{{title}}
{{textarea}}
{{task}}
{{select}}
{{team}}
{{input}}
{{solution}}
{{result}}
ICDS Research Brief Global Marketing
Your Brand in Germany and Your Brand in Brazil Are Two Different Media Spaces
Across 66 FMCG brands in 26 countries, we measured how identical media efforts convert into consumer behaviour — and found that the gap between markets is 30 times more important than any local agency decision.
VM
Vadim Matyushkin Behavioral Scientist & Sociologist · ICDS
A quarterly report across two markets. Germany: a Nivea campaign launching a new skincare range. An identical campaign in South Korea — same budget, same mechanics, comparable peak mention volume in the monitoring data. Germany shows brand search growth. Korea shows sentiment of 0.10 against Germany’s 0.31. Standard marketing conclusion: ‘the Korean audience is less engaged.’ The agency receives a formal note.
A different company, a different market. Evian launches a campaign simultaneously in France and Brazil. France — the mention peak clearly converts into search growth. Brazil — the peak is there, the search is not. Conclusion: ‘the local team failed to communicate the message.’ The team is restructured.
In both cases, the diagnosis is wrong. The agencies and local teams did everything correctly. The problem is that the brand managers were looking at an instrument that does not see what they need to see — and were drawing organisational conclusions from a measurement artefact.
26Countries in the corpus66 FMCG brands · 10 categories · 2023–2024
30×CSRI rangeFrom +2.33 (Thailand) to −31.8 (Philippines)
57–59%Response durabilityIdentical across ALL regions — including the ‘non-responding’ ones
Three Cases: Identical Budget, Different Story
Before the data — three reconstructed cases from our corpus. Each describes a real pattern typical of the corresponding market pair.
Case ANivea · DE → KR
Low Sentiment Is Not a Cold Reception
Standard interpretation: the Korean audience perceives the brand less positively. In Germany the peak carries sentiment of +0.31 and sustained search growth; in South Korea, with identical mechanics, sentiment is +0.10 — a threefold difference.
Correct interpretation: Korean consumers use structurally more restrained evaluative language in public online spaces. The norms of public emotional expression in South Korea and Germany are fundamentally different — and this does not reflect actual brand attitude.
From our corpus: South Korea’s CSRI = +1.72 — one of the highest in the dataset. Korean consumers respond actively to media peaks, seeking brand information. They simply do so quietly.
Case BEvian · FR → BR
The Search Is There — Just Not Where You Are Looking
Standard interpretation: the Brazil campaign failed. In France the peak converts into search; in Brazil the peak is there but the search is not.
Correct interpretation: In Brazil, Google is not the primary channel for brand discovery — consumers use YouTube, WhatsApp groups, and Instagram. Evian is receiving a response, but in channels invisible to the standard dashboard. France’s CSRI is negative (median −1.01), itself a paradox for a developed market — pointing to a worldwide measurement bias.
Brazil CSRI = +1.42 — substantially better than one might expect. The conclusion that follows from the data: ‘our measurement stack is blind in Brazil.’
Case CRed Bull · AT → TH / PH
Paid-Only as a Sufficient Catalyst — but Only in Asia
Standard interpretation: the same playbook should work everywhere. In Western Europe, every mention peak is a mix of editorial and paid content — 0% of peaks are paid-only.
Correct interpretation: In South and Southeast Asia, 27% of peaks are exclusively paid, with no organic component — not because Red Bull underinvests in earned media, but because Asian audiences accept paid influencers as a sufficient and credible source, and respond behaviourally without editorial confirmation. Western audiences wait for a mixed signal.
The practical consequence: a Western playbook transplanted to Thailand will systematically overinvest in earned media where paid alone is sufficient.
The Paradox: Identical Mechanics, Different Conversion
To understand why the three cases above are not anomalies, we need to look at the regional data in full. It reveals a structural paradox.
| Region | N | CSRI median | ΔSearch t0 | ΔSearch t+1 | Durability t+1>t0 | Video / text | Paid-only peaks |
|---|---|---|---|---|---|---|---|
| Western markets | 32 | +0.10 | +1.0 | +2.3 | 59% | 56% / 44% | 0% |
| Asia | 18 | +0.77 | −0.1 | +2.6 | 56% | 80% / 13% | 27% |
| Latin America | 7 | −0.74 | −6.3 | −5.7 | 57% | 71% / 29% | ~0% |
| Africa / Middle East | 7 | −1.88 | −10.4 | −7.4 | 57% | — | — |
Look at the Durability column. In 57–59% of brands across ALL four regions, search activity at t+1 is higher than at t0. This means the delayed consumer response — growing interest in the brand after a media event — operates identically in Nigeria and Germany. The behavioural mechanism is universal.
But CSRI — the index of media-peak conversion into Google search — ranges from +0.77 in Asia to −1.88 in Africa and the Middle East. If the difference were behavioural, we would expect different durability rates. Durability is identical — which means the difference is not behavioural, but instrumental.
Central finding
Consumer response to a media peak is universal. What differs is not how the audience responds, but in which channel it does so, how it expresses evaluation publicly, and what constitutes a sufficient trust signal for it. These three parameters — platform landscape, cultural expressiveness, and trust architecture — are structural. They cannot be changed by campaign quality.
Three Structural Differences Between Markets
1
Where the audience expresses its response: the platform landscape.
70% of Nigerian internet users actively use social media to search for brand information — but through WhatsApp, YouTube, and Instagram, not Google. Brazil shows an identical pattern. For these markets, Google Trends is not a measurement instrument — it is a blind spot. The evidence: post-peak durability in Africa/Middle East (57%) is virtually identical to the West (59%). If the audience were not responding, this figure would be lower. The response is there; the instrument does not see it.
