Why Western Union Loses in AI - Argentina Case Study

An AI Visibility analysis reveals: How ChatGPT, Gemini & Claude interpret Argentina's money transfer market – and why local reality is completely absent

---

---

I. Why This Case Study?

---

Introduction & Objective

International money transfer services are among the most widely used digital services in Argentina—a country characterized by intense currency volatility, strong regulations, and high demand for international remittances.

Yet while users clearly search for reliable platforms, a second, far more decisive filter operates in the background:

---

The Invisible Competition

Modern LLMs (ChatGPT, Gemini, Claude, Perplexity) answer thousands of questions daily such as:

• "What's the best platform for sending money?"
• "How does Wise work in Argentina?"
• "Is Western Union cheaper than Xoom?"

What determines visibility is not actual brand quality, but rather:

→ Which sources the AI uses for its world knowledge
→ Which brands dominate in its internal knowledge graphs

This case study examines precisely this mechanism.

---

The Brands Under Analysis

Brand	Domain	Role
Western Union	westernunion.com	Global market leader
Wise	wise.com	Digital challenger
Prex	prexcard.com.ar	Local player

---

The Central Questions of This Study

Visibility:

• Which domains do AI models cite most frequently?
• How do the models describe these brands?

Semantic Spaces:

• Which topics are linked to these brands?
• Where are the semantic gaps?

Market Dominance:

• Which players dominate the Argentine semantic field?
• Why are some brands nearly invisible?

---

Methodology

For this analysis, we used Waikay.io (developed by Dixon Jones). This tool reveals the "AI Knowledge Graph Sources"—the domain clusters on which LLMs build their answers.

Analyzed AI Models:

• ChatGPT
• Google Gemini
• Anthropic Claude
• Perplexity

---

First Central Finding

There are massive differences between:

✗ Global brand presence
✗ Actual AI world knowledge

---

What This Means

Even established brands can, in AI responses:

• Be described incorrectly
• Appear in the wrong context
• Be completely absent

---

II. Methodology: How AI Visibility Was Measured

This case study is based on a structured analysis of three international money transfer brands that are active or relevant in Argentina:

• Western Union – westernunion.com
• Wise – wise.com
• Prex – prexcard.com.ar

All three brands were examined using a standardized AI Visibility analysis, focusing on two core questions:

1. How do AI systems interpret these brands?
2. On which domains and sources does the AI build its knowledge?

---

2.1 Comparison Model

To ensure comparability, both global and local versions of the domains were considered:

Brand	Global Domain	Local Domain (Argentina)
Western Union	westernunion.com	westernunion.com/ar/es
Wise	wise.com	wise.com/ar
Prex	—	prexcard.com.ar

---

2.2 AI Models Used

The evaluation encompasses four of the most important LLM ecosystems:

• ChatGPT (OpenAI)
• Google Gemini
• Anthropic Claude
• Perplexity / Sonar

All models were tested with the same question set and identical comparison parameters to identify answer consistency and source usage.

---

2.3 Analysis Tool: Waikay.io

For the evaluation, we used the tool Waikay.io (developed by Dixon Jones).

What Waikay.io Reveals:

Analysis Dimension	What Is Measured
Sources View	Which domains an AI uses for knowledge formation
Topic Reports	Which topics a model recognizes as relevant
Commercial Queries	Which brands dominate in transactional questions
LLM Answers	How the AI describes each brand

---

2.4 Analysis Focus

The entire report concentrates on a single guiding question:

Examined Aspects:

Source Structure:

• Which domains are used as knowledge sources?
• How often is each brand cited in Knowledge Queries?

Commercial Visibility:

• How often does each brand appear in Commercial Queries?
• Which brand is preferred in "best platform" questions?

Semantic Gaps:

• Which information do AI systems mix or lose?
• Which semantic gaps lead to misinterpretations?
• Which off-topic sources does the AI draw on when information is missing?

---

What the Methodology Reveals

This analysis goes beyond classic SEO. It doesn't show how brands rank in search engines—but rather how AI systems understand, categorize, and recommend brands.

---

III. How AI Describes the Three Brands (LLM Answers)

This chapter is not about rankings, not about sources, not about convergence maps.

It's about one thing only:

And this is precisely where the discrepancy becomes extremely clear.

