Industries

Which sectors this page compares#

Software has lower short-sleep and zero-exercise shares than education in this survey. The WHI 2026 questions cover everyday habits, recent stress, health-care actions and therapy use. Here, those answers are grouped by the sector people said they work in.

Those are the six largest named sectors, 5,918 of the 9,097 responses between them. Ten smaller named sectors, a broad "Other" category and responses that skipped the question account for the rest. Software dominated the sample in WHI 2025 as it does here, so this is a comparison between insured, salaried workforces in each sector — not a picture of the Indian labour force.

Sectors answered in very different numbers, and that governs what can be said about them. Every ranking on this page is limited to named sectors with at least 200 answers to the question in hand — six or seven sectors, depending on the question. The habit and stress tables include smaller published groups with their answer counts. The screening, therapy and medicine-awareness tables show only sectors that meet this reporting floor; nearby text identifies smaller groups outside the main range. The floor is a reporting choice, not a guarantee that a difference is statistically significant. The broad "Other" category is left out everywhere. It holds more answers than any sector except software, but it pools unlike employers together, so it is not a sector and cannot be ranked against one.

Short sleep is more common in EdTech than software#

Short sleep here means the two shortest options the sleep question offers: under five hours a night, or five to six. The contrast is clear: 38.8% in EdTech and Education versus 27.6% in software among sectors with at least 200 answers.

Of the seven sectors with at least 200 sleep answers, EdTech and Education is highest at 38.8% and IT/Software/SaaS is lowest at 27.6% — 11.2 points apart, with the survey-wide figure of 30.7% of 7,569 answers between them. Two thin samples cluster with EdTech at the top, NGO/Non-Profit at 38.4% on 138 answers and Government/Public Sector at 38.5% on 52, and one sits below software: Hospitality and Travel at 25.4% on 130. Software has the lowest share in that seven-sector comparison.

Sector is not the only way the report cuts the same question. Across the whole survey, short sleep runs 31.8% of 4,629 office-based answers, 30.8% of 1,333 home-based and 27.8% of 1,606 hybrid — a spread of 4.0 points, against 11.2 between sectors (Work Model). The observed spread between sectors is larger than the spread between work arrangements. These separate comparisons do not isolate the effect of either factor. The Sleep page sets out the answer bands and the wider pattern.

Zero exercise days are more common in EdTech than software#

The exercise question asks how many days in a typical week someone gets thirty minutes or more of activity. Answering zero is the measure below.

The order tracks sleep closely. EdTech and Education is again highest of the seven sectors with at least 200 answers, at 38.6%, and IT/Software/SaaS again lowest at 26.9% — 11.7 points apart, around a survey-wide 30.9% of 7,507 answers. Two thin samples bracket that range: NGO/Non-Profit is higher than EdTech at 41.3% on 138 answers, and Pharma matches software exactly at 26.9% on 78. Manufacturing has the second-lowest share among these seven sectors, at 27.9% of 519. BFSI at 29.6% and healthcare at 30.5% are also below the survey-wide 30.9%.

Two different questions putting the same two sectors at the same two ends is a stronger signal than either would be alone, though neither identifies what in the work produces it. The Movement and Exercise page has the full distribution of exercise days.

Finance has the highest any-drinking share; e-commerce has the highest nicotine-use share#

Two questions, one table. "Any alcohol" counts every answer except never. The nicotine measure counts any listed product — cigarettes, smokeless tobacco, vapes, and nicotine-replacement patches or gum — so a positive answer is not the same as smoking.

Among the six sectors with at least 200 alcohol answers, BFSI is highest at 44.6% of 579, against 35.3% of software's 2,703 — a 9.3-point gap, either side of a survey-wide 37.3% of 7,249. Three sectors report more drinking still, and all three are thin: Hospitality and Travel at 55.2% on 125 answers, Real Estate and Construction at 53.5% on 198, and Media and Entertainment at 49.6% on 125.

