State of AI-Native GTM.
The annual reference for go-to-market engineering in AI companies. We analyze 4,682 open roles across 450 companies to map compensation, tech stack, hiring velocity, and geography. Updated live; snapshot available for archives.
Compensation by seniority
USD midpoint of salary bands across live roles, grouped by inferred seniority.
Compensation by role
Median total compensation for common GTM roles in AI companies.
Tech stack adoption
Most-mentioned tools in AI GTM roles, plus the strongest co-occurrences.
- 21% of OpenAI roles also mention Python
- 26% of Python roles also mention AWS
- 33% of Anthropic roles also mention Python
- 79% of AWS roles also mention GCP
- 93% of GCP roles also mention AWS
- 72% of Kubernetes roles also mention Python
- 55% of Salesforce roles also mention OpenAI
- 61% of TypeScript roles also mention Python
Hiring velocity
Top companies by open GTM headcount, hot role titles, and the most active cities.
- Software Engineer256 open
- Enterprise Account Executive118 open
- Senior Software Engineer101 open
- Account Executive98 open
- Staff Software Engineer86 open
- Product Manager82 open
- Sales Development Representative68 open
- Engineering Manager63 open
- Manager63 open
- Forward Deployed Engineer55 open
- San Francisco507 roles
- New York133 roles
- Paris118 roles
- United States106 roles
- London99 roles
- Mountain View68 roles
- Location not specified54 roles
- Hybrid - San Francisco49 roles
- New York City40 roles
- Bangalore39 roles
Geography
Remote, hybrid, and onsite split, plus regional distribution of open roles.
Compensation by region
Median USD midpoint for roles with a disclosed location.
Methodology
This report is built from the SVGTM index of live AI-company job postings. We extract compensation bands, tech tags, locations, and remote status from each posting. Salary numbers are converted to USD where the original currency is known. Medians are computed from midpoint values; p25/p75 bands are shown when sample sizes are large enough. Roles are classified by title keywords, not self-reported level. A single role can mention multiple tools, so tool counts sum to more than the total role count.