---

3.1 Western Union: Global Player – But Weak AI Description

Tested Domains:

• westernunion.com
• westernunion.com/ar/es/home.html

Result: All AIs deliver extremely generic answers—regardless of the Argentine version.

Typical AI Statements About Western Union:

✓ "200+ countries"
✓ "145 years of experience"
✓ "Money transfer worldwide"
✓ "Send · Receive · Tracking"
✓ "Security features"
✓ "Mobile app available"

What's Completely Missing:

✗ Argentina-specific fees
✗ Dólar Blue / parallel exchange rate
✗ BCRA regulation
✗ Problems with the WU app in Argentina
✗ 24h availability (often incorrect in reality)
✗ Cash pickup reality (agents without cash)

3.1.1 What the Content Gap Analysis Shows

The structured analysis of the Western Union Argentina page confirms the AI interpretation:

Critical Gaps:

• ✗ No information on Cumplimiento Normativo (BCRA/AFIP regulatory requirements)
• ✗ No clear details on Velocidad de Transferencia (transfer speed per destination)
• ✗ Missing pages for Uruguay and Chile (important destinations for Argentine users)
• ✗ No details on Opciones de Divisa (which currencies are available)

---

3.2 Wise: Global Digital Brand – Better Understood Than WU

Tested Domain:

• wise.com/ar

Result: The AIs deliver significantly more precise answers than for Western Union.

Typical AI Statements About Wise:

✓ "International transfers at real exchange rates"
✓ "Transparent fees"
✓ "Digital account solution"
✓ "Argentina restricted due to local regulation"
✓ "Modern, transparent, and affordable"

Why Does Wise Appear So Strong?

Wise has a total of 26 Knowledge Citations and is extremely well-documented globally. The brand delivers clear, structured information that AI systems can process well.

But: Many AI answers are based on global information, not on Argentina-specific rules.

---

3.3 Prex: Local Player – Not Recognized by AIs as a "Transfer Brand"

Tested Domain:

• prexcard.com.ar

Result: Prex is not recognized by AIs as a player in the context of "sending money from Argentina."

Typical AI Statements About Prex:

✓ "Digital prepaid card"
✓ "Fintech solution"
✓ "Local payment functions"
✗ "No clear profile as international transfer platform"

---

3.4 Comparison of LLM Descriptions (Quick Matrix)

Brand	AI Description	Quality	Problem
Western Union	global, generic, historical	⭐ medium	no local reality (Argentine context completely missing)
Wise	modern, transparent, digital	⭐⭐ good	often global info, Argentine context partially incorrect
Prex	Fintech/prepaid card	⭐ poor	AI does NOT recognize Prex as transfer service

---

3.5 Why This Matters

These differences in LLM answers reveal the core of the problem:

→ AI doesn't know how money transfer actually works in Argentina
→ Brands are sorted into wrong semantic spaces
→ Local reality is completely absent
→ Globality bias dominates interpretation

---

3.6 Supplement: What the "Facts" Analysis Shows

Beyond the general LLM answers, the Facts View from Waikay.io shows which "facts" AI models claim about Western Union Argentina.

The Numbers at a Glance:

92 Facts total – divided into:

• 46 Facts about "Enviar dinero desde Argentina"
• 46 Facts about "Western Union Argentina"

At first glance, this seems comprehensive. But the analysis of topics shows a different picture.

Problem 1: Extremely Low Thematic Depth

The most-mentioned topics in both fact sets:

Topic	Frequency	What This Means
Argentina	20x	Geographic mention, but no local details
money / money transfers	17x / 13x	Generic term, no Argentina specifics
Western Union	14x / 19x	Brand name repetition without context
website / site	7x / 16x	Technical mention of domain structure
fees	5x / 4x	Mentioned, but never with concrete numbers
exchange rates	4x	Concept known, but no Blue/MEP/Oficial distinction

Problem 2: What's Completely Missing

Not a single fact mentions:

• ✗ BCRA regulations
• ✗ AFIP limits
• ✗ Dólar Blue vs. Oficial
• ✗ RENAPER/CUIT requirements
• ✗ Cepo Cambiario
• ✗ Transfer speed to Argentina
• ✗ Concrete fees in ARS
• ✗ Local payment methods (Rapipago, Pago Fácil)

Problem 3: Model Inconsistency

Different LLMs recognize different numbers of facts:

Model	Facts "Enviar dinero"	Facts "Western Union"
Gemini	8-11	12-11
Claude	7	6
ChatGPT	10	8
Sonar	10	9

Why Is This Important?

