Skip to content
>_ StartupJobList_

Remote Work Reality · HN hiring

HN remote/hybrid mentions: 70.7% in H1 2026

In January to June 2026, 1,294 of 1,831 eligible Hacker News hiring posts mentioned remote or hybrid work (70.7%). In the same months of 2025, it was 1,356 of 1,854 (73.1%).

This measures advertised language in a selected recruiting channel, not the whole job market. “Remote mentioned” includes hybrid work, geographically restricted remote work and company-wide policies. It does not establish fully remote availability or eligibility from your country.

A difference in ad language, not proof of a market shift. Categorization errors may differ by year. The figures do not establish a change in actual remote availability. The error scenarios below are hypothetical, not measured error estimates.

Use the mention to start a conversation

For a candidate, the useful next step is to check permitted countries, time zones, office days and whether the policy covers the role. For an employer, put those terms in the advertisement. A broad remote label alone cannot answer them.

Inspect the counts and download the series · Read the definitions and limits

How has remote language changed?

Annual shares divide the sum of remote/hybrid mentions by all eligible posts in the same months. Gaps mean a complete numerator is unavailable. Only full calendar years appear in this chart. The matched January–June comparison below carries the current partial year.

Annual share of HN hiring posts mentioning remote or hybrid work Full numerical evidence and unavailable categories appear in the evidence table below. 0%25%50%75%100%2012: 16% 20122013: 16.8% 2014: 20.4% 2015: 19.6% 2016: 22% 2017: 23.3% 2018: 26.7% 20182019: 33.4% 2020: 62.5% 2021: 82.7% 2022: 88.6% 2023: 81.5% 2024: 75.7% 2025: 71.9% 2025 Annual share of HN hiring posts mentioning remote or hybrid work Full numerical evidence and unavailable categories appear in the evidence table below. 0%25%50%75%100%2012: 16% 20122013: 16.8% 2014: 20.4% 20142015: 19.6% 2016: 22% 20162017: 23.3% 2018: 26.7% 20182019: 33.4% 2020: 62.5% 20202021: 82.7% 2022: 88.6% 20222023: 81.5% 2024: 75.7% 20242025: 71.9% 2025
Full calendar years only. Partial 2011 and 2026 are omitted here and remain labeled in the evidence table. The isolated 2008 observation is excluded.

The years before the recent comparison

In 2019, 2,669 of 8,000 posts mentioned remote/hybrid work (33.4%). In 2020, 4,221 of 6,750 posts mentioned remote/hybrid work (62.5%). In 2021, 8,090 of 9,780 posts mentioned remote/hybrid work (82.7%). These adjacent years show the longer change in advertised language. They do not isolate a cause: the advertisers, disclosure conventions and categorization errors can all change over time.

A fixed historical comparison: 2022 and 2025

Remote/hybrid language appeared in 6,381 of 7,201 eligible HN hiring posts in 2022 (88.6%), compared with 2,722 of 3,786 in 2025 (71.9%). Under this categorization, the descriptive difference in advertised language is 16.7 percentage points lower in 2025. These are predefined comparison years, not a search for the highest and lowest values.

Shares and advertising counts answer different questions

A percentage describes the mix of posts. A count describes how much advertising appeared in this channel. Neither measures the number of vacancies: one post can cover several openings, and the same employer can return in several months.

Full years: 2022 to 2025

Remote/hybrid mentions changed by -3,659 posts (-57.3%). Total eligible posts changed by -3,415 (-47.4%). These count changes are distinct from the percentage-point change in share.

Matched first half-years

Remote/hybrid mentions changed by -62 posts (-4.6%). Total eligible posts changed by -23 (-1.2%). These count changes are distinct from the percentage-point change in share.

