2.8 second delay before the first listing was visible
On a luxury real estate site where the photography is the product, almost three seconds of empty layout pushed visitors away before a single home appeared on screen.
Case study
From a failing 33 to a perfect 100 in 48 hours.
How Mountain Rose Realty, a luxury brokerage in Telluride, Colorado, rebuilt its WordPress site in Next.js and unlocked faster indexing, higher SEO scores, and AI-ready performance, with zero content loss.
- Performance score 33 to 100
- Load time 4.6x faster (2.8s to 0.6s)
- Blocking time 97% reduced
- 48-hour turnaround
By Rick Janson, JD, MBA·Founder, 10xSearch·
Executive summary
Strong visual brand. Brittle technical core.
Mountain Rose Realty, a luxury brokerage in Telluride, Colorado, ran its brand on a legacy WordPress platform with strong visual design and a brittle technical core. Listings loaded slowly, pages shifted on entry, and Google Lighthouse returned a failing 33 of 100 on Performance.
Inside 48 hours, 10xSearch delivered a pixel-perfect rebuild in Next.js, deployed it to Vercel, and pulled Performance to a perfect 100. SEO climbed from 77 to 91, Accessibility from 84 to 91. Zero content loss, same design, same pages, same URLs.
A faster, more crawlable site is the foundation every AI visibility and AEO strategy depends on. This is what Pillar 7, Technical Crawl Optimization, looks like when it ships.
Performance evidence
Google Lighthouse, desktop.
Audit captures taken at migration cut-over. Figures reflect the public homepage at time of capture and are reproducible against the live site today using the Google Lighthouse audit tool. The Performance score is calculated from Core Web Vitals measurements.
Before
WordPress
- Performance33
- Largest Contentful Paint2.8s
- Total Blocking Time670ms
- Cumulative Layout Shift0.5
- SEO77
- Accessibility84
After
Next.js on Vercel
- Performance100
- Largest Contentful Paint0.6s
- Total Blocking Time20ms
- Cumulative Layout Shift0.012
- SEO91
- Accessibility91
The challenge
What the old site was costing them.
Heavy JavaScript blocked the main thread for ~700ms
Plugin sprawl and unminified scripts froze interaction during the most expensive moment of the page lifecycle, the same moment a buyer decides whether to scroll or bounce.
Visible layout shifts pushed buttons mid-tap
A Cumulative Layout Shift of 0.5 meant the page rearranged itself after first paint, an immediate Core Web Vitals failure and a frustrating mobile experience.
A failing 33 of 100 Performance score
Search crawlers deprioritize slow pages and AI engines cite fast, stable content first. A 33 is not just a UX problem, it is a visibility problem.
48-hour execution timeline
Cloned. Deployed. Verified.
- Day 1
Cloned
Pixel-perfect rebuild of every page in Next.js. Same typography, same imagery, same URL structure. WebP image optimization and code splitting wired in from the first commit.
- Day 2
Deployed
Cut over to Vercel with static HTML pre-rendering and a global CDN. WordPress and PHP server-side rendering retired. Zero downtime, zero lost URLs.
- Hour 48
Verified
Google Lighthouse returned a perfect 100. Core Web Vitals shifted from failing to passing on every metric. Client kept full ownership of the codebase.
Outcomes
What 48 hours of focused engineering delivered.
Perfect 100 Performance score
Top 1% of websites globally on Google Lighthouse Desktop.
4.6x faster Largest Contentful Paint
From 2.8s down to 0.6s, well under the Core Web Vitals 2.5s threshold.
97% reduction in Total Blocking Time
From 670ms down to 20ms, deep into Google's green band.
41x lower Cumulative Layout Shift
From 0.5 down to 0.012, eliminating the mid-tap layout jumps.
SEO 77 to 91
Stronger crawlability foundation across meta, semantics, and structure.
Accessibility 84 to 91
Wider audience reach with cleaner contrast, semantics, and focus states.
Client owns the codebase
No hosting markup, no platform lock-in, AI-assisted updates from here on.
Zero content loss
Every page, every image, every URL preserved through the cut-over.
Proprietary insight
The 10xSearch Technical Crawl Framework.
The conceptual model 10xSearch uses to sequence engagements like this one. Four levers, applied in order:
- Image optimization via WebP and a global CDN
- JavaScript code splitting and deferral
- Static HTML pre-rendering at the edge
- Client-owned codebase on Vercel infrastructure
Note. This is an internal sequencing framework, not a public score. Results vary by baseline, template complexity, and third-party dependencies.
Why a perfect 100 matters for AI search
Fast pages get cited. Slow pages get filtered.
AI answer engines favor fast, stable, structured pages when deciding what to cite.
Core Web Vitals are a direct ranking input for Google and a strong proxy for crawl priority across every modern engine.
Early adopters who fix the technical foundation gain compounding visibility advantages over competitors stuck on legacy stacks.
Frequently asked
About this case study.
What was the actual technical work in 48 hours?
A pixel-perfect rebuild of the existing WordPress site in Next.js, deployed to Vercel with static HTML pre-rendering, image optimization to WebP, JavaScript code splitting, and a global CDN. The visual design, content, and URL structure were preserved exactly.
Are the Lighthouse numbers reproducible?
Yes. The before and after captures were taken on the public homepage at migration cut-over using Google Lighthouse Desktop. Anyone can run the same audit against the live mountainrose-realty domain today and confirm the post-migration score.
Why does a perfect Lighthouse score matter for AI search?
AI answer engines favor fast, stable, structured pages when deciding which sources to cite. Core Web Vitals are a direct ranking input for Google and a strong proxy for crawl priority across every modern engine. Fixing the technical foundation is what unlocks every other AEO investment.
Where does this fit in the 10 Pillars of Search?
This engagement was primarily Pillar 7, Technical Crawl Optimization. The other nine pillars still matter for long-term visibility, but a failing performance score would have capped the upside on every one of them.
Did Mountain Rose lose any content or URLs?
No. The rebuild was pixel-perfect with zero content loss and full URL preservation. Search equity carried over, no redirect chains were introduced, and the brand experience stayed identical from the visitor's point of view.
Want these numbers for your site?
Free scan. Instant quote. Delivered fast.
Book a 30-minute call and we will run Lighthouse against your homepage live, walk through what is dragging your score down, and tell you exactly what a 48-hour rebuild would look like for your stack.