This facts analysis reveals the core problem in three dimensions:

1. Quantity ≠ Quality
→ 92 facts sound like a lot, but they're 92x the same superficial information

2. Globality Displaces Locality
→ The AI knows "Western Union" and "Argentina," but doesn't understand "Western Union IN Argentina"

3. Lack of Consistency Between Models
→ When AI systems don't agree, there's no stable semantic basis

Thus, the facts analysis fully supports the previous chapters:

✓ LLM answers are shallow
✓ Local context is missing
✓ Semantic spaces are wrongly set
✓ Brands are not understood as "Argentine" services

---

IV. Sources View: The Convergence Map

This chapter is the core of the entire case study.

Here we see how AI structures its world knowledge on the topic "sending money from Argentina"—and why certain brands dominate.

The data is based on the Sources View from Waikay.io. The convergence map shows:

• Which domains the AI uses as knowledge foundation
• How frequently they are cited
• Whether they are Knowledge Sources (blue) or Commercial Sources (orange)

---

4.1 The Three Dominant Players

Despite different market roles and business models, only three domains dominate the semantic field:

Brand	Domain	Knowledge Citations	Commercial Citations
Western Union	westernunion.com	30	5
Prex	prexcard.com.ar	27	0
Wise	wise.com	26	7

---

4.2 The Actual AI Dominance Fields (Visual Clusters)

Analyzing the source data reveals three distinct clusters:

1️⃣ Global Money Transfer Cluster

Brands: Western Union, Wise, Xoom, Remitly, Ria

This is the largest cluster. The AI preferentially uses these domains for:

• "How do I send money from Argentina?"
• "Best platforms"
• "Fee comparison"
• "How does Wise/WU/Xoom work?"

2️⃣ Local Banking / Fintech Cluster

Brands: Prex, MercadoPago, Billeteras Virtuales

This cluster provides knowledge, but no recommendations.

Prex appears 27 times, but is not categorized as a global solution.
→ AI assigns Prex to local payment transactions, not international transfers.

3️⃣ Off-Topic & Confidence Sources

Here lies the most interesting part of the analysis:

The AI compensates for missing, stable Argentine sources with completely off-topic PDFs and government pages.

Examples from the dataset:

• supremecourt.gov
• UN PDFs
• Apple Business Pages
• FTC Consumer Warnings
• IPCC Climate Report
• nist.gov
• library.tamu.edu

---

4.3 Why These Clusters Are Decisive

The convergence map reveals three fundamental problems:

Problem 1: Narrow Knowledge Base

AI builds its Argentina knowledge almost exclusively on a few global players:

• Western Union
• Wise
• Prex (local)
• Xoom, Remitly, Ria
• Monito (comparison portal)

Problem 2: Semantic Instability

Because local sources are missing, "red cluster" sources mix in.

This happens almost exclusively in markets with:

• High inflation
• Unstable regulations
• Many patchwork solutions
• Fragmented information sources
• Missing official guideline documents

Problem 3: Missing Reality Recognition

The models don't cleanly recognize what Argentine reality is.

This leads to AI answers that are:

• Incomplete
• Globalized
• Incorrect
• Oversimplified
• Partially hallucinated

---

4.4 The Most Important Insight from Chapter IV

AI has a clear, unambiguous order:

→ Western Union = global knowledge (Knowledge dominant)
→ Wise = commercial decisions (Commercial dominant)
→ Prex = local financial knowledge (but not a "winner" in recommendations)

---

V. Topic Map: Which Semantic Spaces AI Connects

The Topic Map shows how AI models structure the entire market thematically.

Here we see which topics AI classifies as central, important, or secondary—regardless of whether this assessment is correct.

This reveals at a glance:

• Which topics are dominant
• Which topics are missing
• Where false interpretations arise
• How stable or unstable the semantic space "sending money from Argentina" is for AI

---

5.1 What AI Recognizes as Central Topics

In the Topic Map, certain themes are repeatedly prioritized. These form the semantic foundation on which AI builds its answers:

Central Topics (blue = Knowledge Domains):

• Send Money / Money Transfer
• Transferencias Internacionales
• Cómo enviar dinero desde Argentina
• Exchange Rates ARS/USD
• Remesas / International Remittances
• Dólar Blue / Dólar paralelo

Why These Topics Dominate:

These topics are globally well-documented and have many stable sources. They shape the semantic space—even if they don't correctly reflect local realities.

