CrUX Vis or a Custom History API Dashboard: Which Fits Your Weekly Review?
CrUX Vis or a custom History API dashboard for your weekly review? Run one coverage check to see when the fixed layout is enough and when a build actually pays off.

Your team wants to watch Core Web Vitals trends week over week.
You are weighing two options: the CrUX Vis dashboard, or a custom build on the CrUX History API.
The right answer depends on one check: whether the dashboard's fixed layout answers every standing question your weekly review actually asks.
This article walks through that coverage check.
If the prebuilt dashboard covers your questions, you bookmark a shareable URL and build nothing.
If a specific question goes unanswered, that gap—not a general preference for customization—is what justifies a custom build.
What CrUX Vis actually gives you
CrUX Vis is a dashboard that visualizes the weekly historical data from the CrUX History API.
You enter an origin or URL at cruxvis.withgoogle.com, press Go, and get charts.
The URL can then be shared and bookmarked for quick reference, which matters if several people on the team review the same pages.
The dashboard has six pages: Core Web Vitals, Loading Performance, Interactivity, Visual Stability, Ad Metrics, and All Metrics.
Each page shows distributions over time for the past 40 weeks where data is available.
The weekly data updates every Monday, and each data point includes the previous 28 days of data, so the windows overlap.
Each page has a Controls section.
It lets you switch between origin and URL data levels, phone, desktop, tablet, or all devices, the time periods shown, and 75th percentile, distributions, or both.
The metric coverage is broad.
The Loading Performance page includes Largest Contentful Paint, First Contentful Paint, Time to First Byte, LCP resource type, and LCP image subparts.
It also covers Round Trip Time, navigation types, and form factors.
Interactivity covers Interaction to Next Paint.
Visual Stability covers Cumulative Layout Shift.
The Ad Metrics page includes ad count, ad density, and ad weight by CPU and network.
All Metrics puts everything in one place.
What it does not offer is customization.
The documentation is explicit: CrUX Vis does not offer customization options beyond the controls.
Users who want a custom dashboard are pointed to the CrUX History API that powers it, so they can display the data as they want.
Run the coverage check before deciding
Before choosing either path, write down the standing questions your weekly review asks.
Then map each one to a dashboard page and control.
This is the whole decision in miniature.
- If a question maps to one of the six pages with the existing controls, CrUX Vis answers it.
- If a question needs a different layout, a combined view across pages, or automated delivery into a report your team already reads, the dashboard cannot answer it, because its only flexibility is the built-in controls.
- If every question maps, you are done: bookmark the shareable CrUX Vis URL and skip the build entirely.
One caution on the mapping step: a question that seems unmapped may actually be answerable by switching a control.
For example, a question about mobile users specifically is covered by the device control, not a missing feature.
Check the controls before declaring a gap.
Verify data availability first, either way
Whichever option you lean toward, confirm the data exists for your site before committing.
CrUX Vis does not work for all websites.
If your origin is not included in the CrUX dataset, there will be no data to display.
The dataset includes over 15 million origins, but the one you want may not have sufficient data to be included.
Some common issues are worth checking before you conclude data is missing.
Providing the wrong protocol, such as http:// instead of https://, is one.
Omitting the subdomain when it is needed is another.
Some sites redirect, so if http://example.com redirects to https://www.example.com, you should use the latter, the canonical version of the origin.
Also check granularity.
URL-level data may not be available even when origin-level data is available.
If your weekly review focuses on a specific landing page, verify that page has data before building a review process around it.
These availability checks apply to a custom build too, because both surfaces draw on the same CrUX dataset.
A dashboard you build yourself will not conjure data the underlying dataset does not contain.
What a custom History API build involves
If the coverage check surfaces a genuine gap, the CrUX History API is the documented path.
It gives access to historical real-user experience metrics at page and origin granularity.
You can request between 1 and 40 weekly collection periods; 25 are returned by default.
Using it requires a Google Cloud API key provisioned for Chrome UX Report API usage.
The same key works for the daily and history APIs.
The documented request examples include an API key as a query parameter.
Queries are submitted as JSON objects in a POST request to the queryHistoryRecord endpoint.
You specify an origin or a url property, a formFactor such as PHONE, and the metrics you want.
