After ten years of international SEO work, the most common pattern I see isn't teams doing the wrong things. It's teams doing the right things in the wrong order: building strategies on foundations that were never properly laid.
The post-mortems look similar every time. Hreflang is blamed. Content quality comes up. Local authority gets flagged.
These aren't wrong diagnoses, but they're symptoms of the real issue. The conditions for success were never established before the strategy was built. Fix the hreflang and you'll probably still be stuck, because the problem started earlier, in decisions that felt like prerequisites but were treated as afterthoughts.
International SEO doesn't fail at execution. It fails upstream.
Step zero is what happens upstream: the prerequisites that determine whether any strategy built on top of them has a chance of working. It's also the phase nobody budgets for.
Organic isn't a reason to enter a market. It's a test of whether you can win there.
The business case for international expansion rarely starts with SEO. Revenue targets, competitive pressure, board decisions: these drive the direction. That's fine. SEO's role isn't to validate those decisions. It's to honestly assess whether organic can support them, and under what conditions.
The problem is timing. Market research often happens before the SEO conversation begins, but by the time the SEO team is briefed, the decision to enter a market has usually already been made.
What follows isn't an honest assessment of organic viability - it's a case built around a conclusion that was reached elsewhere. That's a different exercise, and it produces different results.
Before any strategy conversation begins, I look at three things:
The first is whether demand fit actually exists: does measurable organic search behaviour exist for your offer category in this market, and at what funnel stage? Volume alone won't answer this.
A market can generate search traffic that never converts because the intent doesn't match the product, because the buying journey happens elsewhere, or because the category simply doesn't exist locally in the way it does at home.
I've worked with clients who had solid keyword volume in a target market - enough to justify investment on paper - and found, once we dug into the SERPs and local consumer behaviour, that the searches were informational with no clear commercial pathway. Volume existed. Demand fit didn't.The second is competitive viability: what does the current SERP landscape tell you about the cost of organic entry? Every market worth targeting is competitive. The real question is whether organic is a viable path given your resources, your timeline, and the specific players already established there.
Some markets have local players with structural advantages - domain authority built over a decade, deep local editorial relationships, brand recognition that no content strategy will displace in twelve to eighteen months. When that's what the data shows, saying so clearly and early is part of the job.The third variable is the one that's changed most in the last few years: what does "ranking well" actually mean in this market's search environment? AI search doesn't roll out uniformly or get adopted uniformly. Features vary by language, by geography, and by query type, with meaningful implications for how organic visibility translates into traffic and conversions.
France, at the time of writing, is excluded from AI Mode entirely, while Germany, the UK and the US are at different stages of AI Overview integration.
A realistic assessment of organic potential has to account for what the actual search experience looks like in your target market, not just what ranking position number three theoretically delivers.
These three variables don't produce a simple yes or no. They produce a calibrated view of what organic can realistically contribute, where the gaps are, and what the team needs to agree on at step zero - before strategy is built rather than after it's launched.
The definition of visibility is now market-dependent. A #3 ranking in one country may still drive meaningful traffic, while the same position elsewhere could sit beneath AI Overviews, shopping modules, video carousels, and local packs that absorb most clicks.
Advanced Web Ranking helps international SEO teams monitor how SERP composition differs by country and query type, including tracking AI search visibility alongside classic rankings. That distinction matters when forecasting organic contribution market by market.
Try AWR free to compare real SERP environments across your international markets.
Knowing the language isn't the same as knowing the market
The most expensive prerequisite gap I encounter isn't technical. It's the assumption that linguistic competence equals market understanding, and the strategic decisions that follow from it.
It doesn't. Language gives you access to the market's vocabulary. It doesn't give you access to the market's behaviour, its trust signals, or the specific reasons someone in that context would choose you over a local alternative.
One French furniture retailer I worked with had done everything right on the linguistic side. Their Dutch-language site was accurate, well-structured, and keyword-optimised for the Netherlands market. Traffic was coming in. Conversions weren't. When we looked at what was actually being featured and promoted on the Dutch homepage, the problem became obvious: the hero products were the same large sectionals and oversized storage units that performed well in France. Dutch apartments are, on average, significantly smaller than French ones. The products weren't wrong - they were just wrong for how people in the Netherlands actually live. No keyword tool surfaces that. Understanding the local housing market does.
