Introduction
This case study examines a hypothetical Australian domestic airline, referred to throughout as the Carrier, following mass cancellations across the December 2025 to January 2026 school holiday peak. The Carrier operates a single-fleet narrowbody network of 18,400 scheduled sectors over the two-month period, serving capital city trunk routes and regional ports. All figures are illustrative but internally consistent.
The analysis classifies the failure and tests whether the service recovery paradox offers a defensible basis for the response, quantifies the operational, attitudinal and financial consequences, and evaluates a redesigned recovery process against the Carrier’s obligations under the Australian Consumer Law. Justice theory provides the evaluative framework, and Figure 1 sets out the recommended process.
Case Background
The Carrier entered the peak with a schedule built for an 84 per cent load factor and no reserve aircraft, a configuration that maximises utilisation but leaves no absorptive capacity when disruption occurs. Crew shortages, two unscheduled maintenance groundings and a week of severe east coast weather then removed capacity from a network that had none to spare, and cancellations concentrated in the fortnight from 20 December, when leisure demand peaks and alternative capacity is least available.
Context shapes how the numbers should be read. The Australian Competition and Consumer Commission (2023) characterises the domestic market as highly concentrated, with limited practical substitution on many routes, so a disrupted passenger frequently has no realistic alternative operator. Australian Bureau of Statistics (2025) data show the holiday travel and accommodation subgroup rising well above headline inflation over the period, so where the Carrier failed its duty of care the burden transferred to passengers at an elevated replacement cost. The failure occurred in conditions that magnified consumer detriment.
Classifying the Service Failure
Failure taxonomy
Bitner, Booms and Tetreault (1990) classify service encounter failures into three groups: failures in the core delivery system, failures to respond to customer needs, and unprompted employee actions. The Carrier’s disruption is primarily a Group 1 failure, since the core promise of carriage at a scheduled time was not delivered, but the critical point is that it cascaded into Group 2. Cancellations are to a degree unavoidable in aviation; what converted an operational event into a reputational one was the Carrier’s inability to respond to the resulting requests for rebooking, refund and information. Severity was compounded by frequency, because repeated failures across a booking shift the passenger’s frame of reference from the individual transaction to the reliability of the relationship (Michel, Bowen and Johnston 2009).
The service recovery paradox
The service recovery paradox proposes that a customer who experiences a failure followed by excellent recovery may report higher satisfaction than one who experienced no failure at all (Hart, Heskett and Sasser 1990). The proposition is attractive because it appears to convert failure into opportunity, but the empirical support is conditional. The meta-analysis by De Matos, Henrique and Rossi (2007) finds the effect holds for satisfaction under limited conditions, generally does not extend to repurchase intention or word of mouth, and weakens sharply as severity rises and failures repeat. The Carrier’s data below confirm the boundary condition rather than the paradox, so recovery is loss mitigation and prevention retains primacy.
Analysis of Complaints and Operational Impact
Operational performance
Cancellation rate is cancelled sectors divided by scheduled sectors:
Cancellation rate = 1,196 / 18,400 x 100 = 6.5%.
On-time performance is departures within 15 minutes of schedule as a share of sectors operated. With 1,196 cancellations, 17,204 sectors operated, of which 10,838 departed on time:
On-time performance = 10,838 / 17,204 x 100 = 63.0%.
Both measures sit well outside the market. Bureau of Infrastructure and Transport Research Economics (2025) reporting places industry on-time departures at 71.2 per cent and cancellations at 3.4 per cent across the comparable period, so excess cancellations relative to that benchmark are (6.5% – 3.4%) x 18,400 = 570 sectors. At an average of 132 passengers per cancelled sector, lower than the 141 on operated sectors because cancellations fell disproportionately on regional services, total passengers affected were 1,196 x 132 = 157,872.
Complaint profile
The Carrier logged 24,180 complaints against 2,425,764 passengers carried, a rate of 24,180 / 2,425,764 x 1,000 = 10.0 per 1,000 passengers, four times the internal target of 2.5. As Table 1 shows, complaints were not distributed evenly across failure types, channels or resolution times.
