How shorter.gg raised its structural health from 67 to 83 while tripling its content
Most sites that triple their page count in two weeks watch their internal structure fall apart: authority pools on the homepage, new pages land as orphans, and topic clusters splinter into islands. shorter.gg did the opposite. Across three real RankForge crawls between June 26 and July 9, 2026, the site grew from 115 to 164 pages and from 470 to 1,517 internal links — and its structural health score climbed 67 → 78 → 83. This is the unedited story of how, including the messy middle where the audit caught a regression, and the modules that got worse. All numbers are from the three real reports. Published with the site owner's consent.
Run the Internal Link Checker on your site — free, no account.
The site, and why this case study is different
shorter.gg is a URL shortener with a link-in-bio and QR-code product — and a genuine content ambition. Over two weeks its team ran a serious content and internal-linking build-out: whole new hub sections (alternatives, comparisons, glossary, use-cases, industries, integrations, how-to guides), not just a handful of blog posts. The interesting question for structural SEO is not “did the score go up.” It's the harder one: can you 3× your content without wrecking the internal structure that carries authority to it? We have three real crawls that answer it.
What this is — and isn't: This is a before/after of the structural health score — internal crawl paths, authority flow, clusters, pillars. It is not a rankings or traffic claim. shorter.gg's Search Console data appears in the reports, but we make no causal claim that structure moved it. The point is narrower and more honest: as the site scaled, did its internal structure hold up — and did the audit describe reality at each step?
Crawl one (67): a scorecard that pointed straight at the problem
The overall health score is never a standalone number — it is the weighted composite of the module scores (that is exactly why fixing one module can move the headline). On June 26 the composite was 67, verdict “moderate structural issues.” Reading the scorecard, three modules were clearly dragging it down while the rest were fine.
MODULE SCORE (/100) ────────────────────── ──────────── Authority Distribution 46 <- "fragile authority flow" Pillar Strength 49 <- one pillar at strength 10 Topical Clusters 56 <- two isolated islands Anchor Gaps 64 Page Speed 69 Internal Link Strategy 73 On-Page SEO 91 Content Visibility 96 GEO Readiness 100 ────────────────────── ──────────── OVERALL HEALTH 67
That is the discipline the tool enforces: don't average, find the floor. On-Page SEO at 91 needed nothing. The path from 67 upward ran through the three low modules — and, usefully, they were the same problem seen from three angles.
Diagnosis 1 — the authority cliff
Internal authority flows along links like an internal PageRank: every page hands a share of its authority to the pages it links to. Plot average internal authority by crawl depth and a healthy site decays gently. shorter.gg fell off a cliff the moment you went one click past the top level.
DEPTH AVG INTERNAL AUTHORITY ───────── ────────────────────────────── Depth 0 92 ██████████████████ homepage Depth 1 93 ██████████████████ top-level hubs (7 pages) Depth 2 13 ██▌ <- the cliff (19 pages)
The knock-on effects were all in the report: depth decay flagged as “sharp,” the homepage hoarding a disproportionate share, and the editorial (in-body) link ratio at just 17.9% — the rest were boilerplate nav and footer links that carry weaker signals. Authority was pooling at the top and never reaching the specific pages that needed it.
Diagnosis 2 — a blog post cosplaying as a pillar
A pillar page is the hub of a topic cluster — the page every supporting article links up to. RankForge picks each cluster's pillar as its strongest page, and on June 26 one cluster's “pillar” was a single blog post, how-to-optimize-and-customize-shortened-urls. Its pillar-strength score was 10 out of 100 — “structurally fragmented.” The report's reason line said it plainly: “Only 0 of 8 related pages link back to the pillar.”
This is a real weakness, not a false alarm: A cluster whose hub receives zero links from its own supporting pages has no hub at all — search engines get no signal that the topic is organized around anything. Average pillar strength across the site was 49, and it was the audit's #1 priority fix.
Diagnosis 3 — two islands, no bridges
Topical clusters earn authority partly by connecting to each other — cross-cluster “bridge” links tell search engines your topics are related facets of one expertise, not disconnected silos. On June 26 shorter.gg had exactly two clusters, both flagged “isolated,” with a single bridge between them. Topical score: 50. Cluster health: 57.
The fix: they didn't just “write more” — they built structure
Here's the part most content pushes get wrong. The lazy version of “fix your clusters and pillars” is to publish a pile of blog posts and hope. shorter.gg instead added structured hub sections — /alternatives, /compare, /use-cases, /glossary, /industries, /integrations, /how-to — each a real hub with supporting pages linking up to it. Pages went 115 → 130 → 164; internal links went 470 → 861 → 1,517. Crucially, the new links were routed to create hubs and bridges, not just dumped.
You can see the placement working in the module scores by July 9. Pillar Strength climbed to 80 with four real pillars (the qr-codes hub at 82, integrations hub a “dominant” 97). Internal Link Strategy hit 98. And Authority Distribution — the original 46 — reached 79, with the homepage's hoarded share of authority falling from 9.3% to 4.8% and the typical page's authority rising as equity finally spread outward.
The messy middle: the audit caught a regression mid-push
This is the crawl most case studies would quietly delete. On July 8, mid-build, the overall score had risen to 78 — but the audit lit up with two brand-new problems the content push had introduced. Nine orphan pages had appeared (pages with zero internal links pointing to them), and the sitemap had ballooned to 114 URLs with only 56% crawl coverage — because 50 of those entries were user-generated short-link slugs (/lcmucH, /zXlwYJ …) that aren't site content at all.
