Markets · Vadodara
International SEO
Market selection, localisation and multi-region architecture — with hreflang only when equivalent pages actually exist.
Fit
Who this is for
A fit
- Businesses entering another country or language with evidence of demand, a way to fulfil, and a willingness to localise — not only to translate.
- Teams choosing ccTLD vs subdomain vs subfolder, and needing that choice tied to operations, not to a blog post about “best practice”.
- Sites that already have more than one locale and are hurting themselves with incorrect hreflang, mixed languages, or copied templates.
Not a fit
- Hreflang theatre: tagging every URL in five languages when four of the languages are auto-translated shells.
- A rollout of twenty country homepages that are the same English page with a flag. That is not a market.
- Anyone who wants “global SEO” as a slogan without naming the first market, the currency, or who will write the local pages.
Problems
What I diagnose
A URL copy is treated as a market. Country folders go live before pricing, shipping, legal pages or local intent exist. Google is asked to rank a page the business cannot actually serve.
Hreflang on non-equivalents. Annotations pointing at pages that are not substitutes: different products, different language quality, or a homepage paired with a category. Hreflang then becomes a source of wrong-country results rather than a fix.
One HTML language for everything. html lang, content language, and the visible copy disagree. Users and crawlers get mixed signals; hreflang cannot paper over that.
ccTLD vs folder chosen for folklore. The hosting map, the CMS, the legal entity and the link profile all constrain the choice. Picking a structure because “Google prefers it” without those facts is how you migrate twice.
Translated thinness. Machine output on commercial templates, untranslated snippets, and local SERPs that want proof the market page never gathered. Localisation is research and content, not a locale switcher.
Delivered
What you actually get
Market-entry memo
Which country or language is actually next, what demand evidence exists, what you can fulfil, and what would have to be true before we build pages. Markets you should not enter yet are named.
URL and language model
ccTLD, subdomain or subfolder, and why, given your CMS, legal setup and existing equity. Includes how language and country combine when they are not the same thing (en-IN vs en-GB, for example).
Equivalence map
Which pages are true alternates. This is the prerequisite for hreflang. Pages without an equivalent stay out of the annotation set.
Hreflang spec — only if earned
Implementation notes (HTML, headers, or sitemap), x-default policy, and QA checks. If equivalents are not there, the spec is “do not ship hreflang yet”, which is a deliverable.
Localisation checklist for commercial templates
What must be rewritten vs translated: titles, proof, pricing, schema, internal links, and the queries that page is allowed to own in that market.
Hreflang annotates equivalents. It does not create a market.
Method
How the work runs
Demand and fulfilment
Search demand is not enough. Can you sell there, support there, and keep the page true? If not, we wait. International SEO is a market-entry decision first.
Structure
Choose the URL model against operations and equity, not against a generic ranking myth. Document the cost of reversing it.
Build equivalents
Localise the commercial templates that have to exist. Language, proof and offers — not a copied nav with a translated footer.
Annotate
Hreflang only across the equivalence map. QA for return tags, wrong codes, and pages that drifted out of being substitutes.
Monitor by market
Coverage and queries per locale, country-wrong landings, and whether x-default is doing what we said. Expand only when the first market is real.
Proof
Examples in progress
There is no published international engagement on this site. A useful example would show the market we chose, the URL model, the equivalence map, and whether hreflang was implemented or deliberately withheld.
This is one page on purpose. There will not be a cluster of hreflang tutorials pretending to be extra service owners. When a market guide exists, it will live as a resource and point here.
Questions
Common questions
Do I need hreflang?
Only if you have equivalent pages in more than one language or country and you need to tell Google which one is for whom. If you have a single English site serving whoever arrives, you do not need hreflang. Adding it without equivalents is how international SEO turns into a debugging hobby.
Subfolder, subdomain, or ccTLD?
It depends on legal presence, CMS, hosting, and how much equity you cannot afford to split. I will not answer with a universal winner. The memo names the option that matches how you operate, and the migration cost if you outgrow it.
Can you just translate the site?
Translation without localisation leaves currency, examples, proof and intent wrong. I will specify what has to be rewritten. I do not run a translation desk; you will need writers or a vendor for the language work.
We want twenty countries at launch.
Then we will argue about the first one. A staggered entry you can staff beats a grid of empty locales. If a stakeholder needs twenty flags on a slide, that is not this engagement.
Name the first market, not the flag grid
Send the site, the country or language you actually want next, and how you fulfil there. I will tell you whether this is international architecture, consulting, or too early.