Plate 04 — Independent Product Case · 2026

Finnair
Disruption Recovery.

A self-service recovery concept for the moment a journey goes wrong. It helps passengers understand what happened, what Finnair has already handled and what they need to do next.

Focus

Product strategy · Requirements Engineering · HTI · Applied AI

Independent product concept based on publicly available information. Not affiliated with, endorsed by, or produced on behalf of Finnair. No internal Finnair data, interviews or usability tests are presented as completed research.

01 — Business context

The disruption itself is expensive.
Uncertainty makes it worse.

Public Finnair reporting shows that large-scale disruption affects both operating result and customer satisfaction. The product question is not how to prevent every disruption but how to make recovery clearer, faster and less dependent on overloaded support channels.

€68M

2025 industrial-action operating-result impact · Jan–Sep

39 → 28

NPS during Q2 2025 disruption period

74.5%

Passengers using modern digital / self-service channels · Q1 2026

2,600+

Flights cancelled in H1 2025 industrial action

02 — Desk research

I looked for the gap between
system action and passenger understanding.

The case uses public sources only: Finnair investor reporting and help pages, EU261 material, public passenger reviews and a six-airline competitor benchmark. Findings are treated as evidence or hypotheses and never as invented internal data.

01

Passengers lack clarity

Public reviews repeatedly point to uncertainty around what has already been handled, what still needs action and what happens next.

02

Support is strained at the worst moment

When disruption affects many passengers at once, phone and chat queues become part of the problem rather than the fastest path to resolution.

03

Rights sit outside the journey

Refund, compensation, meal and hotel information is available, but it is not always presented in the context of one passenger's live booking.

How might we

How might we help passengers recover from flight disruptions quickly and independently?

03 — Product concept

One recovery view.
Four questions, in order.

Finnair Disruption Recovery is not a new app. It is a disruption mode inside the existing digital experience that temporarily replaces the normal booking view when a journey is affected.

  1. 01What happened?
  2. 02What has Finnair already done?
  3. 03What do I need to do?
  4. 04What are my other options?

Core product decision

Show state before choice: tell the passenger what is already true before asking them to make another decision.

FinnairAY1337
Flight cancelled

HEL

Helsinki

LHR

London

What we have already done

Rebooked you on the next available flight

Protected the rest of your itinerary

Issued an €18 meal voucher

Your decision

New departure

18:20 → 19:30

Today

See other optionsRequest refund

Your rights

See what applies to this booking

Still need help?

Your situation is shared with support

04 — Requirements

From evidence
to product behaviour.

Finding
Requirement
Product response

Passengers do not know what has already been handled.

Show one passenger-specific recovery summary.

Recovery Overview

Automatic rebooking can feel like a black box.

Make the proposed itinerary and alternatives easy to understand.

Journey comparison

Rights are difficult to interpret in the moment.

Explain rights using the actual booking and disruption state.

Contextual EU261

Examples of constraints

Operational status, rebooking and compensation must come from deterministic source systems — not generative AI.

The flow must degrade honestly when baggage, partner-airline or other upstream data is unavailable.

Human support remains available for disputes, exceptions and situations that require judgement.

Accessibility, legal accuracy and auditability are product requirements, not polish for later.

05 — MVP

Start with what reduces
uncertainty fastest.

01

Recovery Overview

A single screen that separates what Finnair has already done from what the passenger still needs to decide.

02

Contextual rights

Booking-specific refund, care and compensation information instead of a generic policy page.

03

Human escalation

One clear path to a person when self-service cannot safely resolve the situation.

06 — Recovery flow

Happy path first.
Edge cases stay visible.

Primary scenario: a connecting itinerary is broken by a cancellation and Finnair finds a viable replacement.

01

Disruption detected

02

Passenger notified

03

Full journey rebooked

04

Passenger reviews recovery

05

Accept / change / refund

No replacement exists

Escalate with context already attached.

Baggage status unclear

Show pending honestly; do not guess.

Overnight delay

Surface care / hotel actions proactively.

07 — Applied AI

AI explains.
It does not decide.

Explain my situation

“Why was I moved to this flight?”

The assistant receives already-computed booking and disruption data and translates it into plain language. It never invents missing operational facts.

Deterministic first

Flight status, baggage status, rebooking and compensation stay inside authoritative rule engines and operational systems.

Human when judgement is needed

Disputes, policy exceptions, unusual multi-carrier situations and accessibility needs leave the AI path and move to human support.

08 — Success

Measure resolution,
not screen engagement.

Primary metric

Self-service
disruption
resolution rate

Efficiency

Customer-service contacts per disrupted passenger

Speed

Time from notification to resolved state

Adoption

Automatic rebooking acceptance rate

Guardrail

Incorrect rebooking rate

A self-service system that reduces contacts but creates wrong itineraries would be a false win.

09 — Next

What I would validate
with real Finnair data.

  1. 01

    Analyse contact-centre reason codes to test how much disruption traffic is information-seeking versus decision-requiring.

  2. 02

    Usability-test the “already done vs. your decision” framing with disrupted-travel scenarios.

  3. 03

    Pilot one disruption type first, establish a baseline and compare recovery completion, support contacts and repeat contacts.

What this case demonstrates

Business context → evidence → requirements → scope → product decision.

Research basis: public Finnair reporting, public help material, EU261, passenger-review desk research and competitor benchmarking.

Independent product concept. Not affiliated with Finnair.