This article applies to England and Wales.
WhatsApp evidence by industry: the regulated sectors moved most
In brief: Messaging-app references in employment tribunal judgments are growing fastest in the sectors where informal messaging is most discouraged. Between 2017–19 and 2023–25 WhatsApp mentions rose sevenfold in healthcare judgments and fifteenfold in education, while the sharpest fall in letters was in the public sector. Retail is the exception: paper held.
Last updated: September 2026
By Yerty | Analysis of published Employment Tribunal decisions, 2017 to 2025.
Everyone expects hospitality and construction disputes to run on WhatsApp. They do. The surprise is what happened in hospitals, schools and the civil service.
This is a companion to tribunal evidence is moving from paper to screen, which follows email, WhatsApp and letters through nine years of published judgments and shows the shift is not explained by which claims or industries reach a judgment. This piece asks a different question of the same table: not which industries use messaging apps most, which is obvious, but where the change was largest, which is not.
Summary
- The largest proportional rises in WhatsApp references are in healthcare (×7.2), education (×15) and the public sector (×4.4), the three sectors with the strongest institutional reasons to keep work off personal phones. Hospitality, construction and retail remain highest in level.
- The public sector shed letters fastest, from 81.7% to 68.2% of its judgments.
- Retail did not shed them at all: letters 88.2% to 87.5% while WhatsApp more than tripled. Addition, not substitution.
- Composition is not the explanation. Holding the later period to the earlier industry mix moves the headline series by under a point, and WhatsApp rises in every one of the eighteen industries with enough judgments to report.
WhatsApp by industry, sorted by the size of the rise
Share of substantial judgments mentioning WhatsApp, ■ 2017–19 → ■ 2023–25, by primary respondent industry. Sorted by the multiple; industries with fewer than 50 judgments in 2023–25 omitted.
Education's multiple starts from one judgment in 2017–19; the direction is safe, the multiple is not a figure to quote on its own.
Provisional: from the first-cut population (n 2,039 / 1,960). To be re-issued on the re-run population (request 34).
Share of substantial single-claimant merits judgments (20,000 to 60,000 characters, one judgment per case, England and Wales) mentioning each term at least once. One industry per case, from the primary respondent. Industries with fewer than 50 judgments in 2023–25 (individual employer, housing, food manufacturing, automotive) are excluded from this table because their factors are unstable; they appear in the full table below.
The regulated sectors moved most
In 2017–19 a healthcare judgment mentioned WhatsApp one time in forty. In 2023–25 it is nearly one in five, a higher rate than retail and not far behind construction. Education went from one in ninety to one in six. The public sector remains low in level at 6.6%, but that is more than four times where it started.
None of these are workplaces where messaging apps were adopted for convenience with employer blessing. NHS guidance has discouraged clinical and staff communication over personal messaging apps for years; schools have safeguarding policies to the same effect; the civil service has its own. The judgments suggest that whatever the policy, the messages exist, and when a dispute reaches a tribunal they are produced.
That is a reading, not a demonstration. The data shows where references grew; it cannot show whether messaging itself grew, whether disclosure of it grew, or whether tribunals became more willing to describe it. All three would produce the same table.
Where paper held, and where it fell
The companion piece shows letters falling from 88% to 80% overall. The fall is not evenly spread.
The public sector shed paper fastest. Letters fall from 81.7% to 68.2% of public-sector judgments, 13.5 points, the largest decline of any industry. Technology (−20.6) and financial services (−13.8) fall further in points but on far smaller cells. The public sector is the largest sector in the table where paper has retreated by more than ten points.
Retail did not shed it at all. Letters run 88.2% to 87.5% in retail judgments while WhatsApp more than triples. Retail is the one large sector where the digital rise was addition rather than substitution: the messages arrived and the letters stayed. Charity and non-profit shows the same shape on a smaller base (letters 83.9% to 87.9%).
One plausible reading is that retail disciplinary processes remain heavily paper-driven by design, with letters at each stage of a standardised procedure, while the messaging sits alongside them as evidence of what happened on the shop floor. The data supports the pattern, not the explanation.
The composition test
Because each case carries one industry, the industry mix can be held constant cleanly. Holding the 2023–25 judgments to the 2017–19 industry mix moves email by 0.6 points, WhatsApp by 0.4 and letters by 0.1. The overall shift is not an artefact of which sectors reach a published judgment, and the within-industry table above is what shows it directly: WhatsApp rises in all eighteen industries with enough judgments to report.
What this analysis does not show
The industry is derived, not declared. It is inferred from the respondent's name and Companies House rather than stated in the judgment. Its accuracy has not been measured against a hand-coded sample.
Multi-respondent cases. 8.4% of cases have more than one respondent and are assigned to the primary one.
Factors on small bases are unstable. Education's fifteenfold rise starts from one judgment in 2017–19. The direction is safe; the multiple is not a figure to quote on its own, and the healthcare and public-sector rows, on larger cells, are the better-founded ones.
A mention is not reliance. The count is of what judgments describe, not of what decided them, and it cannot separate more messaging from more disclosure or more description.
