Data & API
Use the data
Published We Love SaaS Benchmarks results are open data: free to read, download and reuse with attribution. They are available as web pages, as JSON and CSV files, through a read-only API, and as tools for AI assistants.
What is available?
One set of files per survey round. A round's files exist from the day it opens and say “not yet published” until its results are released.
| Round | Status | Companies (n) | Files |
|---|---|---|---|
| Founder & Executive Compensation 2026Founder & Executive Compensation | Collecting answers | — | JSON · CSV |
| AI Spend & Tooling 2026 H1AI Spend & Tooling | Collecting answers | — | JSON · CSV |
What is the licence?
Creative Commons Attribution 4.0 International (CC BY 4.0). You may share and adapt the published results, including commercially, as long as you credit We Love SaaS.
Cite a result like this, with the number of companies stated on the result itself:
We Love SaaS Benchmarks, <survey name> <year>, n=<companies>.
The licence covers published results only. The sample report shows made-up numbers and must not be cited as data.
What is in the data, and what is left out?
A fixed set of figures per round: every question for all companies together, and every number split by one company characteristic at a time. No figure exists for a group of fewer than 8 companies.
- Numbers are given as the median with the lower and upper quarter (25th and 75th percentile). Averages, minimums and maximums are not published.
- A row marked withheld means the group was too small; it carries no figure and no count.
- Combinations of two or more characteristics are not published. They are available to contributing companies after signing in.
- Individual answers and the list of participating companies are never published.
How do I use the API?
Three read-only endpoints, no key needed, 60 requests a minute. Every response includes a citation and a source_url.
| Request | Returns |
|---|---|
| GET /api/v1/surveys | Every survey and its rounds. |
| GET /api/v1/surveys/{survey} | One survey: its rounds, questions and definitions. |
| GET /api/v1/results/{round} | Published results for one round. Optional filter, one at a time: country, arr_band, headcount, engineering_fte, funding_type. |
Example: https://benchmarks.welovesaas.io/api/v1/results/founder-exec-compensation-2026?country=NL
The full description is an OpenAPI 3.1 document.
How do I connect an AI assistant (MCP)?
Add this address as a remote MCP server (Streamable HTTP) in an assistant that supports the Model Context Protocol. No sign-in is needed.
https://benchmarks.welovesaas.io/mcp
It offers three read-only tools:
- list_surveys — the surveys and their rounds.
- get_survey — one survey's questions and definitions.
- query_benchmark — published results for a round, optionally for one group.
The tools return exactly what the API returns: published results only, nothing for groups under 8 companies, the same rate limit, and a citation and source link with every result.
Is there a summary for AI crawlers?
Yes: /llms.txt describes the site, the method, the surveys and how to cite them.