Table 1: Complaint volume, channel and resolution performance by failure type, December 2025 to January 2026.
| Failure type | Complaints (n) | Share (%) | Dominant channel | Mean resolution (days) | Escalated to ACA (n) |
|---|---|---|---|---|---|
| Cancellation, rebooked beyond 24 hours | 8,704 | 36.0 | Contact centre | 21.4 | 742 |
| Cancellation, rebooked same day | 4,352 | 18.0 | Mobile app | 9.6 | 148 |
| Delay exceeding 3 hours | 3,869 | 16.0 | Social media | 6.2 | 116 |
| Baggage delayed or mishandled | 2,660 | 11.0 | Web form | 14.3 | 133 |
| Refund and credit processing | 2,176 | 9.0 | Written correspondence | 33.7 | 402 |
| Denied boarding or downgrade | 1,451 | 6.0 | Airport service desk | 12.8 | 174 |
| Staff conduct and information quality | 968 | 4.0 | Social media | 4.1 | 26 |
| Total | 24,180 | 100.0 | Multiple | 16.0 | 1,741 |
The total resolution time is a volume-weighted mean:
Weighted mean = [(8,704 x 21.4) + (4,352 x 9.6) + (3,869 x 6.2) + (2,660 x 14.3) + (2,176 x 33.7) + (1,451 x 12.8) + (968 x 4.1)] / 24,180 = 385,943.4 / 24,180 = 16.0 days.
Against an internal standard of 10 business days, the Carrier resolved the average complaint 60 per cent late. Escalation to the Airline Customer Advocate, the industry-funded body to which participating carriers refer unresolved complaints, ran at 1,741 / 24,180 = 7.2 per cent. The distribution is diagnostic: refund and credit processing generated 402 / 2,176 = 18.5 per cent escalation, the highest of any category, and the two slowest categories, refunds at 33.7 days and delayed rebooking at 21.4 days, account for 45.0 per cent of complaints but 1,144 / 1,741 = 65.7 per cent of escalations. Escalation is thus driven by elapsed time, not by the severity of the operational event, consistent with the finding of Orsingher, Valentini and de Angelis (2010) that perceived timeliness and effort are among the strongest determinants of satisfaction with complaint handling. The Airline Customer Advocate (2025) identifies refunds and credits as the dominant source of referrals across participating airlines, so this is a sector-wide weakness rather than an idiosyncratic one.
Net Promoter Score movement
Net Promoter Score is the percentage of promoters less the percentage of detractors (Reichheld 2003). In November 2025, from 4,120 responses comprising 1,895 promoters, 1,319 passives and 906 detractors:
NPS = (1,895 / 4,120 x 100) – (906 / 4,120 x 100) = 46.0 – 22.0 = +24.
In January 2026, from 5,340 responses comprising 1,282 promoters, 1,816 passives and 2,242 detractors:
NPS = (1,282 / 5,340 x 100) – (2,242 / 5,340 x 100) = 24.0 – 42.0 = -18.
The swing is 24 – (-18) = 42 points in eight weeks. Disaggregating the 2,860 disrupted respondents tests the paradox directly. Among the 1,040 who rated the Carrier’s handling in the top two boxes, 447 were promoters and 333 detractors, giving +11; among the 1,820 who did not, 200 were promoters and 1,238 detractors, giving -57.
Effective recovery therefore produced a 68 point improvement over failed recovery, yet still fell 13 points short of the undisrupted November baseline of +24. The paradox does not hold: for a severe, repeated failure the Carrier’s best recovery recouped roughly five sixths of the attitudinal damage and no more, which places the investment case on prevention first and recovery second.
Cost of failure and of recovery
Direct recovery expenditure across the peak totalled A$22.36 million, spanning accommodation and ground transport, refunds and interline fare differences, additional resourcing, meal vouchers and goodwill credits. On the affected base:
Recovery cost per affected passenger = 22,360,000 / 157,872 = A$141.63.
Churn cost is modelled from stated switching intention. Post-disruption research recorded 18.5 per cent of affected passengers intending to change carrier, converted at the Carrier’s historical 62 per cent intention-to-defection rate, so passengers lost = 157,872 x 0.185 x 0.62 = 18,108. At annual contribution of 412 x 0.234 = A$96.41 per passenger over an average 4.2 year relationship, lifetime contribution is A$404.92, giving:
Churn cost = 18,108 x 404.92 = A$7,332,291, or A$7.33 million.
Total quantified cost is 22.36 + 7.33 = A$29.69 million, before regulatory exposure and lost forward bookings among passengers not themselves disrupted.
Applying Justice Theory to the Recovery
Perceived justice explains why two passengers receiving identical compensation can reach opposite conclusions. Smith, Bolton and Wagner (1999) and Tax, Brown and Chandrashekaran (1998) establish three dimensions, each engaged differently here.
Distributive justice
Distributive justice concerns the fairness of the outcome relative to the loss sustained. The Carrier’s average remedy of A$141.63 must be assessed against actual detriment: a passenger rebooked beyond 24 hours over a holiday period incurs accommodation, meals, forfeited pre-paid arrangements and often lost annual leave. Compensation calibrated to fare value rather than consequential loss is systematically under-scaled for the very group whose losses are greatest, which generates 36 per cent of complaints. Issuing credits in place of refunds compounds this, because a credit transfers residual risk back to the passenger who has just experienced the failure.
