Global brands need market-specific messaging because each market has its own competitors, its own unclaimed territory and its own cultural route in. SOMIN's cross-market read for Land Rover Defender shows it plainly: the brand held 13% share of voice in the UK, 0% in Japan and 22% in the US, so one global message would have been wrong for at least two of the three.
Key takeaways
- AI has made content cheap to produce, so the advantage has moved to knowing what each market needs to hear.
- Translation is the floor: in CSA Research's 29-country survey, 76% of online shoppers preferred product information in their own language.
- One brand can be concentrated in one market, absent in a second and under-indexed in a third, as the Defender read across the UK, Japan and the US showed.
- Keep the brand promise and product truth global; localise the persona, the tension, the pillar, the creators and the timing.
- Read every market's competitors and creators separately, and never pool them into one global number.
Why does one global message stop working in the AI age?
One global message stops working because generative AI lets every competitor fill every market with fluent, generic content, and audiences expect more than that. The cost of making a variant has fallen; the cost of making the wrong one has not.
The expectation is well documented. McKinsey found that 71% of consumers expect companies to deliver personalised interactions and 76% get frustrated when that does not happen. A later McKinsey Quarterly article from January 2025 notes that generative AI now lets marketers develop tailored content at scale at lower cost. Production is no longer the bottleneck.
The new risk is sameness. In SOMIN's ads-optimisation workspace, one insight built on the cultural homogenisation angle reads "Authenticity is the antidote to homogenization": as brands lean on similar AI templates and formats, their messaging converges and loses a distinct voice. A 30-day whitespace study in the same workspace (156 posts, September 2026) surfaced a 39-post pain-point theme on diverse consumer engagement, describing brands that struggle with a one-size-fits-all approach and end up with ineffective cross-market branding.
Is localisation just translation?
No. Translation gets a brand into a market; localisation gets it into the conversation. Language matters, but the competitive set, the audience tension and the cultural route in matter more.
Start with the floor. CSA Research surveyed 8,709 consumers in 29 countries for its 2020 "Can't Read, Won't Buy" study: 76% preferred to buy products with information in their native language, and 40% said they will never buy from websites in other languages. A brand that skips this step loses buyers before the message is even read.
A perfectly translated message can still be the wrong message, though. In the Defender read, the reasons to believe were identical in all three markets: 4x4 heritage since 1948, Terrain Response 2 and a 3.3 ft wading depth on OCTA. The USP shifted only for Japan, reframed towards shared adventure, and promotions were local, from a 0% APR advance payment plan in the UK to Build Your Own and Book an Experience in the US. The product stayed global; the argument did not.
How different can one brand look across three markets?
Very different. SOMIN read 766 Adventurer-tagged posts from Defender, its named competitors and matched creator panels across the UK, Japan and the US over 12 months, and each market pointed somewhere else. The full method is in the Land Rover Defender case study.
- United Kingdom: Defender held 13% share of voice (15 posts) in the 90-day cohort, against 62% for competitors. Its voice was concentrated in extreme off-road, with only 2 posts on lighter community framing. The recommended pillar was Defender Community / Shared Adventure, backed by heritage such as Camel Trophy and Defender Experiences.
- Japan: the owned channel had published 0 posts in 12 months. Competitors held the luxury-sustainability frame and creators held authentic freedom. The lead territory, Confidence in Challenging Conditions, had only 7% competitor share, but step one was simply to start posting again.
- United States: the largest tension in the study, Versatile Adventure-Ready Vehicles, ran to 122 posts. Defender USA held 23% of it while rival luxury brands owned the everyday-luxury reframing, so the verdict was Blocked and the move was to lean in through the accessory ecosystem rather than start fresh.
Each market also surfaced its own lead persona: The Adventurous Innovator in the UK (33% of the corpus), The Adventurous Connoisseur in Japan (64%) and The Adventurous Voyager in the US (44%). These are SOMIN-generated personas built from published content, not survey respondents, but they make the point: "adventure" was three different promises in three markets.