Nicotine does not sort the same way. E-commerce and Retail is highest of the six rankable sectors at 25.4% of 272, ahead of BFSI at 21.4% of 576, and no smaller published group sits above e-commerce. Real estate, FMCG, hospitality and media all sit above BFSI on smaller bases. BFSI is still 6.0 points above software, at 15.4% of 2,694, and EdTech is lowest at 12.6% of 461, against a survey-wide 17.2% of 7,223. So finance is high on both substances and software low on both — but the largest nicotine-product share belongs to e-commerce. The Alcohol and Nicotine pages carry the frequency and product breakdowns.

No sector escapes high stress — the floor is one in three#

The stress question asks for a rating of the previous week on a scale of one to ten. Seven or above is counted as high.

Every sector this page can rank sits between 34.5% and 47.8%: Real Estate and Construction at the bottom on exactly 200 answers, EdTech and Education at the top on 467, with the survey-wide figure at 40.4% of 7,346. There is no sector where fewer than a third of answers report high stress. NGO/Non-Profit is higher than EdTech at 53.6%, on 138 answers.

EdTech's 47.8% is 9.4 points above software's 38.4%, and BFSI is second among the rankable sectors at 44.2% of 586. The spread is 13.3 points — but the bottom of it is still more than a third of answers, which is what separates this measure from the others on the page.

Self-rated stress in the past week is not the exposure occupational cohorts measure, and this survey asks nothing about working hours. Those cohorts are where the clinical weight sits. A collaborative meta-analysis of 13 European cohorts, covering 197,473 workers of whom 30,214 reported job strain, put the hazard ratio for coronary heart disease at 1.23 (95% CI 1.10–1.37) — a small and consistent increase, and explicitly not the doubling earlier work had claimed (Kivimäki et al., The Lancet, 2012). A meta-analysis of 189,729 workers across 35 countries found long working hours associated with the onset of depressive symptoms at 1.14 (1.03–1.25) overall, with the largest estimate in Asian cohorts, 1.50 (1.13–2.01) (Virtanen et al., Scandinavian Journal of Work, Environment & Health, 2018). The Stress page sets out what people said was causing theirs.

Half the e-commerce and logistics answers select none of the listed health-care actions#

One survey question lists what someone might have done for their health in the past year: a routine or preventive doctor consult, a consult for a symptom or illness, a full-body diagnostic test, a fitness assessment, a company-organised health camp or screening — or "None of these". Selecting that last option is the measure below. Some responses selected both “None of these” and an action. This measure records the checkbox selection; it does not establish that someone had no health care.

E-commerce and Retail is highest at 54.0% of 263 answers and Logistics and Supply Chain next at 53.5% of 142, against EdTech and Education at 37.7% of 454 at the foot of the comparison set — 16.3 points, the widest gap among the main habit and care measures compared here. The survey-wide figure is 42.3% of 7,048.

E-commerce combines the highest nicotine-product share in the main comparison with the highest “None of these” selection. Logistics also has relatively high short-sleep and zero-exercise shares, on smaller samples. These are reasons to inspect access to care, not a ranking of clinical need.

Across the whole survey, reported full-body diagnostic testing varies more widely by income: 15.6% of 887 answers in the lowest band versus 54.3% of 138 in the highest (Preventive Care). Both comparisons show variation in reported use; neither identifies why an individual did or did not seek care.

Company screening participation is highest in manufacturing#

Among six sectors with at least 200 answers, 17.5% of manufacturing answers report a company health camp or screening in the past year. The share is 6.8% in e-commerce. This is reported participation, not a measure of whether employers offered screening.

Manufacturing leads company-screening participation in this comparison, but software and healthcare have higher diagnostic-testing shares. Different ways of getting care do not rank sectors in the same order. Among smaller samples, pharma reports company screening at 21.3% of 75 answers and media at 4.9% of 123; consulting reports diagnostic testing at 34.1% of 164, while hospitality reports 14.8% of 122.

Across all 7,048 answers to this question, 10.5% selected company screening, 24.3% a diagnostic test and 24.4% a routine or preventive consult. The five listed actions can overlap, so their percentages must not be added.

Therapy use ranges from 7.4% in e-commerce to 13.8% in EdTech#

The question asks whether someone has ever taken therapy or counselling. It includes current and past use, rather than use in the past week or year.

