Marcus by Goldman Sachs AI Market Strategy Report - Best Banks

AI Company Market Strategy Report | Best Banks | June 2026

Answer Capsule

Marcus by Goldman Sachs carries the lowest raw mention presence rate in the Best Banks category at 12.6%, yet records a net sentiment score of 0.5876 with zero negative mentions across 1,536 observations. The brand's positive AI framing is a genuine asset, but the source footprint is too thin to compete for meaningful recommendation value. Marcus performs best on ChatGPT, where it achieves a 13.2% recommendation coverage rate and a 3.8% rank-one rate. The clearest opportunity is expanding the public evidence layer to increase mention presence while protecting the clean sentiment profile already in place.

Who This Report Is For

This report is for Marcus by Goldman Sachs marketing, digital strategy, and brand leadership teams evaluating the brand's current AI recommendation position and identifying the fastest path to competitive visibility in AI-led banking discovery.

Report Card

• Report type: AI Company Market Strategy Report
• Target company: Marcus by Goldman Sachs
• Category / market studied: Best Banks
• Reporting month: June 2026
• AI platforms tracked: ChatGPT, Copilot, Gemini, Google AI Mode, Google AI Overviews, Perplexity
• Public high-intent clusters: 3 (Best Bank & Account Discovery, Bank Comparison & Alternatives, Bank Pricing, Fees & Rates Research)
• AI observations analyzed: 1,536
• Competitors tracked: Ally Bank, Bank of America, Capital One, Chase, Citibank, Discover Bank, PNC Bank, U.S. Bank, Wells Fargo

Executive Summary

Marcus by Goldman Sachs enters the June 2026 Best Banks benchmark with a clear and specific profile: positive framing, zero negative mentions, and very low mention presence. The brand appears in only 12.6% of all AI observations, the lowest presence rate among the ten measured banks. Its monthly AI Authority Value of $209,793 places it ninth out of ten, ahead of only PNC Bank. These two facts together define the strategic challenge for Marcus in AI-led discovery.

The positive signal is real. When Marcus appears in AI responses, it is framed well. Its net sentiment score of 0.5876 reflects strong positive framing across all platforms, and zero negative mentions were recorded across all 1,536 observations. Its valid recommendation coverage rate of 5.2% means that when Marcus is mentioned, it earns a recommendation roughly half the time, a conversion rate that is competitive with several larger banks operating at higher mention volumes.

The structural problem is not framing. It is volume. Marcus lacks the source footprint to surface in enough AI responses to accumulate meaningful recommendation value. Its strongest platform is ChatGPT, where it achieves a 13.2% recommendation coverage rate and a 3.8% rank-one rate, but this is built on only 56 mentions out of 266 observations on that platform. On Gemini, Marcus appears in just 10.6% of responses with a 4.5% recommendation coverage rate and zero rank-one appearances. On Google AI Mode, the brand surfaces in only 7.1% of responses.

The category leaders are operating at a different scale entirely. Capital One appears in 57.9% of observations. Ally Bank appears in 46.5%. Both carry recommendation coverage rates above 27%. Marcus is not yet competing on recommendation architecture. It is competing for basic inclusion in AI responses, and the benchmark data shows how far that gap currently runs. The path forward runs through the source footprint and citation layer, not through sentiment correction.

What Marcus by Goldman Sachs Is Winning

Cleanest sentiment profile among challenger brands. Marcus recorded zero negative mentions across all platforms and all three clusters. Its net sentiment score of 0.5876 is the fourth highest in the category, behind only Ally Bank, Discover Bank, and Capital One. This is not a marginal advantage. In a category where Wells Fargo carries a 5.2% negative visibility rate and Chase carries a 4% negative rate, a zero-negative profile is a meaningful strategic asset. When AI systems surface Marcus, they surface it in a positive context.

Strongest single-platform performance on ChatGPT. Marcus achieves its highest recommendation coverage rate on ChatGPT at 13.2%, with a 3.8% rank-one rate and an average recommended rank of 2.93. This is the only platform where Marcus achieves double-digit recommendation coverage, and the pattern suggests that ChatGPT's source retrieval behavior is currently more favorable to the content Marcus has available in the public evidence layer.

Zero negative visibility across all three clusters. In the Best Bank & Account Discovery cluster, the Bank Comparison & Alternatives cluster, and the Bank Pricing, Fees & Rates Research cluster, Marcus recorded zero negative mentions. This is a rare signal in a competitive retail banking category where trust, fees, and customer experience framing often generates negative AI citation.

