Skip to content
>_ StartupJobList_

Visa Sponsorship Signals · Hacker News hiring

Visa support language became less common in HN hiring posts

Language indicating sponsorship or transfer support appeared in 670 of 8,000 eligible HN hiring posts (8.4%) in 2019 and 86 of 3,786 (2.3%) in 2025. This includes conditional support and transfer-only language, not guaranteed new sponsorship.

A statement in a hiring post does not establish eligibility or the likelihood of obtaining a visa.

How support language changed over time

The annual share counts posts with sponsorship or transfer support language, including conditional offers. Every eligible hiring post stays in the denominator, including posts that say nothing about support.

Sponsorship or transfer support

Share of eligible HN hiring posts, from 0 to 10%. A labeled gap is not zero. Read the exact numerator, denominator and interval in the evidence table.

2011 is partial. The isolated 2008 observation is excluded. Recent incomplete years are labeled by their actual window. Descriptive intervals in the table do not capture errors in interpreting posts or changes in who advertises on HN.

The last 3 consecutive year-to-year comparisons through 2025 each showed a lower support-language share. That describes the recent annual sequence; it does not make the whole history a straight line or explain why it changed.

Separate the share from the size of the channel

Fewer mentions, or fewer hiring posts?

From 2019 to 2025, posts mentioning support changed from 670 to 86. The total number of eligible HN hiring posts changed from 8,000 to 3,786. Read the two changes together: a smaller number of support mentions and a smaller share answer different questions.

87.2% fewerposts mentioning sponsorship or transfer support
52.7% fewereligible hiring posts overall

The support-language share changed by 6.1 percentage points down, or 72.9% relative to its 2019 level. Those are changes in advertised language, not changes in visa availability or hiring success.

Support mentions and all hiring posts, 2019 = 100

Both series start at the same index value. The common scale runs from 0 to 100; the labels identify each series without relying on color. This compares changes in counts, not the share of posts offering support.

2019 · 670 support posts / 8,000 eligible posts

Support mentions100
All hiring posts100

2022 · 309 support posts / 7,201 eligible posts

Support mentions46.1
All hiring posts90

2025 · 86 support posts / 3,786 eligible posts

Support mentions12.8
All hiring posts47.3
Index = the period’s post count divided by its 2019 count, multiplied by 100. Repeated employers can contribute in several months. These are not counts of unique employers.

The figures cannot tell us how much of the movement came from different employers, roles or work locations appearing on HN. They also cannot separate a change in wording from a change in actual support. No adjustment for those differences is implied.

Three stated benchmarks, with absolute and relative changes kept separate
ComparisonSupport-post changeAll-post changeShare change (pp)Relative share change
2019 to 2022-53.88%-9.99%-4.08 pp-48.76%
2022 to 2025-72.17%-47.42%-2.02 pp-47.06%
2019 to 2025-87.16%-52.68%-6.1 pp-72.88%

The unanswered question

Does less support language mean more refusal?

No. Cases where neither support nor refusal was established are different from explicit refusal. Posts without a clear statement cannot be counted as refusing support. The table keeps both visible instead of distributing unclear posts between yes and no.

Not-established and explicit-refusal shares at the same annual benchmarks
YearNot establishedExplicit refusal
20197,261 / 8,000 · 90.8%Not estimable
20226,838 / 7,201 · 95%Not estimable
20253,639 / 3,786 · 96.1%Not estimable

We cannot give a full-year refusal total where monthly figures are incomplete. The apparent balance between support and refusal must not be inferred from the not-established share. For an individual application, an unstated policy remains a question to ask.

Compare the same months

Language indicating sponsorship or transfer support appeared in 49 of 1,854 eligible HN hiring posts (2.6%) in 2025 H1 and 54 of 1,831 (2.9%) in 2026 H1. This includes conditional support and transfer-only language, not guaranteed new sponsorship. Provisional months are excluded from this comparison.

The difference is 0.31 percentage points higher in 2026 H1. That is less than one percentage point. A matched six-month comparison does not establish a lasting trend or a change in an applicant’s eligibility.

Support, refusal and not established in the latest matched window

2025 H1 · 1,854 eligible posts

Support2.6%
RefusalNot estimable
Not established96.4%

2026 H1 · 1,831 eligible posts

Support2.9%
Refusal2.2%
Not established94.9%
All bars use the same 0–100% scale. These are post shares, not counts of vacancies, candidates or visas.

Use the signal, then clarify the offer

Three questions before you invest time in an application

For candidates who need employer support, the advertisement is the start of a conversation. The category in this report cannot tell you whether a particular role fits your situation.