This is the actual value of re-crawling: A content sprint always creates temporary structural debt — new pages land before their inbound links do. The win isn't avoiding that entirely; it's catching it in the same cycle. By the next crawl, orphans were back to 0 and sitemap coverage was back to 100% (103 clean URLs, no gaps).
Crawl three (83): what actually moved
By July 9 the three constraints from crawl one were resolved or well on their way. Here is the full module scorecard, start to finish — including the two modules that went down, which we are showing on purpose.
MODULE Jun 26 Jul 09 Δ ────────────────────── ────── ────── ──── Authority Distribution 46 79 +33 Pillar Strength 49 80 +31 Internal Link Strategy 73 98 +25 Topical Clusters 56 71 +15 Page Speed 69 76 +7 Content Visibility 96 97 +1 GEO Readiness 100 100 0 On-Page SEO 91 88 -3 Anchor Gaps 64 57 -7 ────────────────────── ────── ────── ──── OVERALL HEALTH 67 83 +16
Two more quiet wins rounded it out. shorter.gg added an llms.txt file (GEO readiness went from “Not found” to “Found · 9 links”), and the JavaScript-rendering warning from the July 8 crawl — three pages that rendered empty for search engines — was gone by July 9, with the crawler confirming “results are complete.” Sitemap coverage returned to 100%.
The honest ledger: what didn't improve
A case study that only shows the numbers that went up isn't a case study, it's an ad. Three things got worse or stayed unsolved, and each is instructive.
- chevron_rightEditorial link ratio fell. In-body editorial links were 17.9% in June and only 10.3% by July 9. The absolute count of editorial links rose, but the programmatic hub sections added so many navigational links that the ratio dropped. The audit still flags this as “only 10% editorial” — a real, open opportunity, not a solved problem.
- chevron_rightTopical Clusters dipped mid-build (module 56 → 87 → 71; topical score 50 → 74 → 65). Splitting two clusters into four spreads the average health and gives each new cluster fewer bridges to start — a reweighting discontinuity you should expect when you add content, not a regression to panic over.
- chevron_rightNot everything is fixed. One weak pillar remains (a blog post at strength 57), one strategic page still sits below median authority, and Anchor Gaps actually slipped to 57. The site is in good structural shape — it is not perfect, and the report doesn't pretend otherwise.
Read this honestly: A structural health score measures internal structure, not rankings or traffic. Structure removes a ceiling on what content can achieve; it does not manufacture demand. What these three crawls demonstrate is specific and useful: a site can scale content aggressively and keep its structure intact — if it routes authority into hubs and bridges deliberately, and re-crawls to catch the debt each push creates.
The play, generalized
- Read the module scorecard and find the floor — the single lowest module — not the average. shorter.gg's floor was Authority Distribution at 46.
- Check average authority by depth for a cliff, and whether your “pillars” actually receive links from their own supporting pages.
- Scale content as structure, not volume: build hub sections with supporting pages that link up, and add cross-cluster bridges — don't just publish posts.
- Expect temporary debt. Every push creates orphans and sitemap noise; the goal is to catch and clear it in the same cycle.
- Re-crawl and confirm the low modules recovered and the composite moved with them — and look honestly at what didn't.
FAQ
Are these real numbers from shorter.gg?expand_more
Yes. Every figure — 67 → 78 → 83 overall, 46 → 79 on Authority Distribution, 470 → 1,517 internal links, the orphan spike, the cluster counts — comes from three real RankForge crawls of shorter.gg dated June 26, July 8, and July 9, 2026, published with the site owner's consent. Nothing is fabricated or illustrative.
Does a structural health score of 83 mean better rankings or traffic?expand_more
No. The structural health score measures internal structure — crawl paths, authority flow, clusters, pillars — not rankings or traffic. Structural fixes remove a ceiling on what a page can achieve; they don't by themselves move rankings. We frame this strictly as a structural before/after.
How can a site triple its content without its structure collapsing?expand_more
By adding content as structure rather than volume: build hub sections whose supporting pages link up to a clear pillar, add cross-cluster bridge links, and route new internal links toward the deep pages that need authority. shorter.gg added whole hub sections (alternatives, comparisons, glossary, use-cases, integrations, how-to) instead of a loose pile of blog posts.
Why did the July 8 crawl look worse in places even though the score went up?expand_more
A fast content push always creates temporary structural debt — new pages go live before their inbound links do. The July 8 crawl caught 9 new orphan pages and a sitemap that had filled with user short-link slugs. Because the audit surfaced both immediately, they were fixed by the July 9 crawl (orphans back to 0, sitemap coverage back to 100%).
Why did the editorial link ratio go down while the score went up?expand_more
The absolute number of in-body editorial links rose, but the new programmatic hub sections added even more navigational links, so the ratio fell from 17.9% to 10.3%. The audit still flags this as an open opportunity. It's a good example of a real metric that didn't improve even as overall structure did — which is why we show it.
Which module drove most of the gain?expand_more
Authority Distribution, which rose from 46 to 79. It was the audit's named primary constraint on June 26 (“fragile authority flow”), and fixing the depth-2 authority cliff — by routing links into deep pages and hubs — lifted it the most. Pillar Strength (49 → 80) and Internal Link Strategy (73 → 98) were close behind.
Put this into practice
Run the Internal Link Checker to see this on your own site, or run the full structural audit for the complete picture — both free, no account required.