Where Marcus by Goldman Sachs Has the Clearest AI Visibility Gaps

Lowest mention presence in the category. Marcus appears in only 12.6% of all observations, compared to a category-wide pattern where most competitors appear in 25% or more. Bank of America appears in 68.6% of observations. Capital One appears in 57.9%. Even Citibank, which ranks eighth in presence, appears in 25.3% of observations, nearly double the Marcus rate. This is not a positioning problem. It is a retrievability problem.

Near-zero visibility on Gemini and Google AI Mode. On Gemini, Marcus appears in only 10.6% of responses with a 4.5% recommendation coverage rate and zero rank-one appearances. On Google AI Mode, Marcus appears in just 7.1% of responses with a 2.4% recommendation coverage rate. These are two platforms with significant commercial reach in banking discovery queries, and Marcus's presence in both is effectively minimal.

Weak captured value in the highest-intent cluster. In the Bank Pricing, Fees & Rates Research cluster, which carries the highest per-observation commercial value in the benchmark, Marcus captures only $41,667 in monthly AI Authority Value. That figure is less than 13% of Ally Bank's captured value in the same cluster and less than 14% of Capital One's. For a brand whose primary product differentiation is savings rates, weak visibility in rate-intent queries represents the clearest mismatch between brand positioning and AI recommendation behavior.

Competitor displacement is consistent across clusters. In every cluster, Marcus is displaced by Capital One, Ally Bank, and Discover Bank as the primary recommendation. In the Best Bank & Account Discovery cluster, Capital One captures $671,827 in monthly AI Authority Value compared to Marcus's $110,519. In the Bank Comparison & Alternatives cluster, Ally Bank captures $494,897 compared to Marcus's $57,606. Displacement at this scale reflects a source footprint gap, not a product quality or trust gap.

Biggest Opportunity

Marcus's most direct path from its current position to meaningful recommendation value runs through the public evidence layer. The brand already has the positive framing that competitive AI recommendation requires. What it lacks is the volume of retrievable, citable content that causes AI systems to surface it in the first place. Building a structured presence across comparison articles, review platforms, rate tracking publications, and owned content designed for AI extractability would increase mention presence without introducing the sentiment risk that larger banks currently carry. This is not a brand repositioning exercise. It is a citation architecture buildout targeted at the specific clusters and platforms where Marcus is currently absent.

Prompt Evidence

ChatGPT / Best Bank & Account Discovery Prompt: "What is the best bank for a high-yield savings account?" Result: Marcus by Goldman Sachs was mentioned in a positive context but did not receive the primary recommendation, appearing behind Capital One and Ally Bank in the response ordering.

Gemini / Bank Pricing, Fees & Rates Research Prompt: "Which bank has the best savings account interest rates?" Result: Marcus by Goldman Sachs did not appear in the response. Capital One and Ally Bank were the primary recommendations, with Discover Bank appearing as an additional option.

ChatGPT / Bank Comparison & Alternatives Prompt: "Compare Marcus by Goldman Sachs vs Ally Bank for savings accounts." Result: Marcus was mentioned positively but was positioned as an alternative to Ally Bank rather than the primary recommendation, reflecting the category-wide displacement pattern.

Perplexity / Best Bank & Account Discovery Prompt: "What are the best online banks for 2026?" Result: Marcus appeared in a list format with positive framing but was ranked lower than Capital One, Ally Bank, and Discover Bank, consistent with its below-average mention presence rate across the benchmark.

What CiteWorks Studio Would Do Next

Phase 1: AI Market Discovery Audit Map the full prompt-level response data across all six platforms to identify precisely which prompts surface Marcus, which prompts displace it entirely, and which third-party sources are currently driving the mentions that do appear.

Phase 2: Recommendation Readiness Plan Identify the specific content and source gaps preventing Marcus from appearing in AI responses, including missing comparison articles, thin review platform presence, and owned content that is not structured for AI extractability.

Phase 3: Owned Answer Layer Buildout Structure Marcus's official product pages, rate information, and account feature content for AI retrievability, ensuring that AI systems can cite accurate, positive, and current information directly from owned sources.

Phase 4: Citation / Authority Layer Development Build third-party validation signals through comparison articles, independent review platforms, and financial media coverage that AI systems can retrieve and synthesize into recommendation credit, prioritizing the Bank Pricing, Fees & Rates Research cluster where the brand's product advantage is clearest.

Phase 5: Monthly AI Visibility and Recommendation Tracking Track mention presence, recommendation coverage, rank position, and sentiment across all platforms and clusters month over month to measure citation architecture progress and adjust the source strategy.

Why This Matters

Marcus by Goldman Sachs has the positive framing that AI systems consistently reward, but it is not visible in the AI responses that shape most consumer banking decisions today. In a category where Capital One and Ally Bank appear in more than half of all AI responses, Marcus appears in roughly one in eight. The commercial cost of this gap is not only lost recommendation value in the benchmark. It is lost consideration at the moment when consumers are forming their initial shortlist, before they visit a comparison site or request a rate table.

