Page experience has been a Google ranking factor since 2021, but measuring it well is more nuanced than running a single test and calling it done. At the same time, AI is opening entirely new pathways for how research gets discovered, including geospatial ones we haven't had to think about before. This month, we're looking at how our industry measures site performance for search, and how a feature as unassuming as Google's new "Ask Maps" points toward a meaningful discoverability opportunity for scholarly content tied to place. Below, read key takeaways and
download the report from SEO consultant Don Fick.
Measuring Performance: Real-World vs. Lab Data
- Real-World Data via CrUX: The Chrome User Experience Report (CrUX) is the best public source for real-world performance data and directly informs Google's page experience ranking factor.
- A Data Threshold to Keep in Mind: CrUX requires at least 1,000 page views in a rolling 28-day period, so newly published or in-development pages won't show up there.
- Lab Data Fills the Gap: Tools like Lighthouse provide one-off or periodic measurements that are especially useful for lower-traffic content that hasn't accumulated enough real-world data.
Core Web Vitals Remain the Foundation
- Three Metrics That Matter: LCP (loading speed), INP (responsiveness), and CLS (visual stability) are the three metrics Google uses to evaluate page experience.
- An Ongoing Discipline, Not a Checklist: Strong results depend on attention throughout design and build, not periodic checks after launch.
Supporting Metrics and Deeper Diagnostics
- Diagnosing the "Why": Load factors like TTFB and FCP, plus responsiveness factors like TBT, page weight, and request count, help explain what's behind a Core Web Vitals issue.
- Getting to Root Cause: Deeper diagnostics, including largest contentful element, JavaScript size and execution time, unused CSS, image payload, and third-party request weight, pinpoint the fix.
Choosing the Right URLs to Monitor
- Represent the Full Stack: Effective monitoring means selecting URLs that touch every layer of delivery: infrastructure, business and application logic, content, frontend, and third-party services.
- Why This Matters for Marketing: Sampling beyond the homepage ensures our monitoring reflects real conditions across the layers that affect speed and stability for readers and search engines alike.
Geocoding for AI Models: Why This Matters Now
- A New Signal from Google Maps: Google's March 2026 launch of "Ask Maps" signals growing use of geospatial data by AI systems, opening a new discoverability opportunity for publishers who geocode academic content.
- A Future-Facing Asset: Geocoding academic articles, news content, or conference proceedings creates new ways for readers, and AI models, to interact with research, positioning well-geocoded archives as a valuable asset as tools like Ask Maps mature.
The Right Approach to Geocoding
- Precision Over Volume: This isn't about tagging every place name; the goal is encoding an article's substantive geography, not over-tagging incidental references.
- Implementation Through Semantic Schema Markup: Publishers can express this through page-level JSON-LD in the HTML <head>, using Schema.org properties like spatialCoverage paired with Place, GeoCoordinates, or GeoShape.
Where This Has the Most Value
- Disciplines Where Location Is Central: This approach matters most for field sciences, public health, environmental studies, archaeology, geology, ecology, epidemiology, and regional social science research.
- A Natural Extension of Our SEO/GEO Roadmap: Geocoding builds on structured data work we've already done with ScholarlyArticle and BreadcrumbList markup, and is worth flagging to clients in the subject areas above.
For the complete analysis, including the full Core Web Vitals benchmarks, URL selection framework, and implementation guidance for geocoding scholarly content,
download the report.
