The machines run one financial set, and the category is drawn as three.

AI models have a clear picture of financial services. Every brand ARA measured scores at least 16 on clarity. The machines know what these companies do. The disagreement is about who belongs beside whom, and the study and the models do not answer that the same way.

ARA drew this category as three segments. Payments and networks. Asset management and alternatives. Market infrastructure and data. The models were never told that. They answer 15 buyer questions and name whoever they name, and the brands they place in the same answer are the set they actually compare.

Those two pictures disagree. BlackRock's strongest measured pairing is Visa. So is S&P Global's. So is State Street's. So is Blackstone's. Only three pairs hold inside a declared segment: Goldman Sachs with Morgan Stanley, Fidelity with Vanguard, and Bloomberg with S&P Global. The models run one large financial set with the payment networks at the centre of it.

The median score is 69. The spread runs from 79 to 54. No brand reaches the top tier. This is the first reading of financial services, so it sets the baseline. The next reading measures what changed against it.

ABOUT THIS READING

ARA scores each brand on five attributes out of 20 each: Findability, Clarity, Sentiment, Identity Fidelity and AI Recommendation. The total falls into one of five tiers. AWESOME is 83 to 100. STRONG is 70 to 82. AVERAGE is 56 to 69. WEAK is 40 to 55. INVISIBLE is 0 to 39.

ARA put 15 buyer questions to four AI models. That produces 60 recommendation moments per brand. A position label sits beside each score. SWELL means the brand holds real equity the machines under-reward. EBB means the machines recommend the brand above its equity. ALIGNED means the two agree.

The Winners.

Visa leads the category at 79. All four models choose Visa. It appears in 17 of the 60 recommendation moments, or 28%, more than any other brand. It takes first place six times. Models name Visa for the most trusted name, for the next ten years and for the crisis question. Visa is absent when buyers ask who to trust with their money.

Mastercard holds the strongest single attribute in the cohort. It scores 16 on AI recommendation against a cohort average of 8.1. All four models choose it, and it appears in 15 of the 60 moments, or 25%. That is presence, not primacy. Mastercard takes first place once. Models name it for innovation, for the next ten years and for the question about who is furthest ahead in AI.

The Losers.

Blackstone sits at the bottom at 54 and is the only brand in the weak tier. One model of four chooses it. It appears once in 60 moments, or 2%, and wins no question. The machines still describe it well, with a clarity score of 17. Blackstone is absent from the most trusted name, from trust with money, from the crisis question and from the next ten years.

Venmo carries the weakest attribute on the board. It scores 0 on AI recommendation against a cohort average of 8.1. No model chooses it in any of the 60 moments. The fundamentals say something different. Venmo scores 15 on findability, 18 on clarity and 16 on identity fidelity, for a total of 62. The machines know Venmo and never recommend it.

The Leaderboard.

The shelf is crowded and the biggest single holding sits outside the measured set

The four models named 131 brands across 280 choices. The 17 measured brands hold 32.5% of that shelf. JPMorgan Chase holds the largest single share at 18 choices, or 6.4%. Bank of America and USAA hold six each. Capital One, Chase and Navy Federal Credit Union hold five each. Moody's holds four and BBVA holds three.

The tail is long. 84 brands are named exactly once and hold 30% of all choices. 17 brands are named exactly twice and hold 12%. Together those 101 brands hold 42% of the shelf. No single rival owns this category in the machines' answers.

That shape sets the opportunity. A brand that earns a repeatable place in these answers takes share from a scattered tail. The toggle below shows the cohort view and the full shelf. The distance between the two views is the category's real competitive set.

15 buyer questions x 4 models / audited brands in colour, every other brand in a stepped grey. Toggle between Cohort Only and Full Shelf.

KEY FINDING 01

Ten of the 17 brands hold more equity than the machines reward.

Ten brands carry the SWELL position. The machines describe these brands accurately and then choose someone else. Six brands are aligned. One brand, Goldman Sachs, is recommended above its equity.

BlackRock shows the pattern at full size. It holds the highest findability score in the cohort at 16, and scores 18 on clarity. Its AI recommendation score is 6. It is named in 6 of 60 moments and takes first place zero times. Morgan Stanley repeats it with 15 on findability, 18 on clarity and 7 on recommendation.

American Express is the cleanest case of the gap. It scores 18 on clarity, 18 on identity fidelity and 16 on sentiment, for a total of 74 and a place in the strong tier. Its recommendation score is 7 and models name it in 4 of 60 moments. State Street and PayPal score 2 each on recommendation. The fundamentals are already paid for in each of these cases.

Brand Fundamentals vs Recommendation.

KEY FINDING 02

Stripe and American Express score highest on identity while every brand scores high on clarity.

Clarity scores run from 16 to 19 across all 17 brands. That attribute does not separate anyone. Identity fidelity runs from 11 to 18. That attribute does.