Procedural justice
Procedural justice concerns the fairness, accessibility and speed of the process, and it is where the Carrier failed most comprehensively. A 16.0 day weighted mean against a 10 day standard, a 33.7 day mean on refunds, and channel fragmentation across five contact points all impose what the literature terms recovery effort on the customer. The escalation pattern in Table 1 shows the consequence: passengers escalate when the process stalls, not when the outcome disappoints.
Interactional justice
Interactional justice concerns the quality of interpersonal treatment, including explanation, empathy and honesty. Complaints about staff conduct and information quality were the smallest category at 4.0 per cent and resolved fastest at 4.1 days, so frontline behaviour was not itself the problem. The deficiency was structural: staff without delegated authority can offer empathy but not remedy, and repeated referral without resolution is read as evasion. The remedy is delegation rather than training alone.
Redesigning the Recovery Process
Figure 1 illustrates the recommended process, built so that the three justice dimensions are addressed by design rather than discretion.
Table 2 evaluates the initiatives that give effect to this process.
Table 2: Recovery strategy evaluation for the Carrier, showing mechanism, incremental cost and expected effect.
| Initiative | Mechanism | Incremental cost (A$m) | Expected effect |
|---|---|---|---|
| Automated re-accommodation engine, offer within 30 minutes | Procedural justice: removes customer effort and waiting | 3.40 one-off, 0.35 per year | Rebooking complaint volume down 28%; mean resolution 21.4 to 6.0 days |
| Frontline settlement authority to A$250 without referral | Interactional and distributive justice: empowers the encounter | 1.30 per peak | First-contact resolution 31% to 65%; escalations down 60% |
| Published duty-of-care standard with defined thresholds | Distributive justice: calibrates remedy to loss, sets expectations | 1.10 per peak | Stated switching intention 18.5% to 12.0% |
| Two standby aircraft and a reserve crew pool over peak | Prevention: restores absorptive capacity | 4.86 per peak | Cancellation rate 6.5% to 3.8%; on-time performance 63.0% to 74.0% |
| Single case-management platform, 10 business day standard | Procedural justice: ends channel fragmentation | 1.75 one-off, 0.60 per year | Weighted mean resolution 16.0 to 8.5 days |
| Proactive communication protocol with cause and entitlement | Interactional justice: explanation before request | 0.45 one-off, 0.20 per year | Conduct and information complaints down 55% |
| Total | All three justice dimensions plus prevention | 5.60 one-off, 8.41 recurring | See benefit calculation below |
Two streams quantify the benefit. Reducing the cancellation rate from 6.5 to 3.8 per cent removes 1,196 – (0.038 x 18,400) = 497 cancellations, avoiding 497 x 132 = 65,604 disrupted passengers at A$141.63 each, an avoided recovery cost of A$9.29 million. Reducing stated switching intention from 18.5 to 12.0 per cent at the same conversion retains 18,108 – (157,872 x 0.12 x 0.62) = 6,362 relationships at A$404.92 each, an avoided churn cost of A$2.58 million.
Total quantified benefit is 9.29 + 2.58 = A$11.87 million against recurring cost of A$8.41 million, a benefit-cost ratio of 11.87 / 8.41 = 1.41 and a net annual benefit of A$3.46 million, so payback on the A$5.60 million of one-off investment is 5.60 / 3.46 = 1.6 years.
Obligations under Australian Consumer Law
The Australian Consumer Law in Schedule 2 to the Competition and Consumer Act 2010 (Cth) applies to air carriage as a supply of services. Section 60 guarantees services rendered with due care and skill, and section 62 guarantees supply within a reasonable time where none is fixed. Section 64 renders void any term purporting to exclude those guarantees, so a fare described as non-refundable cannot displace them.
Where the failure is major, section 267 gives the election to the consumer, who may terminate and recover a refund or keep the service and recover compensation for the reduction in value. A cancellation leaving a passenger without carriage for more than 24 hours during a holiday peak, on a route where the ACCC (2023) reports limited substitution, will ordinarily meet the major failure threshold, so the Carrier’s practice of issuing travel credits carries direct exposure. Of the 157,872 affected passengers, 21 per cent, or 33,153, sought a refund at an average fare of A$214, and credits were issued instead in 41 per cent of those cases:
Potential remediation exposure = 33,153 x 0.41 x 214 = 13,593 x 214 = A$2,908,902, or A$2.91 million.
