International search work is mostly a set of architecture decisions made early and lived with for years: whether each market gets its own domain, a subdomain or a folder, how language versions are declared to search engines, and what happens to a visitor who arrives from the wrong country. Get those wrong and no amount of content fixes them cheaply. An international SEO consultant is hired to make those calls, and the way to test one is to ask how they would make them and why. This page covers what the job contains, what Google's own documentation says, and how to vet the person before a rollout commits you.
The architecture decision comes first
Google documents three workable URL structures for locale-specific content and is candid about the trade-offs of each. A country-code top level domain such as a national extension gives the clearest geotargeting signal and makes server location irrelevant, but it is expensive, it targets a single country, and each one has to be built up separately. A subdomain on a generic domain is easy to set up and allows different server locations, but users may find the targeting ambiguous. A subdirectory on the main domain is the lowest maintenance option and keeps everything on one host, though it makes separating sites harder and ties you to one server location. Google specifically advises against segmenting locales with URL parameters, calling that approach not recommended. A consultant who leads with a recommendation before asking about your legal entities, market priorities, existing domain strength and engineering capacity is guessing. One who walks you through those trade-offs against your own constraints is doing the job you are paying for.
Language versions, hreflang and the signals Google reads
Where the same content exists for several languages or regions, Google's guidance is to use different URLs for each version rather than swapping content on one URL, and to annotate them with hreflang so a searcher is offered the right variant. That matters most when versions are similar or duplicated across URLs, which is exactly the situation for regional variants of one language. Google says it identifies target locales through several signals, with country-code domains the strongest, alongside hreflang annotations, server location and contextual clues such as local addresses, phone numbers and currency, and it notes that it ignores locational meta tags. Two warnings in the same documentation are worth writing into a brief. Google advises against automatic language redirects, because they can stop users and search engines from seeing all the versions, and it says relying on IP analysis to detect a visitor's location is difficult and generally not reliable. A consultant proposing an IP-based redirect as the centrepiece of an international rollout is proposing the thing Google's own documentation tells you not to do.
Vetting the consultant
Google's hiring guidance applies here and is worth using verbatim. Ask for examples of previous work and the outcomes attached to them, ideally rollouts across more than one market rather than a single translated site. Ask what results they expect and in what timeframe. Ask them to explain their recommendations clearly, because Google warns against consultants who will not explain their methods, and be sceptical of unsolicited approaches and claims of a special relationship with Google. Any guaranteed ranking should end the discussion, since Google states that no one can guarantee a number one ranking. Add the international-specific probes: who writes the local-language content and are they native speakers writing rather than editing machine output, how will hreflang be generated and validated as pages change, who owns the decision when the local sales team wants a page that conflicts with the global template, and what is the plan if a market underperforms and has to be wound down. Grant read-only Search Console access during an audit.
Scope, sequencing and what to measure
The expensive mistake in international programmes is launching every market at once. A staged rollout, with one market proving the template and the workflow before the next begins, converts an architecture bet into a series of smaller ones. Sequence by revenue opportunity and by whether you can actually service the market, not by translation cost. Measure each market separately, because a global traffic chart hides the market that is failing behind the one that is working, and set up separate reporting for the localised properties from the start. Expect slow initial movement: Google's own guidance says some changes take effect within hours while others can take several months, and a new market has no existing footprint to build on. Judge the first market on whether the architecture, the annotation and the content workflow hold up under change, because those are what every subsequent market will inherit.
Questions people ask about international seo consultant
Country domains, subdomains or subfolders?
Google documents all three as valid. Country-code domains give the clearest geotargeting but are expensive and single-country; subdomains are easy to set up but can look ambiguous to users; subfolders are the lowest maintenance but harder to separate later. Google advises against using URL parameters for locales.
Do I need hreflang?
If the same content exists in more than one language or regional version, Google's guidance is to keep each version on its own URL and annotate them with hreflang so searchers are pointed to the right one. It matters most where the versions are similar or duplicated across URLs.
Should the site redirect visitors based on their country?
Google advises against automatic language redirects because they can stop users and search engines from seeing every version, and it says IP-based location detection is difficult and generally not reliable. A visible suggestion the visitor can decline is the safer pattern.
How many markets should we launch at once?
One, then the next. A staged rollout proves the architecture, the hreflang workflow and the content process before you multiply them. Measure each market separately, and expect movement over months rather than weeks, since a new market starts with no existing footprint.