---

5.2 What AI Does NOT Recognize (Critical Gaps)

Here the problem begins—and this is precisely what makes this case study valuable.

Critically Missing Topics (red = Missing Knowledge):

Missing Topics	Why This Would Be Important
Argentine Regulation	AFIP limits, anti-money laundering restrictions, identity verification
Daily Transfer Limits	Real usability of services
Digital Identity	RENAPER, CUIT/CUIL – central for Argentina
Local Payment Methods	Mercado Pago, Rapipago, Pago Fácil
Onboarding Processes	How does registration actually work?
Specific Fee Models	Argentina-specific costs
Argentine Banks	Correspondent banks, local partnerships
Exchange Rate Realities	MEP, Blue, Tarjeta – three parallel systems

These Gaps Explain:

→ Why AI overvalues Wise
→ Why AI reduces Western Union to global facts
→ Why AI doesn't interpret Prex as a transfer service
→ Why AI includes off-topic sources (UN, PDFs, SupremeCourt)

---

5.3 Why These Gaps Emerge

The AI builds its Topic Map from the interaction of four source types:

1. Stable Sources (Knowledge)
2. Commercial Sources (Commercial)
3. Local Sources (Local Knowledge)
4. Global Norms (Universal Concepts)

Since there are only few clean local sources, three effects emerge:

A) Global Topics Are Overweighted

The AI insists on universal terms like "Remittances" and "Send Money"—because these terms have stable, global sources.

B) Local Reality Is Underweighted

Important Argentine topics are completely missing from the graph—not because they're unimportant, but because they are insufficiently documented.

C) AI Compensates for Gaps with Off-Topic Documents

→ Exactly the off-topic sources described in Chapter IV.

---

5.4 Why the Topic Map Is Central to This Case Study

The Topic Map answers the question:

And this reality is currently:

❌ Global
❌ Generic
❌ Stable (but wrong)
❌ Barely localized
❌ Free of regulatory details
❌ Compensated with partial information gaps

This Means Concretely:

✓ AI doesn't understand Argentina
✓ AI prioritizes global brands because they provide the most stable sources
✓ Local providers like Prex don't exist in the global context
✓ Western Union is recognized as a global player, but not as a local market leader
✓ Wise is often portrayed above its actual market weight

---

What This Means for Companies:

---

VI. Commercial Intent: Who Dominates "Best Platform" in AI?

While "Knowledge Queries" show which sources AI draws on for facts, the Commercial Intent area reveals something different:

Typical Commercial Queries:

• "Which is the best platform for sending money?"
• "Which fees are lowest?"
• "How do I compare X with Y?"
• "Which app is most reliable?"

This is where real competition emerges—not for knowledge, but for recommendations.

And the result is clear.

---

6.1 Dominance in the Commercial Area

Commercial Queries (Number of AI Citations):

Brand	Commercial Citations
Wise	7
Western Union	5
Prexcard	0

What This Means:

✓ Wise wins in commercial context
✓ Western Union is visible but loses relevance
✓ Prexcard plays no role in "best-of" answers

---

6.2 Why Does Wise Dominate?

Despite a smaller global market share, Wise delivers exactly what AI systems are looking for:

What Wise Does Right:

✓ Clearly structured landing pages
✓ Transparent fees (always visible, always comparable)
✓ Global comparison tables ("Wise vs. Western Union")
✓ Strong semantic signals for "low fees," "international money transfer," "best way to send money"

For AI Systems, Wise Is an "Easy Case":

• Clear facts
• Few contradictions
• Good semantic density
• Strong internal linking
• Clear commercial signals

---

6.3 Why Does Western Union Lose Despite Market Leadership?

Western Union possesses:

• Strong global brand authority
• Long history (145 years)
• Massive offline presence

But AI sees something different:

What's Missing at Western Union:

❌ Generic texts ("200 countries," "145 years") instead of concrete benefit communication
❌ Hardly any local details about Argentina
❌ Little context on fee structure
❌ Few modern semantic signals
❌ Very global rather than local optimization

---

6.4 Why Is Prexcard Completely Invisible?

AI doesn't recognize Prexcard as:

• International player
• Money transfer solution
• Commercial alternative
• Recommendation candidate

Why?