Full table
| Industry | n 2017–19 | n 2023–25 | Letters | ||
|---|---|---|---|---|---|
| Healthcare | 237 | 256 | 70.5% → 83.6% (+13.1) | 2.5% → 18% (+15.5) | 92.4% → 82.4% (-10.0) |
| Public sector | 197 | 211 | 67.5% → 85.8% (+18.3) | 1.5% → 6.6% (+5.1) | 81.7% → 68.2% (-13.5) |
| Retail | 204 | 152 | 59.8% → 68.4% (+8.6) | 5.9% → 19.7% (+13.8) | 88.2% → 87.5% (-0.7) |
| Transport and logistics | 174 | 167 | 60.9% → 76% (+15.1) | 4% → 10.2% (+6.2) | 88.5% → 82% (-6.5) |
| Facilities and support | 129 | 143 | 82.9% → 81.8% (-1.1) | 3.1% → 9.8% (+6.7) | 93% → 86.7% (-6.3) |
| Hospitality | 135 | 130 | 67.4% → 77.7% (+10.3) | 5.9% → 27.7% (+21.8) | 85.9% → 76.9% (-9.0) |
| Manufacturing | 106 | 94 | 67.9% → 78.7% (+10.8) | 1.9% → 8.5% (+6.6) | 93.4% → 80.9% (-12.5) |
| Professional services | 88 | 105 | 70.5% → 85.7% (+15.2) | 2.3% → 15.2% (+12.9) | 86.4% → 83.8% (-2.6) |
| Education | 90 | 91 | 81.1% → 85.7% (+4.6) | 1.1% → 16.5% (+15.4) | 91.1% → 78% (-13.1) |
| Construction | 87 | 80 | 74.7% → 76.3% (+1.6) | 3.4% → 21.3% (+17.9) | 85.1% → 76.3% (-8.8) |
| Technology | 75 | 67 | 85.3% → 83.6% (-1.7) | 2.7% → 6% (+3.3) | 90.7% → 70.1% (-20.6) |
| Charity and non-profit | 56 | 58 | 75% → 89.7% (+14.7) | 1.8% → 10.3% (+8.5) | 83.9% → 87.9% (+4.0) |
| Recruitment and staffing | 42 | 65 | 81% → 87.7% (+6.7) | 2.4% → 10.8% (+8.4) | 71.4% → 70.8% (-0.6) |
| Individual employer | 62 | 46 | 58.1% → 73.9% (+15.8) | 3.2% → 30.4% (+27.2) | 80.6% → 67.4% (-13.2) |
| Financial services | 51 | 40 | 78.4% → 90% (+11.6) | 9.8% → 17.5% (+7.7) | 86.3% → 72.5% (-13.8) |
| Housing | 38 | 33 | 65.8% → 78.8% (+13.0) | 2.6% → 18.2% (+15.6) | 94.7% → 81.8% (-12.9) |
| Food manufacturing | 34 | 37 | 64.7% → 70.3% (+5.6) | 5.9% → 13.5% (+7.6) | 88.2% → 81.1% (-7.1) |
| Automotive | 34 | 36 | 50% → 75% (+25.0) | 0% → 27.8% (+27.8) | 88.2% → 75% (-13.2) |
Provisional: from the first-cut population (n 2,039 / 1,960). To be re-issued on the re-run population (request 34).
Fourteen further industries had fewer than 30 judgments in one or both periods and are not shown.
Frequently asked questions
Which industries saw the biggest rise in WhatsApp evidence at tribunal?
By multiple, education (1.1% to 16.5% of substantial judgments), healthcare (2.5% to 18.0%) and professional services (2.3% to 15.2%). By level, hospitality (27.7%), construction (21.3%) and retail (19.7%) remain highest in 2023–25.
Is WhatsApp evidence common in NHS and healthcare tribunal cases?
Increasingly. WhatsApp is mentioned in 18.0% of substantial healthcare judgments in 2023–25, up from 2.5% in 2017–19, despite long-standing guidance against using personal messaging apps for work communication.
Which sector has moved away from letters the most?
The public sector: letters appear in 68.2% of its judgments in 2023–25, down from 81.7%. Technology and financial services fall further in points but on much smaller numbers of judgments.
Is there any sector where paper evidence has not declined?
Retail. Letters appear in 87.5% of retail judgments in 2023–25 against 88.2% in 2017–19, while WhatsApp more than tripled over the same period. Charity and non-profit shows the same shape on a smaller base.
Is the rise in messaging-app evidence just a change in which industries go to tribunal?
No. Holding the later period to the earlier industry mix moves the headline series by under a point, and WhatsApp rises within every one of the eighteen industries with enough judgments to report.
Analyse the data
This cut is one industry at a time. Run it for your sector, or against a named respondent.
Intelligence Hub →Manage your evidence
Messages are evidence in every sector, including the ones that discourage them. Keep the thread, not the screenshot.
Case Hub →Sources
Analysis of published Employment Tribunal decisions, 2017 to 2025, drawn from the public register on GOV.UK under the Open Government Licence v3.0. Population as the companion piece.
Related reading: tribunal evidence is moving from paper to screen, can WhatsApp messages, emails and recordings be used as evidence at an employment tribunal.