Stripe holds the top clarity score at 19 and an identity score of 18. American Express matches the 18 on identity. Bloomberg follows at 17. These brands keep their own language when a model describes them.

Fidelity sits at the other end with 16 on clarity and 11 on identity fidelity. Visa scores 18 on clarity and 13 on identity. A brand can be understood perfectly and still sound like its category rather than itself. The chart below plots both scores for every brand.

Understood vs Distinct.

KEY FINDING 03

Venmo and Nasdaq both score zero, and the zeros mean opposite things.

Two brands are never named. Venmo and Nasdaq do not appear in any of the 60 recommendation moments, at any rank. Neither brand is named lower than a rival. Neither brand is named at all. The reason differs for each.

Venmo's five payment peers are named on ten questions that Venmo never enters. Visa, Mastercard, American Express, Stripe and PayPal each hold ground Venmo does not. Nasdaq's position is the reverse. On eleven of the fifteen questions no model names a market infrastructure company. Nasdaq has no peer in those answers to be measured against.

Absence and low rank are different problems with different repairs. So are these two absences. Venmo is missing from answers its own peers hold, and a published record can change that. Nasdaq is missing from answers no company like it reaches, and publishing does not change that. Venmo totals 62 and Nasdaq totals 57. Only one of them has a way in from here.

KEY FINDING 04

Thirteen brands can publish their way forward. Four cannot.

For every question a brand is missing from, one test decides what to do about it. Do other brands in the same position hold that answer? If they do, the gap belongs to the brand, and evidence closes it. If no brand like it appears at all, the gap belongs to the position, and no amount of publishing closes it.

Run that test across the cohort and the category splits three ways. Payments and networks reaches 12 of the 15 questions. Asset management and alternatives reaches 11. Market infrastructure and data reaches 4. For Bloomberg, S&P Global, Nasdaq and State Street, eleven of the fifteen questions return no company like them at any position.

S&P Global is the clearest case in the cohort. It holds four questions and has no brand problem anywhere. It already holds every answer its position allows, and it scores 13 of 20. That is not underperformance. That is the ceiling of what a market data company reaches in this battery. The only way past it is to be understood as something else.

KEY FINDING 05

Visa takes first place six times and no other brand takes more than three.

Presence and primacy are separate contests. Visa wins both. It appears in 17 of 60 moments and takes first place six times. Stripe takes first place three times from 6 namings. Charles Schwab takes it twice from 5 namings.

Mastercard shows the split most clearly. It scores 16 on AI recommendation, the strongest attribute score in the cohort, and appears in 15 of 60 moments. It takes first place once. A high recommendation score measures presence and rank across many questions. It is not a count of first answers.

Goldman Sachs is named in 10 of 60 moments and takes first place twice. It holds the only EBB position on the board. The machines recommend Goldman Sachs above the equity its other four attributes support.

KEY FINDING 06

The segment that cannot grow is the only one where the brands collide.

ARA read every brand's plan for the year ahead against the same fifteen questions. Across all 17, only two questions are contested by more than one brand. S&P Global and State Street both aim at which company sets the standard. Bloomberg and Nasdaq both aim at trusted research and data.

All four are market infrastructure brands. The segment reaches four questions, and its four brands are now pointed at the same small ground. The thirteen brands with room to move are not competing with each other for it. The four without room are.

KEY FINDING 07

The two most named brands are both absent when buyers ask who to trust with money.

Visa and Mastercard lead the category on namings. Both are absent from the same question. Neither appears when a buyer asks who to trust with their money. Neither appears on the fraud question or on the research and data question.

The models name these two brands for capability. Visa wins the most trusted name, the next ten years, the technology question, the crisis question and the question about what is hardest to replace. Mastercard wins innovation, the crisis question, the next ten years and the AI question.

Blackstone is absent from the trust questions as well, and from the crisis question and the next ten years. These questions are not open. Four measured brands already hold the trust-with-money question: American Express, Vanguard, Fidelity and Charles Schwab. Vanguard also holds the fairness question. Visa and Mastercard are absent from ground their own cohort occupies.

WHAT THIS MEANS

Fundamentals are not the constraint in financial services. All 17 brands score at least 16 on clarity, and the median total is 69. Recommendation is the constraint. The cohort averages 8.1 on that attribute while Visa scores 19. Ten brands hold more equity than the machines return to them.

The shelf gives that gap a shape. The 17 measured brands hold 91 of 280 choices. The other 189 go to 114 companies outside the set, led by JPMorgan Chase with 18. Five of the six largest holdings outside the set take deposits, and the category boundary excludes banks by design. That boundary is the right one, and this is its price on the shelf.

For a brand in this category the first question is not how to rank higher. It is which kind of absence it has. Sixteen of the seventeen miss at least one answer their own peers hold. Evidence moves those. Four sit in a segment that reaches only four of the fifteen questions. Only a change in what the brand is understood to be moves those. The two need different plans, different owners and different budgets. This reading sets the baseline for both.