Problem 1: Wrong Categorization
Prexcard is local, not global → Semantics are "Wallet," not "International Transfer"

Problem 2: Missing Commercial Signals
No pages with "Comparison," "Fees," "Best way to send money"

Problem 3: No International Validation
No international review signals, no stable AI citation

---

6.5 The Structural Logic of AI

AI evaluates Commercial Intent based on the following building blocks:

Evaluation Criterion	Wise	Western Union	Prex
Fee Comparison	✓	○	✗
Reliability Signals	✓	✓	○
Easy Explainability	✓	○	✗
Global vs. Local Reach	✓	✓	✗
Stable Source Base	✓	○	○
Total	5/5	3/5	1/5

Interpretation:

Wise meets all five criteria → Rank 1
WU meets three → "Okay, but not the best option"
Prex meets one at most → "No recommendation"

---

6.6 Brief Conclusion on Chapter VI

This chapter shows:

✓ Who dominates recommendations in the AI ecosystem
✓ Why Wise appears as "best platform"
✓ Why WU loses despite size
✓ Why Prex is commercially invisible

---

What This Means for Companies:

---

VII. The Argentina Leak: Why AIs Suddenly Cite Government Sites and UN PDFs

When you first see the source map, something immediately stands out:

Examples from the Data:

• supremecourt.gov
• nist.gov (cryptography certificates)
• earthday.org
• UN documents / PDFs
• IPCC climate report
• Apple Business Pages
• library.tamu.edu

This seems chaotic—but it's a highly relevant signal.

---

7.1 What This "Leak" Means

These off-topic sources don't emerge randomly. They emerge because the AI, on the topic:

This is a typical case of semantic instability.

The AI Attempts to Compensate for What's Missing

When enough genuine sources are missing, the model does three things:

1. Fallback to trustworthy, global documents
→ Government sites, UN, academic libraries

2. Search for structured PDFs
→ Regulatory metadata, security certificates, general "compliance documents"

3. Replacement of missing contexts with global standards
→ US regulation instead of AR regulation

---

7.2 Why This Happens More Strongly in Argentina Than in Other Countries

Argentina has three characteristics that fundamentally confuse AIs:

1️⃣ Complex Currency System Without Documentation

Argentina has multiple parallel exchange rates (Peso, Blue Dollar, MEP, Tarjeta), but neither Western Union nor Wise explain these on their Argentine pages.

The AI cannot distinguish between:

• Official Dollar
• Dólar Blue
• Dólar MEP
• Dólar Tarjeta

—not because these concepts change rapidly, but because no reliable source documents them in a structured, AI-readable way.

2️⃣ Missing Official Digital Knowledge Sources

Only a few reliable sources explain:

• Fees
• Limits
• Regulation
• Travel statutes
• Identification (RENAPER)
• Foreign exchange controls

3️⃣ Many Contradictory Information Sources

• Blogs say A
• Media says B
• Banks say nothing

→ AI doesn't know what "ground truth" is.

In such markets, the model "flees" to stable, global PDFs—even when they're thematically wrong.

---

7.3 What This Leak Reveals About the Entire Market

This Means Concretely:

Problem 1: Too little high-quality, structured content
Problem 2: Local players (like Prex) provide hardly any context
Problem 3: Global players provide only generic info
Problem 4: Argentina-specific facts are missing almost everywhere
Problem 5: There's no single actor providing clean knowledge signals

This Is Why the AI:

✗ Describes Western Union incorrectly
✗ Miscategorizes Prex
✗ Overestimates Wise
✗ Doesn't recognize important local factors at all

---

7.4 Why This Leak Is One of the Strongest Insights of the Entire Case Study

Because it shows:

And even more importantly:

What This Means:

No single company can solve this problem alone—but every company that provides clean, structured, local information gains massively in AI visibility.

---

Practical Consequence:

Companies operating in unstable markets like Argentina must:

1. Create their own knowledge sources (no expectation that AI will "automatically understand")
2. Explicitly document local contexts (BCRA limits, fees, processes)
3. Send consistent signals (not contradictory info on different pages)
4. Provide structured data (Schema.org, FAQs, clear comparison tables)

---

VIII. Comparison of the Three Brands (Western Union, Wise, Prex)

This chapter summarizes the most important findings of the analysis.

