×
Homepage Performance POC
A diagnostic breakdown of basspro.com homepage performance, plus live optimization on real production assets, showing what Harper can improve from the edge.
1
Diagnose
Full Lighthouse audit of the homepage: what's slow, why, and how it impacts shoppers and search rankings.
2
Optimize
Image compression, hero preloading, script deferral, layout shift prevention: all applied automatically from the edge.
3
Zero Code Changes
Harper deploys as a proxy layer in front of the CDN. The application, backend, and deployment pipeline stay untouched.
Lighthouse Scores
Current
Projected with Harper*
*Projected scores based on the combined impact of image optimization, script deferral, preload injection, 103 Early Hints, and CLS prevention. Should be validated with a staged rollout and 28-day CrUX monitoring.
Core Web Vitals & Performance Metrics
CrUX field data (real users, 28-day rolling) and lab metrics from PageSpeed Insights, April 1, 2026.
Lighthouse Diagnostics
Every item below was flagged by Google's Lighthouse audit on basspro.com. Each issue directly impacts user experience, SEO ranking, and conversion rates.
Live Image Optimization
Original images fetched at full quality from Bass Pro's CDN, then optimized through Harper's pipeline using Sharp (libvips). Each image is converted to WebP, resized to the target viewport width, and compressed at quality 85. Click any image to enlarge.
Revenue Impact Model
Adjust the inputs to match Bass Pro's real numbers. Defaults use third-party estimates and industry benchmarks. The model uses conservative conversion-lift assumptions so Bass Pro's team can plug in their actual data and see the impact. Click any value to type an exact number.
...
Projected Annual Revenue Recovered
Your Numbers
Conv. Rate Lift
...
New Conv. Rate
...
From Conversion Lift
...
additional annual revenue
From Reduced Bounce
...
Research Behind the Model
Google / Deloitte (2020) — 0.1s faster mobile load speed increased retail conversion by 8.4%. web.dev → · Deloitte →
Portent (2022) — Sites loading in 1s convert at 2.5× the rate of sites loading in 5s. Portent →
Akamai (2017) — 100ms delay in page load reduces conversion by up to 7%. Akamai →
Google (2016) — 53% of mobile sessions are abandoned if load exceeds 3 seconds. Google Blog →
Methodology: This model uses a conservative 2.1% conversion lift per 100ms of LCP improvement — approximately 4× lower than the Google/Deloitte finding of 8.4% per 100ms. Bounce rate recovery assumes 4.5 percentage points recovered per second of LCP improvement, capped at 20pp. Revenue from recovered bounced sessions uses the improved conversion rate. These are not guarantees — they're a framework for Bass Pro's team to model the opportunity using their real numbers.
What Harper Delivers: Zero Code Changes
Edge proxy in front of the origin. Bass Pro deploys nothing.
Image Optimization
WebP/AVIF, responsive resize, DPR-aware, cached with auto-invalidation.
→ 80 to 90% image weight reduction
Hero Preloading
Inject preload + fetchpriority so LCP image loads immediately.
→ 1 to 3s LCP improvement
Layout Shift Prevention
Add width/height to every image element.
→ CLS from 0.96 to <0.3
Script Deferral
Defer analytics, replay, chat, ads until after LCP.
→ 700 ms render-blocking removed
Resource Hints
Preconnect, dns-prefetch, font-display: swap.
→ 200 to 500 ms FCP improvement
103 Early Hints
Akamai EdgeWorker emits 103 before HTML arrives.
→ Overlap TTFB with asset loading
Implementation Path
POC to production: what happens at each stage, who does what, and how long it takes.
1
Deploy Image Optimizer to Harper Fabric
Harper deploys the image optimization service to our Fabric infrastructure. This includes the Sharp-based image pipeline, the caching layer with 24-hour TTL and auto-invalidation, and the API endpoint. No action needed from Bass Pro.Timeline: 1 to 2 days. Owner: Harper.
2
Configure Akamai EdgeWorker
We write the EdgeWorker code that intercepts HTML responses and rewrites them in-flight: image URL rewriting, preload injection, width/height attributes, script deferral, and resource hints. Bass Pro's Akamai team reviews and deploys. Origin stays untouched.Timeline: 3 to 5 days. Owner: Harper writes, Bass Pro's Akamai team deploys.
3
Enable 103 Early Hints
Harper populates a page-to-asset mapping table. The EdgeWorker queries it and emits 103 Early Hints before the origin HTML is generated. Bass Pro confirms which image is the LCP element on each key page template.Timeline: 1 to 2 days after Step 2. Owner: Harper configures, Bass Pro confirms hero images.
4
Script Audit & Deferral Configuration
Harper classifies every third-party script as critical or deferrable. We provide Bass Pro's team with the full classification for review and sign-off before activating deferral.Timeline: 2 to 3 days. Owner: Harper provides classification, Bass Pro validates.
5
QA & Staged Rollout
Bass Pro runs their standard QA process. All critical flows validated. We start with a percentage-based rollout (e.g., 5% of traffic) on Akamai before scaling to 100%.Timeline: 3 to 5 days. Owner: Bass Pro QA team, Harper monitors.
6
Measure, Monitor & Iterate
PageSpeed Insights before/after comparisons and CrUX data over 28 days. Harper provides a dashboard tracking LCP, CLS, INP, TBT, and image byte savings. Ongoing optimization is continuous.Timeline: Ongoing. Owner: Harper provides reporting and recommendations.