Case Study: Rebuilding Site Architecture to Boost Organic Impressions by 250%
How migrating off a slow monolithic setup onto an edge-delivered hub-and-spoke architecture improved crawl efficiency, Core Web Vitals, and organic impressions, with a playbook you can reuse.
Legacy monolithic hosting and messy URL hierarchies quietly cap organic growth. Pages load late, crawlers waste budget on junk paths, and blog posts compete with each other instead of reinforcing a topic. This case study breaks down a rebuild onto a modern edge stack with hub-and-spoke information architecture, and what moved in Google Search Console afterward.
The pattern matches how we approach performance-minded website work: fix the substrate, then make the content graph make sense.
Executive Snapshot
| Dimension | Before | After |
|---|---|---|
| Delivery | Origin-heavy content management system (CMS), slow Time to First Byte (TTFB) for distant users | Edge-cached HTML on Cloudflare Pages-class hosting |
| Information architecture | Flat blog sprawl, weak parent/child relationships | Topic hubs with spoke guides and deliberate internal links |
| Structured data | Partial or missing templates | Automated Article / BlogPosting, FAQPage, Organization |
| Organic outcomes | Stalled impressions, crawl noise | ~250% impression lift; ~3× inbound leads from organic |
flowchart LR
Legacy["Legacy stack<br/>slow TTFB · broken equity · CMS bloat"] --> Rebuild["Edge hub architecture"]
Rebuild --> Imp["+250% search impressions"]
Rebuild --> Leads["~3× inbound leads"]1. The Challenge
The legacy site had the usual compound fracture: slow server responses, unoptimized dynamic URLs, broken internal link equity after years of CMS churn, and front-end scripts that crushed mobile Core Web Vitals. Content existed, but it was hard to crawl cleanly and hard to trust as a coherent topical graph.
- Speed failures: Time to First Byte (TTFB) and Largest Contentful Paint (LCP) failed on key landing templates, especially on mobile.
- URL chaos: Multiple variants competed (parameters, leftover staging paths, inconsistent trailing slashes).
- Topic sprawl: Blog posts targeted overlapping keywords without a parent hub.
- Weak entities: Schema was inconsistent or absent, so rich results and entity clarity suffered.
- Noisy sitemaps: Included non-indexable or redirected URLs, confusing discovery.
2. Solution Architecture
Stack migration
We re-platformed marketing delivery onto a modern JS framework with static/edge HTML on Cloudflare Pages (the same class of approach outlined in monolithic CMS vs serverless edge). The point was not novelty. It was immutable assets at the edge, predictable TTFB, and fewer moving parts on the public origin.
Hub-and-spoke redesign
Sprawling posts became structured clusters: one cornerstone per topic, supporting spokes with prerequisite/next links, and hub landers that read like a curriculum. Parent pages earned the head terms. Spokes earned long-tail and supporting intent. Internal links stopped being decorative footers and became the ranking graph.
flowchart TD
Hub["Topic hub / cornerstone"] --> S1["Spoke: foundations"]
Hub --> S2["Spoke: technical"]
Hub --> S3["Spoke: applied / case study"]
S1 --> Hub
S2 --> Hub
S3 --> Hub
S1 --> S2
S2 --> S3Semantic schema automation
Templates emitted Article/BlogPosting, FAQPage, BreadcrumbList, and Organization JSON-LD (JavaScript Object Notation for Linked Data) from the same source of truth as the UI. No hand-pasted blocks drifting out of date. That same discipline supports classic rich results and cleaner entity extraction for AI answers.
Migration hygiene
- Full URL inventory: From crawl + analytics + Search Console.
- 301 map with intent: One-to-one for keepers; intentional consolidation where thin URLs merged into stronger parents.
- Parity checks: Titles, H1s, primary content, canonicals, and indexability.
- Clean sitemaps on day one: Old sitemap URLs retired.
- Search Console watch: Daily for two weeks, then weekly.
3. Business Impact
Once crawl and rendering stopped fighting the content, impressions scaled on the topics the hubs were built to own. Traffic quality mattered as much as volume: inbound leads roughly tripled as landing pages aligned with commercial intent and loaded fast enough to convert.
| Metric | Directional outcome | Primary drivers |
|---|---|---|
| Organic impressions | ~+250% | Indexable URL clarity, hubs, internal links, refreshed templates |
| Inbound leads (organic) | ~3× | Faster Largest Contentful Paint (LCP) and Interaction to Next Paint (INP), clearer calls to action (CTAs), intent-matched spokes |
| Core Web Vitals | Money templates into “good” field ranges | Edge HTML, less main-thread JavaScript, image discipline |
| Crawl efficiency | Fewer soft-404 / duplicate wastes | Canonicals, sitemap hygiene, retired thin URLs |
Treat the percentages as a case outcome under that site’s competition and baseline, not a guarantee. Architecture removes ceilings. Content and authority still have to earn the clicks.
4. What Actually Moved the Needle
- Fewer, stronger URLs: Consolidation beat publishing more mediocre posts into a broken graph.
- Hub clarity: Searchers and crawlers could tell which page owned which topic.
- Speed as trust: Better vitals helped engagement and gave technical SEO room to work.
- Measurement loop: Google Search Console (GSC) plus crawls caught redirect mistakes early instead of months later.
For the operating cadence behind those checks, see the modern SEO tool stack. For originality standards on the new hubs, use the information gain framework.
5. Reusable Playbook
- Benchmark: Google Search Console for 16 weeks, top landing pages, Core Web Vitals, crawl errors.
- Design the target IA: Hubs, spokes, redirects, and kills.
- Build edge-fast templates: With schema baked in.
- Migrate carefully: Tested 301 map and parity QA.
- Watch coverage: Submit sitemaps, monitor, fix fast.
- Then scale content: Only after the new graph is healthy.
If your organic channel feels “stuck” despite steady publishing, suspect architecture before you blame the algorithm. Start from the Complete Guide to SEO, review selected work, or talk through a rebuild.
Frequently asked questions
Was the 250% impression lift only from hosting changes?
No. Edge delivery fixed crawl and vitals foundations, but hub-and-spoke information architecture (IA), internal links, schema, and content consolidation did equal or greater work. Hosting alone rarely produces durable organic gains.
How long after migration did results show?
Indexation and crawl patterns moved within weeks. Material impression growth compounded over subsequent months as Google re-evaluated the cleaner URL graph and updated hubs. Plan for a transition window, not overnight magic.
Did rankings drop during the migration?
A careful 301 map, parity checklist, and Google Search Console monitoring kept the dip short. Migrations without redirect discipline are where traffic disappears. Treat redirects as a first-class deliverable.
Can a small marketing site use the same playbook?
Yes at smaller scale: faster templates, clearer hubs, honest schema, and fewer thin URLs. You do not need a thousand stock keeping units (SKUs) to benefit from architecture hygiene.
What would you do differently next time?
Freeze non-essential content launches during cutover, automate more parity checks, and instrument impression and conversion dashboards before week one so causality is clearer for stakeholders.