This is what I call Digital Expatriation: immersing yourself in a market's information environment - its SERPs, its content formats, its consumer context - before opening a keyword tool or writing a single brief. The goal is to become a temporary resident of that market's world, not a translator of your own.
What gets missed without it, follows consistent patterns across the markets I've worked in.
Content that passes linguistic review but fails behavioural expectations. Veruska Anconitano has documented specific examples across European markets:
German readers expect two to three citations per article as a baseline credibility signal.
Brazilian audiences default to short video and local case studies.
Japanese readers weigh a mention in a local trade publication more heavily than international domain authority.
These aren't stylistic preferences. They're trust architecture - and getting them wrong produces content that is correct but not credible.
Intent also shifts across markets in ways that translated keyword lists won't reveal. The same search pattern, the same surface-level query, can carry very different intent depending on where it originates. What someone in Germany means when they search for a product category isn't always what someone in the UK means, even when the words translate cleanly.
The divergence only becomes visible when you study local SERPs and local content. This is why I always recommend spending time inside a market's search environment before doing any keyword research - not to find keywords, but to understand what the market is actually trying to do.
Native speakers help. But native speakers and native in-market expertise are not the same thing. The latter requires understanding local buying behaviour, local competitive dynamics, and local content formats, not just local vocabulary.
At step zero, this distinction is the difference between entering a market and approximating one.
Studying local SERPs manually is valuable. Doing it consistently across multiple markets is difficult without reliable tracking infrastructure.
Advanced Web Ranking allows SEO teams to monitor localized search results at scale, helping surface differences in intent, SERP layouts, and feature prevalence between regions.
Try AWR free to explore how search intent and SERP structure vary across markets.
The decisions that break international SEO were made before the strategy started
These aren't configuration tasks. They're architectural decisions: the kind that are expensive to undo, invisible until they cause problems, and almost never resolved before the strategy begins. They belong in step zero. They almost never are.
Each of the three below has a decision at its centre - one that can be made clearly before the strategy begins, or left unclear and inherited as a problem mid-execution.
Decision 1 - Language, region, and authority are three separate signals
They're almost always treated as one, which creates compounding problems that little subsequent optimisation can fix.
Language describes what the content is.
Region describes who it's for.
Authority describes whether a market-specific version should rank.
A strategy that conflates them - building a single "French content" approach without distinguishing between France, Belgium, Switzerland, and Canada - produces a structure that sends confusing signals and can't be cleanly untangled later.
Hreflang sits inside this decision, not above it. It's a hint to Google, not a directive. It helps resolve which version to serve to which audience, but it does not build local relevance and does not transfer authority between versions. Teams that treat hreflang implementation as the end of their international SEO work have, in practice, skipped the harder part.
The decision that needs to be made before any architecture is built is which signal serves which purpose - for each market on your list. Once that's documented, hreflang can do what it's actually designed to do: resolve ambiguity within an already-coherent structure.
-> The clearest sign this decision was never made: local market teams are getting traffic from the wrong country versions, and explaining why requires a whiteboard and forty-five minutes.
Decision 2 - SEO must sit inside the localisation workflow, not downstream of it
Translation processes built without SEO input actively undo optimisation work. The systems regularly overwrite title tags, flatten internal linking structures, and erase metadata that was carefully optimised in the source version. The localisation team isn't making mistakes - they're following a process that was never designed with SEO in mind.
Localisation that actually carries SEO value into a new market includes product placement decisions, internal linking architecture, and offer framing adapted to local intent. Everything else is translation, which is a different and more limited thing.
I've worked with clients who had published hundreds of localised pages using templated structures - market-specific landing pages, local branch or agency pages - that looked complete from the outside. Technically indexed, correctly hreflang'd, broadly on-topic. But the SEO work that had gone into the source version simply hadn't survived the localisation process. Title tags had been overwritten. Internal linking had been flattened back to a generic structure. The pages ranked briefly, and then didn't. The localisation team had done their job. Nobody had built a process where both jobs could happen at once.
The decision to make before any localisation work begins is: who owns SEO requirements within the localisation process, and what does sign-off look like? Without a named owner and a defined handover point, the default - in every system I've seen - is that the localisation workflow wins.