The three providers—Western Union, Wise, and Prex—show completely different visibility patterns in AI systems. Contrary to their market size or local popularity, they are understood, weighted, and categorized differently by AI models.

To make the results tangible, here is a clear comparison table across all relevant dimensions.

---

AI Visibility Matrix – Tabular Comparison

Analysis Dimension	Western Union	Wise	Prex
Knowledge Sources (citations in knowledge queries)	30 citations	26 citations	27 citations
Commercial Sources (citations in commercial queries)	5 citations	7 citations	0 citations
LLM Recognition (brand profile)	Very generic, global, barely any Argentina reference	More precise, functional, clear advantages in answers	Not recognizable as money transfer service
Topic Depth	Low – AI mentions standard facts ("200 countries")	Medium – better explanations of fees, speed	Very low – AI often confuses or doesn't recognize Prex
Local Knowledge (Argentina-specific)	Weak – almost no local specifics	Moderate – partially local info	Very weak – AI doesn't understand role & offering
Entity Stability (brand stability in AI knowledge graph)	High – but global, not local	Medium – clear classification	Low – unstable, often misclassified
Commercial Ranking	Inferior despite market size	Dominant	No relevance
AI Interpretation Quality	"Global shadow profile" – no Argentina depth	"Functional profile" – good answers	"Dead profile" – barely usable answers

---

Brief Contextualization of Results

1) Western Union – Visible, But Misinterpreted

WU has the most Knowledge citations, but AI interprets the brand almost exclusively from global sources.

Consequence: The LLMs deliver extremely generic, superficial answers without local context.

2) Wise – Smaller, But Commercially Dominant

Wise has the most commercial citations (7).

This means: AI recommends Wise more frequently in answers with purchase or comparison intent.

In Argentina, this creates an asymmetric overpresence.

3) Prex – Local Player Without AI Relevance

Prex receives many Knowledge citations (27)—but: AI doesn't recognize Prex as a money transfer provider.

Thus, Prex has no Commercial Impact (0 citations).

---

Interpretation: What the Table Really Means

The matrix shows three fundamentally different AI profiles:

Western Union: Visible, But Imprecise

→ Knowledge is global, not local
→ High presence, low relevance

Wise: Functionally Dominant

→ Higher probability in ranking or "best platform" answers
→ Lower presence, higher conversion probability

Prex: Practically Invisible

→ Despite local significance, Prex plays no role for AI
→ High presence, but wrong category

---

What This Means in Practice

• Western Union is often cited but rarely recommended
• Wise is cited less often but recommended more frequently
• Prex is often cited but misunderstood

This comparison is the link between the analytical chapters (IV-VII) and the strategic contextualization in the next chapter.

---

IX. Strategic Interpretation of Results

This chapter summarizes what the AI data really means.

It's no longer about numbers, scores, or tables—but rather:

---

9.1 How AI Forms Semantic Spaces

LLMs don't work with "keywords," but with concept spaces.

A concept space emerges from three layers:

1. Primary Sources (Dominant Sources)

→ Domains that are frequently cited
→ Determine the foundation of world knowledge

2. Secondary Signals (Topics, examples, recurring patterns)

→ Typical terms, narratives, situations
→ Stored as "normal" for a topic

3. Stabilizing Additional Sources

→ PDFs, government sites, reports
→ Used when primary sources have gaps

---

In the Case of the Argentine Money Transfer Market:

The primary sources are extremely narrow, yet they dominate the entire perception:

• Western Union
• Prex
• Wise
• Ria, Xoom, Remitly (commercial layer)

---

9.2 Why Argentina Is Semantically Unstable

Argentina creates a special problem for AI:

✗ Complex currency system poorly documented by transfer providers
✗ Two parallel exchange systems without clear explanation (official dollar vs. Blue dollar)
✗ Unclear legal situation for transfers—not explained on brand websites
✗ Many unofficial solutions (cuevas, informal exchange houses)—never mentioned
✗ Missing structured information on local regulations

This Leads to Three Structural Breaks in the AI Knowledge Space:

1. Missing Clear Entities

There are no stable, permanent terms and rules.

