tracking FID as of today.

Real-time dashboard

Since no new FID data will be collected, our real-time dashboards and technical tabs no longer display FID information. The FID metric has been removed from the dropdown menu in our technical dashboard, aligning with Google's removal of FID.

This will prevent site owners from looking at the FID metric while having better metrics around (INP), FID is phased out entirely.

Competitors

Our competitor module utilizes Google CrUX data. Although this API will contain FID data for 6 more months, we have opted to remove FID from our competitor dashboards for the same reasons mentioned above.

New score calculation

With FID being removed, we needed to adjust the calculation of our UX scores both within our Audience dashboard and competitors' listings.

Public tool

We previously adjusted the UX score within our public and free UX score checker with today's change in mind. We provided detailed explanations and calculations for this pagespeed UX score in August 2023.

Private tool

Now that FID has been removed, we must update the UX score calculations in our authorized tool. The table below illustrates the previous and new weights assigned to different metrics, with weights represented in percentages.

MetricGood thresholdsUnitOld weightNew Weight
Cumulative Layout Shift0.1 2530
Interaction to Next Paint200ms2530
Largest Contentful Paint2500ms2525
First Contentful Paint1800ms1510
Time to First Byte800ms55
First Input Delay100ms50

The reasoning behind these new weights are as follows:

  • There exists a strong correlation between LCP, FCP, and TTFB.
  • For instance, a poor TTFB adversely impacts FCP and LCP.
  • While LCP is a Core Web Vitals metric, from a user experience standpoint, FCP holds greater significance than TTFB.
  • CLS and INP represent a distinct aspect of user experience compared to LCP.

Hence, we have chosen to allocate 30% each to INP and CLS, with 25% assigned to LCP. This exception for LCP is based on the understanding that a portion of LCP is already encompassed within FCP and TTFB.

Impact on historic data

Since scores are calculated retrospectively, UX scores on historical metrics will automatically adjust. This enables safe comparison of UX and FUX scores over extended periods. For example when comparing UX trend lines from before and after March 12.

However, if you've captured screenshots of previous scores, you may find yourself comparing scores based on different metric weights. In such cases, adjusting the date range to match the conditions in the screenshot and creating a new screenshot is recommended.

Share blog post