2
How the audience expresses evaluation: cultural expressiveness.
The median sentiment index of East Asia (South Korea, Singapore) is +0.10. The Anglophone West: +0.39. Africa: +0.40. A fourfold difference at comparable volumes and budgets — not a different attitude toward brands, but different norms of public emotional expression. Reading this sentiment without adjusting for the Cultural Expressiveness Baseline (CEB) means systematically underestimating campaign success in East Asia.
3
What constitutes a sufficient signal: the trust architecture of content.
ESG campaigns and social initiatives generate peaks in the West (25% US/UK, 24% Europe) and barely function as a catalyst in South Asia (7%) — not because brands do not invest there, but because those audiences have not developed a stable appetite for ‘brands with a position’ as a media catalyst. The paid-only pattern is symmetrical: 0% of peaks are purely paid in the West (audiences wait for editorial confirmation), versus 27% in Southeast Asia, where editorial confirmation is not required for a behavioural response.
CSRI: Search Conversion Index by Country
Based on a corpus of 65 observations (66 brands minus one outlier), we calculated the Country Search Responsiveness Index (CSRI) — the index of media-peak conversion into Google search — for 11 countries with more than one observation.
| Country | N | CSRI mean | CSRI median | Reliability | Interpretation |
|---|---|---|---|---|---|
| USA | 12 | +0.36 | +0.10 | Moderate | Benchmark for a mature Western market. 10 pp of peak → ~1 point of brand search at t+1. |
| India | 8 | +1.19 | +1.35 | Moderate | High conversion. Response is delayed: near zero at t0, but +8.1 at t+1. |
| Germany | 5 | +0.47 | −0.30 | Low | High dispersion — 5 brands, very different results within the sample. |
| France | 4 | −0.94 | −1.01 | Low — artefact? | Unexpectedly negative CSRI for a developed market. Likely worldwide measurement bias. |
| Mexico | 4 | −1.34 | −0.74 | Low — artefact | Search migrates to YouTube and local platforms, not Google. |
| Brazil | 3 | +0.95 | +1.42 | Low | Positive CSRI — but N=3. Interpret with caution. |
| Nigeria | 3 | −3.30 | −2.76 | Low — artefact | 70% of Nigerian users search for brands in social media, not Google. |
| South Korea | 2 | +1.72 | +1.72 | Very low | High CSRI alongside restrained sentiment — an important combination. |
| Thailand | 2 | +2.33 | +2.33 | Very low | Highest CSRI in the dataset. Strong collective information-sharing culture. |
| Indonesia | 2 | −4.14 | −4.14 | Very low — artefact | Google does not dominate. Response via WhatsApp and YouTube. |
| Philippines | 2 | −31.8 | −31.8 | Very low — artefact | Extreme artefact. Google Trends is effectively blind in this market. |
How to read CSRI
CSRI +0.10 (USA): every 10 percentage points of peak above baseline yield ~1 point of brand-search growth at t+1. CSRI +1.35 (India): the same 10 pp yield 1.35 points — India converts peaks into search 13.5× more efficiently, in the data Google captures at all. The real gap is larger.
Correction formula:
Adjusted Effectiveness = ΔBrand_search(t+1) / CSRI_CountryA brand generating +4 points in India outperforms a brand generating +8 points in Germany after adjusting for the structural baseline.
Practical Implications for Global Teams
The three structural differences do not require new agencies and do not require different creative. They require a different set of questions before the campaign launches.
Three questions to ask before launching a global campaign
In which of our markets is Google the primary channel for brand discovery? For markets where the answer is no — what are we measuring instead?
Have we established a Cultural Expressiveness Baseline (CEB) for each key market? A sentiment score of +0.15 in Japan and +0.15 in Australia are not the same thing.
Are we expecting an ESG initiative to generate earned media in a market where that mechanism does not exist? Or expecting a paid influencer to be an insufficient catalyst where it is in fact sufficient?
The more fundamental shift concerns attribution. When a campaign ‘does not work’ in an emerging market, the first question should not be ‘what did the local team do wrong’ but ‘does our measurement stack see this market.’ In most cases, the answer is: not fully.
Methodology note
How the study was built
The research is based on two ICDS series articles: (1) “Media Events in FMCG: Patterns Across Product Categories and Countries” (C3) — N=66 FMCG brands, 26 countries, 10 categories, 8 trigger types. (2) “Does Geography Shape Signal? Cross-Country Variation in Consumer Search Responsiveness to Brand Media Peaks” (Article 7) — N=65 observations, 26 countries. CSRI = ΔBrand_search(t+1) / (Peak_Deviation% / 10). Social media monitoring data: commercial platform. Search data: Google Trends. Consumption trend data: Kantar Brand Footprint, DataReportal 2025.
Limitation. Average of 2.5 observations per country. Country-level CSRI estimates should be treated as directional signals rather than stable measurements. A minimum of 10 observations is required for reliable country-level estimates.
Full research articles: icds.institute.
VM
About the author
Vadim Matyushkin
Behavioral Scientist & Sociologist · ICDS
Vadim Matyushkin is a psychologist and sociologist with close to twenty years of research into digital behavior, trust, and information dynamics, and a researcher behind the Institute of Communication and Data Science (ICDS). His work has supported organizations including Coca-Cola, PepsiCo, Mars, Danone, Nestlé, and Bayer in moving from self-reported survey data toward direct behavioral evidence.
ICDS — Institute of Communication and Data Science is an independent research institute focused on understanding how trust, behavior, and reputation are formed in digital environments. ICDS operates as an intellectual institution — not an agency, not a platform, and not an educational provider.