  1. What kind of support?

    Does the employer mean new sponsorship, an existing visa transfer, or support only under stated conditions?

  2. Which role does it cover?

    A post may advertise several openings. Ask whether the statement applies to the role and work location you are considering.

  3. What is still unspecified?

    If the post says nothing about support, ask directly. Do not read the not-established category as refusal, or a positive statement as a guarantee.

For employers: answer these questions in the advertisement so candidates do not have to infer the policy. This is a checklist for reading recruiting language, not immigration advice.

Counts behind the charts

Each category shows count / eligible posts, share and a descriptive 95% Wilson interval. Not estimable means the full period cannot be measured from the monthly figures. It does not mean nobody offered or refused support. Low-support cells retain counts but omit intervals.

Annual and matched-period visa-language counts and descriptive intervals
PeriodEligible postsSupportRefusalNot established
2011 Mar–Dec (partial)1,586 Not estimableNot estimableNot estimable
20122,266 Not estimableNot estimable2,200 / 2,266
97.1%

96.3–97.7%
20133,015 Not estimableNot estimable2,923 / 3,015
96.9%

96.3–97.5%
20143,892 Not estimableNot estimable3,777 / 3,892
97%

96.5–97.5%
20156,550 242 / 6,550
3.7%

3.3–4.2%
Not estimable6,274 / 6,550
95.8%

95.3–96.2%
20167,406 429 / 7,406
5.8%

5.3–6.3%
Not estimable6,945 / 7,406
93.8%

93.2–94.3%
20178,921 756 / 8,921
8.5%

7.9–9.1%
Not estimable8,112 / 8,921
90.9%

90.3–91.5%
20189,160 819 / 9,160
8.9%

8.4–9.5%
Not estimable8,231 / 9,160
89.9%

89.2–90.5%
20198,000 670 / 8,000
8.4%

7.8–9%
Not estimable7,261 / 8,000
90.8%

90.1–91.4%
20206,750 400 / 6,750
5.9%

5.4–6.5%
Not estimable6,291 / 6,750
93.2%

92.6–93.8%
20219,780 360 / 9,780
3.7%

3.3–4.1%
Not estimable9,341 / 9,780
95.5%

95.1–95.9%
20227,201 309 / 7,201
4.3%

3.8–4.8%
Not estimable6,838 / 7,201
95%

94.4–95.4%
20233,888 153 / 3,888
3.9%

3.4–4.6%
Not estimable3,692 / 3,888
95%

94.2–95.6%
20243,692 128 / 3,692
3.5%

2.9–4.1%
Not estimable3,503 / 3,692
94.9%

94.1–95.5%
20253,786 86 / 3,786
2.3%

1.8–2.8%
Not estimable3,639 / 3,786
96.1%

95.5–96.7%
2026 H11,831 54 / 1,831
2.9%

2.3–3.8%
40 / 1,831
2.2%

1.6–3%
1,737 / 1,831
94.9%

93.8–95.8%
2025 H11,854 49 / 1,854
2.6%

2–3.5%
Not estimable1,787 / 1,854
96.4%

95.4–97.1%
2026 H11,831 54 / 1,831
2.9%

2.3–3.8%
40 / 1,831
2.2%

1.6–3%
1,737 / 1,831
94.9%

93.8–95.8%

Downloads contain only monthly visa-language aggregates: period, category, eligible-post denominator, category count, share, interval and provisional status. Empty CSV cells and JSON nulls mean unavailable, never zero.

Source, definitions and limits

Source: Hacker News “Who is Hiring?” threads. This independent analysis is not endorsed by Hacker News. Eligible posts are top-level hiring posts naming an organization and an opening. A post can contain multiple roles. This is a selected recruiting channel, not a representative survey of employers or software jobs.

Support includes language indicating sponsorship, transfer-only and conditional support. Refusal means explicit unavailability language. Not established means neither support nor refusal was established by the post’s wording. These three categories describe posts, not an employer’s complete policy. No country-specific eligibility, causal policy explanation or visa-success probability is inferred.

Annual shares divide the sum of category counts by the sum of eligible posts, never the average of monthly percentages. A total requires every expected monthly category count. The not-established category is kept separate from explicit refusal. 2011 begins in March; 2008 is not continuous history. Provisional periods: 2026-07, 2026-08.

Descriptive intervals do not include categorization error, repeated employers or selection into HN. Historical vocabulary, negation, conditional and transfer-only language can affect categorization differently across eras. Differences in language are not proof of a change in the wider hiring market. We do not claim company-weighted trends or statistically significant market changes.

Data through . Later posts and edits are not included. The methodology page explains the broader analysis.