2. Semantic Gaps in the Market

Neither Western Union nor Wise cover all relevant concepts:

• Límites
• Trabas
• Impuesto PAIS
• 70/30 rule
• Mercado Pago Restricciones

3. Loss of Local Realities

AIs tend to interpret Argentina through "global logic":

→ Stable markets
→ Clear fees
→ Simple banking structures
→ Digital identity = unambiguous

9.3.1 Example: The Missing Compliance Documentation

The Content Gap Analysis of Western Union Argentina shows the problem concretely:

What's Missing:

• No page on "Cumplimiento Normativo" (BCRA limits, AFIP rules, identity verification)
• No information on transfer speed per destination
• No explanation of currency options (USD, EUR, ARS)

What Happens Then?

The AI finds no reliable source on Argentine regulations and falls back on:

• UN PDFs on international money laundering prevention
• NIST documents on cryptography standards
• FTC Consumer Warnings (US-based)

---

9.3 Why Western Union Is Misinterpreted

Western Union is extremely present in Argentina. However, AI recognizes mainly the global brand, not the local reality.

This Causes the AI to:

✗ List global facts ("200+ countries," "145 years")
✗ Repeat generic security notices
✗ Not know local fee models
✗ Not consider AR-specific restrictions
✗ Not describe real user journey

---

9.4 Why Prex Appears Strong But Isn't Recommended

Prex appears strongly in Knowledge Sources because:

✓ The domain provides many structured contents
✓ The pages are technically clean
✓ The brand appears entity-stable (clear thematic focus)

BUT:

Prex is not primarily an international transfer solution.

AI doesn't always recognize this clearly. An overweighting in the knowledge space emerges that doesn't correspond to reality.

---

9.5 Why Wise Wins Despite Smaller Knowledge Base

Wise works excellently internationally. AI "loves" Wise because:

✓ The domain has extremely many structured financial contents
✓ International logic is consistent
✓ Fee transparency is clearly communicated

The Problem:

Wise can only be used to a limited extent in Argentina. However, AI adopts international narratives—without local restrictions.

---

9.6 Why Local Realities Are Completely Missing

No AI mentioned in answers:

Missing Concepts	Argentine Reality
dólar blue	Parallel currency
impuesto PAIS	Tax burden on transfers
cepo cambiario	Foreign exchange restrictions
restricciones del BCRA	Regulatory limits
cuevas	Informal exchange houses
límites por CUIT	Identity-based limits
métodos informales	Real user practices
reales diferencias de tarifas	Hidden costs
AR-specific identity verification	RENAPER, CUIT/CUIL
Volatility within 24h	Exchange rate chaos

---

9.7 Consequence: Context Loss + Topic Gaps = Wrong Answers

AI answers seem plausible at first glance. But they're based on:

✗ Global knowledge
✗ Missing AR context
✗ Incomplete source clusters
✗ Wrong priorities
✗ Compensating off-topic documents

This Leads to Systematic Misinterpretations:

• Incorrect fee comparisons
• Ignored regulations
• Overestimation of global players
• Underestimation of local solutions
• Erroneous "best platform" recommendations

---

9.8 Conclusion of This Chapter

The data clearly shows:

This Explains:

→ Why Western Union is misrepresented
→ Why Prex appears too strong
→ Why Wise is overestimated
→ Why the market remains semantically unstable
→ Why off-topic PDFs appear
→ Why local realities are missing
→ Why AI recommendations are often wrong in Argentina

---

What This Means for Companies:

In semantically unstable markets like Argentina, brand strength doesn't determine AI visibility—but rather:

1. How clearly the brand is documented
2. How consistent the signals are
3. How well local realities are reflected
4. How stable the source structure is

---

X. Conclusion & Action Recommendations

The analysis of the three brands—Western Union, Wise, and Prexcard—clearly shows how AI systems currently categorize the money transfer market in Argentina.

Two global players (WU, Wise) and one regional provider (Prexcard) form the core of AI answers, although the real market is much more diverse and numerous other services are active.

---

The Central Finding

This creates a distorted picture:

• Global brands appear more stable than they actually are in Argentina
• Regional providers like Prexcard are partially miscategorized
• Other services don't appear at all, despite being used in practice

---

This Leads to Three Consequences:

1️⃣ AI Answers Are Heavily Simplified

They orient themselves to the best-known domains and deliver generic descriptions that barely address local rules, restrictions, or everyday user problems.