Section 18 prohibits misleading or deceptive conduct and section 29 prohibits false representations about a right or remedy; advising a passenger that a fare is non-refundable when section 267 confers an election engages both, and misrepresentation of consumer guarantees has been a standing ACCC (2024) enforcement priority. The Aviation White Paper (Department of Infrastructure, Transport, Regional Development, Communications and the Arts 2024) proposed an Aviation Industry Ombuds Scheme with powers to direct remedies, signalling that the voluntary Airline Customer Advocate model will be superseded by a stronger accountability regime.
Recommendations
- Restore absorptive capacity before the next peak by funding the two standby aircraft and reserve crew pool. This is the highest-value action, since it removes 497 cancellations at source and recovery cannot restore what prevention protects.
- Replace credits with refunds as the default remedy for cancellations exceeding 24 hours, eliminating an estimated A$2.91 million of remediation exposure, removing the largest source of escalation and aligning practice with section 267 rather than fare conditions voided by section 64.
- Delegate settlement authority of A$250 to the frontline, converting an interaction into a resolution and addressing interactional and distributive justice where remedy is cheapest and most credible.
- Consolidate all complaint channels onto one case-management platform with a published 10 business day standard, since escalation is driven by elapsed time rather than by the operational event.
- Publish a duty-of-care standard with defined meal, accommodation and transport thresholds, converting discretionary goodwill into a stated entitlement that evidences compliance.
- Report prevention and recovery jointly to the board each quarter, using cancellation rate against the BITRE benchmark, resolution time, escalation rate and disrupted-passenger NPS, since better recovery that masks worse reliability is not an improvement.
Conclusion
The Carrier’s difficulty was operational in origin and procedural in consequence. A cancellation rate of 6.5 per cent against an industry 3.4 per cent produced 157,872 disrupted passengers, but it was a 16.0 day weighted mean resolution time and the substitution of credits for refunds that converted a capacity failure into 24,180 complaints, 1,741 escalations and a 42 point NPS collapse.
The recovery paradox does not assist the Carrier: passengers who received its best recovery still returned an NPS of +11 against an undisrupted baseline of +24, confirming that for severe and repeated failures recovery mitigates loss rather than creating value. The proposed program returns A$11.87 million against A$8.41 million of recurring cost with payback of 1.6 years, yet its stronger justification is that with an ombuds scheme carrying directive powers in prospect, the discretion the Carrier currently exercises over remedies is one it is unlikely to retain.
References
Airline Customer Advocate 2025, Annual review of airline customer complaints, Airline Customer Advocate, Sydney.
Australian Bureau of Statistics 2025, Consumer price index, Australia, ABS, Canberra.
Australian Competition and Consumer Commission 2023, Domestic airline competition in Australia, ACCC, Canberra.
Australian Competition and Consumer Commission 2024, Compliance and enforcement policy and priorities, ACCC, Canberra.
Bitner, MJ, Booms, BH & Tetreault, MS 1990, ‘The service encounter: diagnosing favourable and unfavourable incidents’, Journal of Marketing, vol. 54, no. 1, pp. 71-84.
Bureau of Infrastructure and Transport Research Economics 2025, Domestic airline on time performance, BITRE, Canberra.
De Matos, CA, Henrique, JL & Rossi, CAV 2007, ‘Service recovery paradox: a meta-analysis’, Journal of Service Research, vol. 10, no. 1, pp. 60-77.
Department of Infrastructure, Transport, Regional Development, Communications and the Arts 2024, Aviation white paper: towards 2050, DITRDCA, Canberra.
Hart, CWL, Heskett, JL & Sasser, WE 1990, ‘The profitable art of service recovery’, Harvard Business Review, vol. 68, no. 4, pp. 148-156.
Michel, S, Bowen, D & Johnston, R 2009, ‘Why service recovery fails: tensions among customer, employee and process perspectives’, Journal of Service Management, vol. 20, no. 3, pp. 253-273.
Orsingher, C, Valentini, S & de Angelis, M 2010, ‘A meta-analysis of satisfaction with complaint handling in services’, Journal of the Academy of Marketing Science, vol. 38, no. 2, pp. 169-186.
Reichheld, FF 2003, ‘The one number you need to grow’, Harvard Business Review, vol. 81, no. 12, pp. 46-54.
Smith, AK, Bolton, RN & Wagner, J 1999, ‘A model of customer satisfaction with service encounters involving failure and recovery’, Journal of Marketing Research, vol. 36, no. 3, pp. 356-372.
Tax, SS, Brown, SW & Chandrashekaran, M 1998, ‘Customer evaluations of service complaint experiences: implications for relationship marketing’, Journal of Marketing, vol. 62, no. 2, pp. 60-76.