If you omit the metrics property, all available metrics are returned.
If you omit formFactor, values are aggregated across all form factors.
The response structure mirrors the daily CrUX API, except values come in arrays with plural key names, such as histogramTimeseries instead of histogram and p75s instead of p75.
The collectionPeriods array lists the first and last date of each aggregation window.
By default, 25 collection periods are returned.
You can change this by setting collectionPeriodCount in the request to a number between 1 and 40.
The API is updated each Monday around 04:00 UTC and contains data up until the previous Saturday, with the standard two-day lag.
Understand the data shape before you chart it
Two documented characteristics of the data affect any dashboard, custom or not.
First, the windows overlap.
Each collection period contains the previous 28 days of aggregated data, and periods are weekly, so three weeks of data carry into each subsequent period and only one week is new.
The documentation describes this as similar to a moving average.
The practical consequence: a sharp week-over-week change in your chart reflects one new week of data inside a 28-day window, so single-period swings are smoothed.
Plan your review thresholds with that in mind rather than treating each point as an independent week.
Second, eligibility can change over time.
An origin or URL may be eligible for only some collection periods.
In those cases the API returns NaN for histogram densities and null for percentiles for the periods with no eligible data.
For URLs or origins that fall in and out of eligibility, you may notice many missing entries.
A worked example of the coverage check
This example is illustrative, not a real customer result.
Suppose a lean marketing team runs a Monday review with three standing questions.
Are Core Web Vitals trending in the right direction for the site overall?
Is the mobile experience keeping pace with desktop?
And did last month's homepage redesign change loading behavior on that specific page?
Mapping each question: the trend question maps to the Core Web Vitals page.
The mobile-versus-desktop question maps to the device control on the same page.
The homepage question maps to URL-level data on the Loading Performance page.
First verify the homepage URL has URL-level data, since it may be absent even when origin-level data exists.
All three map, so the team bookmarks the CrUX Vis URL for the origin and one for the homepage, and builds nothing.
Now suppose a fourth question existed: pipe the p75 LCP trend into the same report the team uses for pipeline metrics, refreshed automatically each week.
That question needs a different layout and automated delivery, which the dashboard's controls cannot provide.
That single unmapped question—not a vague wish for more charts—would justify a History API build.
That build requires an API key and handling of NaN and null periods.
A short decision checklist
Work through these in order.
Each question has its own consequence.
- Does your origin have data in the CrUX dataset, using the canonical https form with the subdomain? If no data appears, fix the origin form first; if it still shows nothing, the origin may lack sufficient data, and neither option will help.
- Does each standing weekly question map to one of the six pages with the existing controls? If yes for all, bookmark the shareable URL and stop. If one question needs a different layout, combined view, or automated reporting, that specific gap points to a custom build.
- For page-specific questions, does the specific URL have URL-level data, not just the origin? If not, decide whether origin-level data answers the question or whether the page-level question goes unanswered in either option.
- For a custom build, does the team have a Google Cloud API key provisioned for Chrome UX Report API usage, and a plan for handling NaN and null entries from eligibility gaps? If either is missing, that is the next thing to verify before estimating the build.
Where the two options land
For a lean team whose weekly review fits the six pages and four controls, CrUX Vis is the lower-effort choice.
It is shareable by URL and updated every Monday, with no build of your own to maintain.
For a team with a documented, specific question the fixed layout cannot answer, the History API offers the same underlying data with display flexibility.
The cost is an API key and your own handling of overlapping windows and eligibility gaps.
The decision is not about which tool is better in the abstract.
It is about whether your actual weekly questions fit the documented layout.
Run the coverage check, verify data availability in the canonical origin form, and let the specific unmapped question—not customization for its own sake—drive the build.
If you are comparing data surfaces for an engineering build, the choice between the History API and BigQuery is covered separately in CrUX History API or BigQuery?
If the broader question is how monitoring data flows into your existing systems, the glossary entry on run history explains how automated workflow runs are recorded and reviewed.
Source references: developer.chrome.com; developer.chrome.com.
How Meshline can help. Connect automation, Organic Marketing (demand generation), and customer lifecycle management (Revenue Intelligence).
Bring topic planning, content publishing and performance feedback into the conversation about your workflow. Book a Meshline demo.