Olivia Moore & Marc Andrusko & Seema Amble — a16z - The AI Application Spending Report: Where Startup Dollars Really Go - October 2025 logo

Olivia Moore & Marc Andrusko & Seema Amble — a16z - The AI Application Spending Report: Where Startup Dollars Really Go - October 2025

Free

Where Startup Dollars Really Go

FreeFree tier
Type
Open Source
Founded
2009
Company
Andreessen Horowitz (a16z)

About Olivia Moore & Marc Andrusko & Seema Amble — a16z - The AI Application Spending Report: Where Startup Dollars Really Go - October 2025

The AI Application Spending Report by a16z partners Olivia Moore, Marc Andrusko, and Seema Amble analyzes spending patterns across over 200,000 Mercury customers from June to August 2025. It identifies the top 50 AI-native application-layer companies based on actual spend (ACH, card, wires), excluding infrastructure and cloud services. The report reveals that horizontal applications (e.g., LLM assistants, meeting notetakers) claim 60% of the list, while vertical tools serve specific roles. It provides real-time signals of where early-stage startups are investing in AI, highlighting trends like vibe coding platforms, creative tools, and customer service solutions.

Key Features

Ranking of top 50 AI-native application companies by spend data
Analysis of horizontal vs. vertical AI application categories
Insights from over 200,000 Mercury customers over June–August 2025
Identification of emerging categories like meeting notetakers and vibe coding platforms
Comparison of AI spend ROI against traditional software tools

Pros & Cons

Pros
  • Data-driven ranking based on actual spend, not web traffic or hype
  • Covers both horizontal and vertical AI applications with clear separation
  • Provides real-time signal of startup buying behavior in AI
  • Includes methodology notes for transparency
Cons
  • Data limited to Mercury customers, may not represent the entire startup ecosystem
  • Only captures spend via Mercury (ACH, card, wires), missing other payment methods
  • Time window is only three months (June–August 2025), may miss seasonal effects
  • Does not include infrastructure or cloud providers, focusing solely on application layer

Best For

Investors seeking real-time data on AI startup adoptionFounders benchmarking their company against top AI-native peersProduct managers understanding which AI applications gain early tractionAnalysts tracking AI spending trends in the startup ecosystem

FAQ

What is the AI Application Spending Report?
It is a report by a16z partners that analyzes spending data from over 200,000 Mercury customers to rank the top 50 AI-native application-layer companies and identify trends in where startup dollars are going in AI.
How was the data collected?
The data was collected from Mercury's database of over 200,000 customers, looking at spend patterns from June to August 2025 via ACH, IO card spend, and wires. It does not include Mercury Personal customer or reimbursement spend.
What types of companies are included?
The report includes AI-native application-layer companies, excluding those that primarily sell cloud services (e.g., Azure), GPUs (e.g., CoreWeave), or infrastructure tools. Both horizontal (60%) and vertical (40%) applications are covered.
What are the top findings from the report?
Horizontal applications like general LLM assistants (OpenAI, Anthropic, Perplexity) and meeting notetakers (Fyxer, Otter AI) dominate spend. The report also notes that the race for general assistants is not won, and AI is turning specific roles into broadly deployed skills.