What should stay global, and what should change by market?
Keep what is true about the brand and the product; localise what depends on the audience and the competition. SOMIN's Coca-Cola read across India, the United Kingdom, Mexico and South Africa shows the split well.
In every market, Coca-Cola's voice sat in celebration and sport, alongside its rivals. Foodie creators were elsewhere. Home-cooking enablement made up 62.9% of Indian Foodie-creator content while the cola category had 0% there, and authentic regional culinary discovery took 76.8% of creator content in Mexico and 82.9% in South Africa, again with the category absent. The recommendation in the Coca-Cola case study was the same strategic move everywhere, lead with a culinary whitespace pillar, but with a different persona and cultural route in each market.
| Layer | Keep global? | What to localise |
|---|---|---|
| Brand promise and brand codes | Yes | Tone can flex within the brand voice |
| Product truth and reasons to believe | Yes | Which proof point leads in each market |
| Persona | No | One lead persona per market, built from local content |
| Tension and content pillar | No | Chosen from each market's own whitespace |
| Creators and cultural route in | No | Local creator panel, local food, local references |
| Offer and promotion | Rarely | Local financing, pricing and calls to action |
| Timing | No | Local festive, sporting and retail calendars |
How does AI make market-specific messaging affordable?
AI makes it affordable by doing the market reading, not only the writing. The old blocker was research: one study per market, weeks per study. Agents that read each market's ads, posts and creators can build a separate evidence base for every market in the time a team used to spend on one.
That is how SOMIN's AI agents run the loop: Listen, Decode, Decide, Make and Ship, once per market. The Defender briefing drew on six workspaces, ten studies and three insight chains, one per market and never copied between borders. Each step stays traceable: a pillar points to a tension, the tension to a study, and the study to the posts behind it.
The discipline matters more than the speed. In SOMIN's audience-research workspace, the insight "To truly connect, dare to diverge from the trending path" rests on herd behaviour theory: brands follow prevailing advertising trends without checking whether those trends fit their own audience. The tension behind it carried 2,061 engagements across 65 posts. A global team that copies its lead market's winning campaign into every other market is running the same herd behaviour inside one company.
Where should a global team start?
Start small and make the comparison honest. Three or four markets are enough to see whether a global message is landing evenly or not.
- Name the local competitive set in each market; it is rarely the same list everywhere.
- Add a local creator panel as a proxy for what the audience is engaging with.
- Measure share by content strand inside each market, as a share of each cohort's own output, never as raw counts pooled across markets.
- Call whitespace only where brand and competitors are both quiet and creators, or another audience signal, are present.
- Write one lead pillar per market and check it against a real asset the brand has permission to use.
- Keep the global layer thin: promise, product truth and brand codes.
For the measurement side, our guide to share of voice by market walks through the metrics, and SOMIN for brands shows how brand teams run the same read on their own category. Both start from the same rule: read each market on its own terms.
FAQ
What is market-specific messaging?
Market-specific messaging is a brand message built for one market's competitors, audience tensions and culture, rather than translated from a global master. The brand promise and product truth stay the same; the persona, pillar, creators and timing change.
Does market-specific messaging dilute a global brand?
Not if the global layer is kept clear. In SOMIN's Defender read the reasons to believe were identical in the UK, Japan and the US; only the lead pillar, the USP framing and the promotions changed.
How many markets should we localise for first?
Start with the markets that matter most to revenue and that look least like your lead market. Two to four markets are enough to show whether one message is landing evenly.
Can AI write market-specific messaging on its own?
AI can draft variants quickly, but it needs a market-specific evidence base to draft from. Without local competitor and creator data it tends to produce fluent copies of the same global idea, which is the homogenisation problem in a new format.
Pick one campaign that currently runs unchanged in three markets and pull the last 90 days of local competitor and creator content for each. If the shares look alike everywhere, keep the message. If they look like Defender's 13%, 0% and 22%, the next brief should be three briefs.
