How WRC Hearings Are Won and Lost: Patterns From 12,847 Decisions

The dataset

The WRC publishes all adjudication decisions on its website. We collected, cleaned, and embedded 12,847 decisions spanning 2015 to early 2026. Each decision was chunked, tagged by claim type, outcome, award amount, and adjudicator, then embedded into a vector database for semantic analysis.

This is the dataset that powers the CaseDesk Precedent Engine. What follows are some of the patterns we've found.

1. Procedure beats substance

The single most consistent pattern across 12,847 decisions: the employer's adherence to fair procedures matters more than the underlying reason for dismissal. In 73% of cases where the employer lost, the adjudicator cited procedural failures — not the merits of the dismissal itself.

The most common procedural failures:

• No written warnings prior to dismissal (cited in 41% of employer losses)
• No opportunity for the employee to respond to allegations (38%)
• No right of appeal offered (29%)
• Investigation conducted by the same person who made the dismissal decision (22%)

What this means for practitioners:When preparing a case, the first question should not be “was the dismissal justified?” but “did the employer follow a fair process?” CaseDesk flags procedural gaps automatically.

2. Documentation gaps are decisive

In cases where the employer claimed to have issued warnings, adjudicators consistently distinguished between documented and undocumented warnings. Verbal warnings with no written confirmation were treated as unverified in 89% of cases.

More striking: in cases where the employee could demonstrate a period with no employer documentation at all — a “documentation gap” — the adjudicator ruled in favour of the employee 64% of the time, compared to 47% overall.

What this means for practitioners:CaseDesk's chronology builder automatically detects documentation gaps and flags them. This is not a nice-to-have feature — it surfaces one of the strongest predictors of case outcome.

3. Contradictions in the employer's own documents are gold

When the employer's documents contradict each other — for example, a dismissal letter citing reasons not mentioned in the investigation report — the claimant success rate jumps to 71%. Adjudicators interpret internal contradictions as evidence of an unfair process or a post-hoc rationalisation of a decision already made.

The most common contradictions:

• Dismissal letter reasons vs investigation report findings
• Number of warnings claimed vs warnings actually documented
• Contract terms (notice period, role) vs actual practice
• Employer's submission vs documentary evidence

What this means for practitioners:Finding these contradictions manually takes hours. CaseDesk's contradiction detector runs continuously as documents are added and surfaces conflicts with severity ratings within seconds.

4. Award amounts are more predictable than you think

Among successful unfair dismissal claims, the median award is €14,200. But there are clear patterns in what drives higher or lower awards:

• Length of service: Each additional year of service correlates with approximately €1,800 in additional compensation
• Mitigation efforts: Claimants who demonstrated active job-seeking received awards 23% higher than those who did not
• Severity of procedural breach: Cases with multiple procedural failures averaged €18,400 vs €11,200 for single failures
• Employer size: Awards against larger employers were on average 15% higher

5. Unrepresented claimants lose winnable cases

Among claims that were factually similar to successful cases (based on semantic similarity scoring), unrepresented claimants had a success rate of 31% compared to 52% for represented claimants. The gap was not in the merits — it was in presentation.

The most common issues with unrepresented submissions:

• Evidence submitted without labelling, dates, or context
• No structured chronology of events
• Legislation not cited or cited incorrectly
• Submissions structured as narrative rather than legal argument

This gap is what motivated the development of CaseDesk. The platform can't replace legal representation, but it can ensure that every submission — whether from a solicitor or a claimant — meets the structural standard that adjudicators expect.

Conclusion

WRC hearings are not unpredictable. They follow patterns — patterns that are invisible when you're reading decisions one at a time, but become clear when you analyse 12,847 of them together.

The firms that win consistently are the ones that identify procedural gaps, find contradictions in the employer's documents, and present their case against the right precedents. CaseDesk automates all three.