-> If the SEO team's first month after a market launch is spent rebuilding what already existed in the source version, that decision was never made.
Decision 3 - Governance: what you can actually influence
This one is worth being honest about. In most multi-market projects I've been brought into, governance isn't a decision the SEO team gets to make. The structure was decided before the project began, inherited from a previous market expansion, or locked in by organisational constraints that no consultant is going to redesign.
The decision available to you isn't whether to redesign governance. It's whether to make the existing governance explicit enough to be useful. Centralised strategy with decentralised execution tends to work when three things are documented:
who can approve local content decisions without escalation,
who owns the relationship with in-market contributors,
and what happens when local and global priorities conflict.
You don't need a governance overhaul to establish those three things. You need a conversation and a document.
When those aren't clear, performance problems become very difficult to diagnose, because the conflict producing them is structural, not visible. Global teams push for consistency. Local teams push for relevance. Both are right. Neither can see the other's constraints. The result is stalled performance and a lot of unproductive internal meetings.
-> A useful diagnostic: if a local market team made a content decision tomorrow that contradicted global strategy, would anyone know? Would there be a clear process for resolving it? If the answer to either is no, that's where the governance conversation starts - not with an org chart, but with those two questions.
Authority doesn't travel. It has to be earned in each market.
A strong domain does not automatically carry its authority across markets and language versions. This is one of the most persistent assumptions in international SEO - and it's understandable. Google does use domain-level signals, and the logic that a trusted domain should transfer some of that trust across its own subfolders and subdomains is not absurd. The problem is what doesn't transfer: local relevance, local recognition, and local editorial relationships.
Authority built in one market doesn't move to another the way a bank balance does. It's closer to a professional reputation - legible in the context where it was earned, largely illegible everywhere else. A brand that has spent years building domain authority in the UK is not starting from a position of strength in Germany. It's starting from zero local relevance, with a global domain that may or may not be recognised by local publishers, local journalists, or local AI systems.
The step zero task here is straightforward, even if the work isn't: before launching in a market, establish what local presence you actually have versus what you're assuming.
Local E-E-A-T has to be built per market - not through translated outreach to international publications, but through genuine relationships with local publishers and local media. Here are the questions to answer before a strategy is commissioned, not after it stalls:
Can you name three local publications in this market that have cited your brand?
Three local entities - partners, associations, media outlets - that reference you in the local language?
Does your brand appear in local-language AI-generated results when users query your category?
If the answer to all three is no, the work that needs to happen first isn't a content strategy. It's a local presence strategy, and it belongs in step zero.
The AI search dimension here compounds the stakes. Weglot's analysis of 1.3 million citations across Google AI Overviews and ChatGPT found that translated websites receive 327% more visibility in AI search than untranslated ones, with the advantage extending even to the original-language content. AI platforms don't process hreflang. Local relevance has to be legible at the content and entity level - your brand needs to be recognised as a genuinely local actor, not just translated into the local language. Which means local authority building is now also an AI visibility decision, with compounding returns for brands that invest in it early.
Two markets with a genuine step zero investment consistently outperform ten markets built at surface level. The question isn't how many markets you can launch into - it's how many you can genuinely be present in.
The work nobody wants to bill for
The teams that consistently perform across markets aren't the ones with the biggest budgets or the most aggressive timelines. They're the ones that did the preparation that's easiest to skip - the market fit analysis before the strategy was commissioned, the governance conversation before the first page was localised, the authority-building work that produces no deliverable for months.
The preparation doesn't feel like progress, and it's hard to put on a Gantt chart, hard to sell internally, and easy to defer until the strategy is already running, at which point it’s no longer preparation, it’s damage control.
I've never seen a multi-market SEO strategy recover from a skipped foundation. I've seen quite a few fail very elegantly, and right on schedule.
Article by
Alizée Baudez
Alizée Baudez is an international SEO consultant specialising in multi-market strategy, localisation workflows, and AI search visibility. She works with brands across industries, from fintech and SaaS to e-commerce and higher education, across Europe, North America, and the Middle East. Her Digital Expatriation framework, a methodology for understanding markets from the inside before building strategy, shapes how she approaches every new market entry. She also teaches SEO at the graduate level and consults on French-market entry for international companies.
stay in the loop