Between October 5 and November 4, 2026, we tracked 1,323 active software engineering postings on public job boards. When we looked at who posted them, one thing stood out: 833 of them, about 63%, came from just ten companies.
| Company | Active postings |
|---|---|
| Cloudflare | 216 |
| Stripe | 141 |
| MongoDB | 104 |
| Databricks | 103 |
| 58 | |
| Anthropic | 57 |
| Braze | 43 |
| Datadog | 39 |
| OpenAI | 38 |
| 34 |
The other 143 companies in our data posted the remaining 490.
Why a raw count misleads
If you count every posting equally, these ten companies decide what "the market" looks like. Their preferred stack becomes the top skills. Their locations become the top cities. Their seniority mix becomes the level breakdown. A job seeker reading that report learns a lot about ten employers and very little about the other 143.
Large posting counts also don't line up neatly with headcount. Many of these listings are the same role repeated across cities, evergreen postings that stay open year round, or pipeline roles that collect applications ahead of a real opening. We can't tell from the outside which is which, and we don't think anyone can.
What we do instead
Any company with more than 25 active postings in the last 30 days, across all job categories, is treated as a high-volume poster. In each category, all of its postings together count as 5. Cloudflare's 216 software engineering postings count the same as a company that posted 5.
With that weighting, last month's software engineering market comes to 355 weighted postings from 153 companies, and the picture changes:
- The top named employers are Elastic, Fivetran, Moloco, Palantir and Ramp, companies hiring at a pace that matches a real team.
- Python (10% of postings), TypeScript (9%), AWS (8%) and React (7%) lead the skills, measured across all of those employers rather than a handful.
- About 27% of postings say remote. Two thirds don't state a workplace at all, which is a finding in itself.
We still list the high-volume posters on every market page, so nothing is hidden. They just don't get to speak for everyone.
The same rule applies to your analysis
When VulcanCV analyzes your resume, it builds a pool of postings to compare you against. High-volume posters go to the back of that pool, unless you named one of them as a company you want to work for. Then they come first, because that is a real signal from you.
You can see the full software engineering numbers, updated every month, on the software engineering market page.