Field Data vs. Lab Data: Which One Should You Trust?

You run a Lighthouse audit and get a performance score of 92. You check PageSpeed Insights and it says your LCP is "Poor." What's going on?

The answer lies in the difference between lab data and field data — and understanding both is essential for making smart performance decisions.

Lab Data: The Controlled Experiment

Lab data is collected in a controlled, simulated environment. Think of it as a performance test in a laboratory — same conditions every time.

How Lab Data Is Collected

• A tool (Lighthouse, WebPageTest) loads your page
• CPU and network are throttled to simulate a mid-tier mobile device
• The page is loaded from a single geographic location
• No real user behavior is involved

Lab Data Metrics

• First Contentful Paint (FCP)
• Largest Contentful Paint (LCP)
• Total Blocking Time (TBT)
• Cumulative Layout Shift (CLS)
• Speed Index
• Time to Interactive (TTI)

Pros of Lab Data

• Reproducible — consistent environment for A/B testing
• Actionable — comes with diagnostics and fix recommendations
• Proactive — test before deploying to production
• Works on any page — no traffic required

Cons of Lab Data

• Not real users — simulated conditions may not match reality
• Single device/location — misses geographic and device diversity
• No interaction data — can't measure real user interactions
• Variability — scores can fluctuate 5-15 points between runs

Field Data: The Real World

Field data (also called Real User Monitoring or RUM data) is collected from actual users visiting your site with real devices on real networks.

How Field Data Is Collected

• Chrome browsers (with user consent) report performance metrics to Google
• Data is aggregated into the Chrome User Experience Report (CrUX)
• Covers 28 days of rolling data from real visits
• Available at origin level and URL level

Field Data Metrics (Core Web Vitals)

• Largest Contentful Paint (LCP)
• Interaction to Next Paint (INP)
• Cumulative Layout Shift (CLS)

Pros of Field Data

• Real user experience — reflects what actual visitors see
• Device diversity — includes slow phones, old tablets, fast desktops
• Geographic spread — captures users from all locations
• Interaction metrics — INP measures real user interactions
• SEO-relevant — Google uses CrUX field data for ranking signals

Cons of Field Data

• Requires traffic — no data for new or low-traffic pages
• Slow to update — 28-day rolling window means changes take weeks to appear
• Less actionable — tells you there's a problem but not always why
• No staging support — can't test unpublished pages

When They Disagree

It's common for lab and field data to tell different stories. Here's why:

Good Lab, Bad Field

Your Lighthouse score is 90+ but field CLS is Poor.

Possible causes:

• Ads or third-party scripts cause layout shifts only for real users
• Slow devices in the field can't run JavaScript as fast as your lab machine
• Users on 3G networks experience much worse loading
• A/B testing tools inject content that shifts layout

Bad Lab, Good Field

Lighthouse gives you 60 but field data says all Core Web Vitals are Good.

Possible causes:

• Lighthouse simulates a slower device than most of your actual users
• Your audience is primarily desktop users on fast connections
• Browser caching helps repeat visitors (lab always tests cold cache)
• Your CDN serves content faster to real users than to Lighthouse's test location

The Recommendation: Use Both

Scenario	Best Data Source
Debugging a specific issue	Lab data
Testing a fix before deploy	Lab data
Checking SEO ranking signals	Field data
Understanding real user pain	Field data
Setting performance budgets	Both
Comprehensive audit	Both
CI/CD automated testing	Lab data
Executive reporting	Field data

Building a Complete Monitoring Strategy

1. Lab monitoring — Run Lighthouse audits on a schedule to catch regressions early
2. Field monitoring — Track CrUX data to understand real user experience
3. Alerting — Set thresholds on lab metrics and get notified when they cross
4. Trend analysis — Track both lab and field metrics over time to spot patterns

Don't Choose — Monitor Both

BadPageSpeed runs automated Lighthouse audits on your pages so you always have fresh lab data. Pair it with CrUX field data for a complete picture of your site's performance.

Start monitoring with lab data →