Dataset · V 2026.1.0 ← Read the benchmark article

2026 Fintech AI Visibility Benchmark -- Dataset

AI visibility scores for 50 fintech and payments companies across ChatGPT, Gemini, Claude, and Perplexity. This is the full dataset for the benchmark article -- all scores, all companies, all verticals.

50Companies
8Verticals
4Engines
16Queries
64Total Runs

CSV contains company, vertical, V score, tier, query appearances, A%, E%. Collection date: June 2, 2026. License: CC BY 4.0.

How to Cite This Dataset
DIGI CONVO. (2026). 2026 Fintech AI Visibility Benchmark Dataset.
Version 2026.1.0. Published June 2, 2026.
https://digiconvo.ai/research/datasets/fintech-ai-visibility-benchmark-2026
License: CC BY 4.0

Key Findings

8 / 50 Highly Visible

Companies appeared consistently across 3 or more engines. Stripe (85), Wise (85), Stripe Connect/Treasury (85), Adyen (80), nCino (80), Kyriba (80), Unit (75), Mambu (75).

8 / 50 Invisible

Zero appearances across all queries and all four engines. Rapyd, Remitly, WorldFirst, i2c, Sift, Featurespace, Provenir, Peach Finance.

28 / 50 Gray Zone

Appeared in some engines, for some queries -- but not consistently. This is the dominant condition in fintech AI visibility.

39 pts Biggest swing

One company scored 31 on generic queries and 70 on matched-framing queries. Same engines, different question, different outcome.

Tier Distribution -- 2026 Fintech AI Visibility Benchmark Highly Visible 16% · 8 companies Partially Visible 56% · 28 Low Visibility 12% · 6 Invisible 16% · 8 Source: DIGI CONVO 2026 Fintech AI Visibility Benchmark · digiconvo.ai/research/fintech-ai-visibility-benchmark-2026

Complete Score Table

Loading...
Company Vertical V Score Tier Query Appearances (of 8) A% E%

Stripe appears in two verticals: Stripe Connect/Treasury (Embedded Finance / BaaS) and Stripe (Payments Processing). These are scored separately because they compete in distinct buyer contexts.

Benchmark collected: June 2, 2026. V = 0.40·A + 0.25·E + 0.20·D. Maximum score approximately 85 under 2026 methodology. A% = Appearance Rate. E% = Engine Breadth.

** PayQuicker is a current DIGI CONVO client. This benchmark applies identical methodology to PayQuicker as to all 50 companies. No score adjustments were made. Results are reported as measured.

What Surprised Us

Four findings from the data that were not predicted before collection.

FindingWhy it matters
Remitly scored 0Strong consumer brand; zero enterprise AI shortlist presence
Finxact dominated Q9, vanished from Q10The clearest single framing boundary in the dataset
ComplyAdvantage and Socure owned completely different queriesSpecialist framing appears more powerful than broad coverage
One company moved 39 points on a query wording changeQuery alignment may matter more than domain authority

Scoring Methodology

V = 0.40·A + 0.25·E + 0.20·D

ComponentSymbolWeightWhat it measures
Appearance RateA40%% of a vertical's 8 possible runs where the company was named
Engine BreadthE25%How many of the 4 engines cited the company in at least one query
Query DepthD20%D receives full credit once a company appears in at least one relevant benchmark query

Under the 2026 baseline model, the maximum achievable score is approximately 85. The remaining 15% is reserved for Citation Strength and Position Score components planned for the 2027 benchmark.

2 queries per vertical × 8 verticals = 16 total queries. All queries unassisted -- brand names excluded from the prompt.

Prompt template: "List the leading [query text] solutions. For each, briefly explain why you would recommend it and what distinguishes it from competitors."

Q# Vertical Query
1Embedded Finance / BaaSbest embedded finance platform for building card programs
2Embedded Finance / BaaSbest unified embedded finance platform
3Payments Processingbest payment processor for SaaS companies
4Payments Processingbest global payment processor for enterprise
5Payout / Disbursementbest payout platform for direct selling companies
6Payout / Disbursementbest global payout orchestration platform
7Cross-Border Paymentsbest cross-border payments platform for businesses
8Cross-Border Paymentsbest B2B international payment API
9Banking Infrastructurebest cloud-native core banking platform
10Banking Infrastructurebest banking-as-a-service infrastructure for neobanks
11Risk / Fraud / AMLbest AML compliance platform for fintech
12Risk / Fraud / AMLbest identity decisioning platform for digital banking
13Credit / Loan Originationbest loan origination platform for banks
14Credit / Loan Originationbest digital lending platform for SMB lenders
15Treasury / Cash Managementbest treasury management system for midmarket companies
16Treasury / Cash Managementbest API-first cash management platform
  • Engines: ChatGPT, Gemini, Claude, Perplexity
  • Session type: Private / incognito browsing, logged-out
  • Location: US-based IP address
  • Collection date: June 2, 2026
  • Method: Responses captured manually and reviewed for company mentions using a standardized scoring rubric
  • Query type: All unassisted -- brand names excluded from all prompts
  • Total runs: 16 queries × 4 engines = 64 runs

This benchmark uses 2 queries per vertical. The 2027 benchmark will expand to 10 queries per vertical for broader buyer-intent coverage. Scores reflect a single-day collection window. AI engine responses are non-deterministic and change over time. This is a dated snapshot, not a permanent ranking.

Planned for 2027

  • Position Score -- first vs. buried in response
  • Citation Authority -- trade press vs. vendor blog weighting
  • 10 queries per vertical
  • Multi-week collection window

Downloads

Score Dataset (CSV) 50 companies. V Score, Tier, Query Appearances, A%, E%. UTF-8.
CC BY 4.0 Download CSV
Score Dataset (JSON) Same data in JSON format. Preferred by developers and AI systems.
Browser-generated
Query Log All 16 queries, 4 engines, returned companies per run.
Raw runs data

Version History

VersionDateNotes
2026.1.0 June 2, 2026 Initial release. 50 companies, 8 verticals, 16 queries, 2 queries per vertical.

Update Policy

This benchmark is updated annually. The 2027 edition will expand to 10 queries per vertical, add Position Score and Citation Authority components, and use a multi-week collection window. A methodology note will be published if AI engine behavior shifts significantly before the next annual update.