2️⃣ Knowledge About Argentina Is Primarily Derived From Global Sources

This means:

• Fees, limits, regulation, peso logic, or dólar blue are barely understood
• Models replace missing context with international PDFs, government sites, or random off-topic sources
• This creates gaps or even clear misinterpretations

3️⃣ Market Interpretation Remains Incomplete

Even established players—whether local or international—are only considered in AI answers when stable, recurring sources exist.

---

Concrete Action Recommendations: Western Union Argentina Example

The Content Gap Analysis shows what's missing—and what must be built:

Phase 1: Content Creation (High Priority)

1. Create Regulatory Clarity

• Create "Cumplimiento Normativo" page
  • Document BCRA limits
  • Explain AFIP requirements
  • Describe RENAPER/CUIT identity verification

2. Complete Missing Destinations

• Create pages for Uruguay
• Create pages for Chile
• Create pages for Peru

3. Increase Transparency

• Document transfer speed per destination
• Clearly present currency options (USD, EUR, ARS, Blue?)
• Present payment methods clearly

Phase 2: Structural Improvements

1. Hub Page "Envíos desde Argentina"

• Visual overview of all destinations
• Filter options (speed, cost, method)
• Comparison table for different destinations

2. Local FAQ Section

• "¿Cómo funciona con el dólar blue?"
• "¿Cuáles son los límites mensuales?"
• "¿Qué documentos necesito?"

Phase 3: Semantic Optimization

1. Implement Structured Data

• Schema.org for financial services
• FAQ markup for common questions
• Breadcrumb navigation

2. Consistent Terminology

• Uniform terms across all pages
• Use local terms (cepo, dólar MEP, BCRA)

---

What Companies Can Learn From This

1. AI Needs Stable, Clear Brand Signals

Without clear page structures, unambiguous explanatory pages, fee comparisons, local guides, and recurring mentions, no reliable brand image emerges.

Concrete:

• FAQ pages with structured data (Schema.org)
• Comparison tables (Wise vs. Western Union)
• Local landing pages with specific information
• Consistent terminology across all channels

2. Local Specifics Must Be Explicitly Explained

Especially in countries like Argentina, clear, well-documented presentation is crucial:

What Must Be Documented	Why
Limits	Daily/monthly transfer limits
Identity Verification	RENAPER, CUIT/CUIL requirements
Payment Methods	Mercado Pago, Rapipago, bank transfers
Currency Logic	Official rate vs. Blue vs. MEP
Risks & Fees	Transparent cost breakdown

3. International Players Should Strengthen Local Sub-Pages

Western Union is an example:
Strong global presence, but weak local AI profile → leads to generic answers.

Solution:

• Argentina-specific content pages
• Local blog posts on regulations
• Fee calculator with ARS/USD/Blue
• User stories from Argentina

4. Regional Providers Must Make Their Role Visible

Prexcard is an example:
Widespread in Uruguay and partly in Argentina—but AI doesn't recognize the international utility.

Solution:

• Clear positioning as "international transfer solution"
• Comparison pages ("Prex vs. Western Union")
• Use cases for cross-border payments
• External validation (reviews, media mentions)

---

What This Case Study Means for marcus-a-volz.com

This case study clearly shows how large the gap between real market and AI knowledge can be—especially in complex markets like South America.

And This Is Precisely Where the Opportunity Lies:

Helping companies build their brand knowledge so that AI systems:

• Understand them (correct categorization)
• Correctly classify them (right semantic spaces)
• Reliably cite them (Knowledge + Commercial Queries)

---

The Service Portfolio From This:

1. AI Visibility Audits

→ How do AI models currently interpret your brand?

2. Semantic Gap Analysis

→ What knowledge and topic gaps exist in your market?

3. Market Intelligence for AI Systems

→ How do you structure content so AI understands it?

4. Local Knowledge Frameworks

→ How do you build local expertise that AI can interpret?

5. Commercial Intent Optimization

→ How do you position yourself in "best platform" answers?

---

This Makes It Clear:

---

Closing

If your organization wants to understand how AI models interpret your brand across different countries and platforms—and how to strengthen this visibility—feel free to reach out.

I help companies build a clear, reliable presence in markets where traditional SEO signals are no longer enough.