Full-year eligible posts and remote/hybrid mentionsPaired bars start at zero. Eligible posts on the left; remote mentions on the right. Exact counts are in the evidence table. Missing counts have no bar. 0k5k10k2012 eligible posts: 2,2662012 remote mentions: 36220122013 eligible posts: 3,0152013 remote mentions: 5072014 eligible posts: 3,8922014 remote mentions: 7952015 eligible posts: 6,5502015 remote mentions: 1,2852016 eligible posts: 7,4062016 remote mentions: 1,6332017 eligible posts: 8,9212017 remote mentions: 2,0802018 eligible posts: 9,1602018 remote mentions: 2,44620182019 eligible posts: 8,0002019 remote mentions: 2,6692020 eligible posts: 6,7502020 remote mentions: 4,2212021 eligible posts: 9,7802021 remote mentions: 8,0902022 eligible posts: 7,2012022 remote mentions: 6,3812023 eligible posts: 3,8882023 remote mentions: 3,1702024 eligible posts: 3,6922024 remote mentions: 2,7962025 eligible posts: 3,7862025 remote mentions: 2,7222025Full-year eligible posts and remote/hybrid mentionsPaired bars start at zero. Eligible posts on the left; remote mentions on the right. Exact counts are in the evidence table. Missing counts have no bar. 0k5k10k2012 eligible posts: 2,2662012 remote mentions: 36220122013 eligible posts: 3,0152013 remote mentions: 5072014 eligible posts: 3,8922014 remote mentions: 79520142015 eligible posts: 6,5502015 remote mentions: 1,2852016 eligible posts: 7,4062016 remote mentions: 1,63320162017 eligible posts: 8,9212017 remote mentions: 2,0802018 eligible posts: 9,1602018 remote mentions: 2,44620182019 eligible posts: 8,0002019 remote mentions: 2,6692020 eligible posts: 6,7502020 remote mentions: 4,22120202021 eligible posts: 9,7802021 remote mentions: 8,0902022 eligible posts: 7,2012022 remote mentions: 6,38120222023 eligible posts: 3,8882023 remote mentions: 3,1702024 eligible posts: 3,6922024 remote mentions: 2,79620242025 eligible posts: 3,7862025 remote mentions: 2,7222025
Each pair: eligible posts left, remote/hybrid mentions right. Mobile labels are spaced farther apart; all yearly bars remain. Scale starts at zero; partial years are excluded. Counts can rise or fall as employers change where they advertise, hiring activity changes, or the mix of advertisers changes. These aggregates cannot distinguish those explanations.

An absent policy is not an onsite requirement

The share outside “remote mentioned” contains two different categories: an explicit negative statement and an unknown policy. They should not be combined into an onsite estimate.

Full-year comparison

Unknown policy: 390 (5.4%) in 2022, then 245 (6.5%) in 2025. Explicit not-remote statements: 430 (6%), then 819 (21.6%) in those same periods.

Matched first-half comparison

Unknown policy: 119 (6.4%) in 2025 H1, then 142 (7.8%) in 2026 H1. Explicit not-remote statements: 379 (20.4%), then 395 (21.6%) in those same periods.

Changing disclosure can move these categories without a change in working arrangements. These are separate groups of advertisements, not the same employers tracked through a policy change. Missing category counts stay unavailable and are never reconstructed by subtraction.

Compare the same six months

January to June 2025 and January to June 2026 each contain six complete, nonprovisional months. Matching months avoids comparing half a year with a full year.

Remote/hybrid mention share in matched first half-yearsH1 2025: 73.1 percent. H1 2026: 70.7 percent. Scale zero to one hundred percent. H1 202573.1%H1 202670.7%0%100%
Shares of eligible top-level hiring posts, not shares of jobs, hires or people.

For candidates: read the restrictions

A remote mention is a reason to inspect the advertisement, not a promise you qualify. Check permitted countries, time zones, office attendance, employment arrangements and which specific role the policy covers. Ask whether a company-wide statement applies to the opening you want.

For employers: state the policy clearly

Separate fully remote and hybrid arrangements. Name geographic restrictions and required office days for each role. Unstated terms make the post harder for a candidate to evaluate, but this analysis does not measure application outcomes.

Keep the unknowns visible

Unknown policy accounted for 6.4% of eligible posts in H1 2025 and 7.8% in H1 2026. Unknown posts are not counted as onsite. “Not remote” means an explicit negative statement. A change in disclosure can move the reported categories. Do not treat a missing policy statement as an onsite requirement.

The evidence behind the charts

Unavailable means one or more monthly category counts cannot be reported; it is different from the Unknown policy category. We sum counts, never monthly percentages.