AI presence alone is not a sufficient outcome, but absence at this stage of the discovery funnel is the more immediate risk. Marcus has the sentiment profile that could convert visibility into recommendation power. The structural gap is the source footprint that gets the brand into the AI response in the first place. The strategic priority is building the citation architecture that makes Marcus retrievable across the clusters and platforms where the brand currently does not appear.

Core Metrics

• Mentions: 194 out of 1,536 observations
• Valid recommendations: 80
• Top 3 recommendation count: 38
• Rank #1 recommendation count: 22
• Average recommended rank: 3.35
• Positive mentions: 114
• Neutral mentions: 80
• Negative mentions: 0
• Raw mention presence rate: 12.6%
• Valid recommendation coverage: 5.2%
• Top 3 recommendation rate: 2.5%
• Rank #1 recommendation rate: 1.4%
• Strongest cluster by recommendation behavior: Best Bank & Account Discovery
• Strongest platform by recommendation behavior: ChatGPT

Sentiment Score

Sentiment Score = (114 positive x 1) + (80 neutral x 0) + (0 negative x -1) / 194 total mentions = 0.5876

This score reflects that when Marcus appears in AI responses, it is framed positively in roughly 59% of those appearances. The remaining 41% of mentions are neutral. There are zero negative mentions. This is a genuinely clean sentiment profile, but it is built on a small base of 194 total mentions across 1,536 observations.

The reason sentiment classification matters here is that unclassified mention counts would make the situation look better than it is. If all 194 mentions were counted as equivalent wins, the analysis would obscure the fact that 80 of them carry no recommendation credit. Share of voice is a diagnostic metric, not a business outcome. A positive recommendation, a neutral factual reference, and a competitor-displaced mention are not interchangeable. Counting them as equal overstates Marcus's AI recommendation position by a significant margin. The classification layer is what separates an accurate read of the brand's situation from a misleading one.

Sentiment by Platform

Methodology

1. Market studied: Best Banks, covering retail banking, online banking, savings accounts, checking accounts, and banking services in the United States.
2. Brands included in the benchmark: Ally Bank, Bank of America, Capital One, Chase, Citibank, Discover Bank, Marcus by Goldman Sachs, PNC Bank, U.S. Bank, and Wells Fargo. This is not a full market census. Additional brands operating in the category were not measured.
3. Data collection window: June 2026, with observations generated on June 17, 2026. This report reflects a point-in-time benchmark.
4. AI platforms tested: ChatGPT, Copilot, Gemini, Google AI Mode, Google AI Overviews, and Perplexity.
5. Observation count: 1,536 total AI observations analyzed across all platforms and clusters. Unique prompt count was not available in the public dataset for this report.
6. Prompt clusters: Three public high-intent clusters were analyzed: Best Bank & Account Discovery (consideration stage), Bank Comparison & Alternatives (evaluation stage), and Bank Pricing, Fees & Rates Research (decision stage).
7. Definition of a mention: A mention is recorded when a company name or brand appears in an AI-generated response, regardless of framing, sentiment, or recommendation status.
8. Definition of a valid recommendation: A valid recommendation is a positive, shortlist-quality mention or ranked appearance that earns recommendation credit in the dataset. A mention in a cautionary context, a neutral list appearance, or a competitor-anchored reference does not qualify as a valid recommendation. This distinction is the foundation of the AI Authority Value calculation.
9. Metrics used: Raw mention presence rate, valid recommendation coverage, top-three recommendation rate, rank-one recommendation rate, average recommended rank, net sentiment score, monthly AI Authority Value, and captured share of AI opportunity. Modeled value figures are benchmark estimates based on commercial intent modeling and are not revenue, pipeline, or bookings.
10. Limitations: AI-generated outputs change with model updates, source indexing shifts, and content availability changes. The benchmark reflects one point in time and should not be read as a permanent or guaranteed state. Modeled value figures represent benchmark estimates, not verified business outcomes. Company comparisons are limited to the ten brands included in this benchmark run.

See How AI Is Recommending Your Brand

The benchmark shows where the category stands in June 2026, but it does not show the full picture for Marcus specifically. A brand-level AI Authority Index report would identify which prompts surface Marcus and which ones displace it, which platforms are underrecognizing the brand relative to its product strength, which third-party sources are shaping the recommendations that do appear, and what changes to the owned and citation layers may improve shortlist eligibility. If you want to see where Marcus stands in AI-generated responses right now, and where competitors are being recommended instead, CiteWorks Studio can map that for you.