EdTech and healthcare have the two highest shares among seven sectors with at least 200 answers. E-commerce has the lowest. Smaller groups extend the range: NGO/Non-Profit reports 21.7% of 138 answers and pharma 5.1% of 78.

Therapy was also reported in 2025: software 13.2%, healthcare 16.8%, BFSI 9.3% and manufacturing 7.8%, compared with 9.4%, 13.1%, 11.4% and 8.9% here. These are reported lifetime-use shares in successive samples, not individual changes in treatment.

High stress and therapy use answer different questions. EdTech has high shares on both, but a lifetime therapy measure cannot show how many people received help for stress in the previous week. For employers, the useful question is whether staff can seek confidential help without needing to know which type of professional to approach first.

Medicine awareness varies more by sector than caffeine or eating out#

Three further sector comparisons sit outside the habits above. The first has a real spread.

Awareness means having heard of the medicines, knowing what they are, or having considered or used one. Healthcare is highest of the six rankable sectors at 36.4% of 588 and Manufacturing lowest at 21.4% of 471, a 15.0-point spread around a survey-wide 26.1% of 6,968. Consulting is higher than healthcare at 42.7%, on 164 answers. This is awareness and not use (GLP-1 Awareness).

The other two have smaller spreads. Workday caffeine use is common: among those that can be ranked it runs from 74.3% of software's 2,686 answers to 80.6% of manufacturing's 489, a 6.3-point spread around a survey-wide 76.6% of 7,211. Eating out or ordering in three or more times a week is flatter still, 21.6% to 25.5% across seven sectors against 23.5% of 7,614 survey-wide. These sector shares are closer together than those for short sleep, stress or the listed care actions.

Alcohol use fell across four sectors; exercise changed less#

WHI 2025 published figures for seven sectors. Four have published 2025 sector samples of at least 200 and at least 200 answers per current measure: healthcare, BFSI, manufacturing and software. Consulting is too thin in both — 111 respondents in 2025 against 164 to 173 answers here. EdTech and Education, and E-commerce and Retail — the sectors at the extremes of several tables above — were measured last time too, on 161 and 119 respondents, which is below the floor this page ranks on.

The 2025 stress column carries two figures because WHI 2025 published conflicting values for the same four sectors with the same listed sector sample sizes. For healthcare they are 18.4 points apart and the 2026 value falls between them: against 57.1% it is a drop of 15.9 points, against 38.7% a rise of 2.5. Healthcare's change cannot be given a direction, and this page does not give it one.

The other three moved the same way whichever 2025 figure is taken. High stress rose between 1.1 and 9.6 points in BFSI, between 2.9 and 6.0 in software, and between 7.8 and 11.3 in manufacturing. A rise in all three is what this column establishes; the size of it is not settled.

BFSI remained near the top on any alcohol use: second of the five sectors reported in 2025 at 47.8%, and highest in the current main comparison at 44.6%, a fall of 3.2 points. Software's reported share fell most among these four sectors, by 7.0 points. Short-sleep figures are shown for context, but the answer-band definitions differ between editions.

The differences in published exercise shares are smaller than the largest alcohol differences. The share answering zero days rose by about 3.9 points in software, 3.5 in healthcare and 1.6 in BFSI, and manufacturing's 28% to 27.9% is unchanged within the precision of the 2025 figure, which is printed to whole per cent.

What a sector comparison can and cannot tell an employer#

What this page establishes, for salaried insured workforces in Loop's client book:

What it does not establish is that working in a sector causes any of this. Sectors differ in age, sex, income, role, city and working arrangement, and none of those is held constant in these comparisons. That position at work tracks health is an old and well-evidenced idea — the Whitehall II cohort of 10,314 British civil servants found a steep gradient in morbidity by grade of employment (Marmot et al., The Lancet, 1991) — but that study graded rank inside a single employer, which is a different cut from the one made here. The report's nearest equivalent is its job-function and income analysis.

For a benefits team the usable output is narrower than a league table and more practical. Your own sector's answers say which of these measures to ask about first; the care-seeking table says where the gap is likely to be widest; and the Preventive Care page shows how reported testing also varies by income across this survey.