Annual and matched half-year policy counts and shares
PeriodEligible postsRemote mentionedNot remoteUnknown
2011 Mar-Dec (partial)1,586UnavailableUnavailableUnavailable
20122,266362 (16%)Unavailable1,783 (78.7%)
20133,015507 (16.8%)172 (5.7%)2,336 (77.5%)
20143,892795 (20.4%)270 (6.9%)2,827 (72.6%)
20156,5501,285 (19.6%)2,203 (33.6%)3,062 (46.7%)
20167,4061,633 (22%)4,592 (62%)1,181 (15.9%)
20178,9212,080 (23.3%)5,772 (64.7%)1,069 (12%)
20189,1602,446 (26.7%)5,849 (63.9%)865 (9.4%)
20198,0002,669 (33.4%)4,692 (58.7%)639 (8%)
20206,7504,221 (62.5%)1,952 (28.9%)577 (8.5%)
20219,7808,090 (82.7%)991 (10.1%)699 (7.1%)
20227,2016,381 (88.6%)430 (6%)390 (5.4%)
20233,8883,170 (81.5%)415 (10.7%)303 (7.8%)
20243,6922,796 (75.7%)621 (16.8%)275 (7.4%)
20253,7862,722 (71.9%)819 (21.6%)245 (6.5%)
2026 Jan-Jun (partial)1,8311,294 (70.7%)395 (21.6%)142 (7.8%)
2025 H11,8541,356 (73.1%)379 (20.4%)119 (6.4%)
2026 H11,8311,294 (70.7%)395 (21.6%)142 (7.8%)

Download these aggregate series

Remote policy only. JSON nulls and blank CSV cells mean unavailable, never zero. No individual posts or contact details.

Download CSV Download JSON

Method, uncertainty and limits

Source: public top-level posts in Hacker News “Who is Hiring?” threads that name an organization and an opening. Independent analysis by StartupJobList, not endorsed by Hacker News. See our methodology.

Remote mentioned includes hybrid-only wording, restricted remote and company policy statements. Negation, mixed-role posts, and older “telecommute” vocabulary can affect categorization. A company policy does not necessarily apply to every role. The categories cannot measure return-to-office behavior at the same employers.

HN is a selected recruiting channel, not a representative survey of employers or workers. Repeat employers can influence post-weighted shares. We make no company-weighted or causal claim, and no claim about worldwide access. Changes in Unknown may reflect changing disclosure or categorization as well as changing work policies.

Categorization errors can differ by era. A descriptive proportion interval cannot capture those errors, repeat employers, or selection into HN. We therefore do not interpret these differences as formal market significance.

How sensitive is the direction?

These scenarios allow each period’s mention share to be wrong by up to 5 or 10 percentage points, independently and in either direction. They describe possible net effects of false positives and false negatives, not measured error rates. Bounds are clipped to 0–100% per period.

Latest matched H1, ±5 points per era: the change could range from -12.5 to 7.5 percentage points. The direction is not guaranteed under this assumption.

Latest matched H1, ±10 points per era: the change could range from -22.5 to 17.5 percentage points. The direction is not guaranteed under this assumption.

Fixed 2022–2025 comparison, ±5 points per era: the change could range from -26.7 to -6.7 percentage points. The direction survives this assumption.

Fixed 2022–2025 comparison, ±10 points per era: the change could range from -36.7 to 3.3 percentage points. The direction is not guaranteed under this assumption.

These scenarios are not confidence intervals. The size of categorization error in each period is not established by these aggregate counts. An overall accuracy figure alone cannot resolve that uncertainty.

Data cutoff: 2026-08-24. Provisional months (2026-07, 2026-08) remain in monthly downloads but not comparisons. The isolated 2008 observation is excluded from this report. 2011 covers March onward; later partial years are labeled.

For context, HN Trends has tracked remote language in hiring posts. This report makes post denominators, matched periods and unavailable categories explicit. For worker arrangements rather than advertising language, see the Survey of Working Arrangements and Attitudes. Its survey population and measures differ; the percentages are not interchangeable.

Correction history: first edition. Counts and comparisons are recalculated when the underlying data changes.