International SEO
Structure international sites so each market ranks on its own terms — without duplicate-content and hreflang problems.
Expanding beyond Australia introduces a specific set of search problems: the same content appearing in multiple markets, the wrong market's pages ranking in the wrong country, and sites split across domains that never build enough authority to compete. None of these are exotic problems, but they compound quietly until organic performance in every market is mediocre.
Domain structure: ccTLD, subdomain or subdirectory
The first structural decision is how to house each market:
- ccTLD (example.com.au, example.co.nz) — the strongest local signal, but each domain builds its own authority from scratch. Expensive, but right for some markets.
- Subdirectory (example.com/au/) — consolidates authority on one domain, easier to manage, and the right default for most businesses expanding from Australia.
- Subdomain — usually the worst of both: weak local signal and split authority.
The right answer depends on your markets, your content operations and your competitive situation — not on habit.
ccTLD vs subdomain vs subdirectory
The structural decision everything else inherits. It is expensive to reverse, which is why it gets decided on evidence, not habit.
| ccTLD vs subdomain vs subdirectory | ccTLD | Subdomain | Subdirectory |
|---|---|---|---|
| Local trust and CTR | Strongest local signal | Weak local signal | Moderate — needs geo signals |
| Authority inheritance | None — authority starts from zero | Partial — split from the root domain | Full — builds on existing domain authority |
| Operational cost | Highest: renewals, compliance, hosting | Low | Lowest — one platform, one deploy |
| Best for | Deep single-market commitment | Rarely the right answer | Most multi-market programmes |
Hreflang, done correctly
Hreflang is how you tell search engines which version of a page belongs to which market and language. It's simple in concept and notoriously error-prone in practice: wrong codes, pages pointing at each other inconsistently, tags on pages that don't exist. Done correctly, it stops Australian pages ranking in the UK and vice versa, and lets each market's content compete on its own merits. Done incorrectly, it's a silent failure — nothing breaks, results just stay wrong.
The hreflang audit
Hreflang is the single most common international failure — implemented, broken silently, and discovered months later in the wrong market’s SERPs.
- Cluster mapping12
- page pairs verified
- market coverage complete
- Markup validation16
- return links
- x-default rules
- XML vs on-page parity
- Fallback and error handling12
- x-default set
- no redirect loops
- canonical conflicts
Geo-targeting without duplicate content
Markets that share a language (Australia, UK, US) will naturally have near-identical content. The fix isn't machine-translated "uniqueness"; it's understanding which pages genuinely need localisation — pricing, terms, products, spelling, local proof — and canonicalising or consolidating the rest honestly, rather than hoping Google sorts it out.
Local market nuances
Search behaviour, competitors and SERP features differ by market. Rankings in Australia tell you almost nothing about what will work in the US or Southeast Asia. Market-level keyword and competitor research is part of the work, not an optional extra.
What localisation actually involves
Market demand research
Keyword data per market — translated intent is not the same intent. Germans search differently than translated Americans.
Local entity signals
Local addresses, currency, units, regulatory references — the trust signals that tell engines and users the page is genuinely local.
Regional search ecosystems
Baidu, Naver, Yandex and local Google variants each index differently. The plan names which engines matter per market and why.
What this looks like
Engagements usually start with the structural question — how the site should be organised for the markets you actually plan to enter, not the ones you might — followed by implementation of hreflang and geo-targeting as engineering requirements, and a measurement setup that reports each market separately. The aim is clean expansion: each market building its own visibility without cannibalising the others.
What the engagement produces
- Domain structure recommendation with the evidence behind it
- Full hreflang implementation map, validated both ways
- Per-market keyword and intent research
- Localisation requirements per template and market
- Duplicate-content and canonicalisation rules across markets
- Measurement setup: per-market Search Console and SoV baselines
International SEO: common questions
You can, and it will underperform. Translation handles words; localisation handles intent, units, regulations, local proof and local entities. The per-market research is what tells you which pages need to exist at all in each market — often a different set than the home market.
Correct hreflang clustering plus disciplined canonicalisation: pages that are genuinely the same language and intent consolidate, pages that differ by market cross-reference. The failure mode to avoid is signalling two near-identical pages as alternates of each other.
Yes — market teams own language and local accuracy; the engagement supplies the search evidence and the technical structure. The handoff document tells each market team exactly which pages to build and what each must contain.
Tell Me What You Need to Win.
I will tell you honestly whether I am the right person for it.
Or write directly: rahulraj.sengupta@gmail.com