"Will AI replace travel agents?" is the wrong question. The right question is: "Which agents will AI replace, and which will it make irreplaceable?"
The answer depends on what agents actually do — and which parts of their work AI can replicate. This analysis separates hype from evidence to give travel professionals and agency leaders a realistic picture of the future.
What Travel Agents Actually Do
Breaking down agent activities reveals which are automatable and which require human capability:
Task Analysis
| Agent Task | Time Allocation | AI Capability | Human Advantage |
|---|---|---|---|
| Destination research | 15-20% | Very high — AI searches faster | Low — AI has access to more data |
| Price comparison | 10-15% | Very high — instant across suppliers | Low — price comparison is algorithmic |
| Booking administration | 15-20% | Very high — process automation | Low — routine process work |
| Customer needs analysis | 10-15% | Medium — can ask questions | High — empathy, reading between lines |
| Product recommendation | 10-15% | Medium — pattern matching | High — nuance, creativity, experience |
| Relationship building | 10-15% | Low — can't build trust authentically | Very high — human connection |
| Problem solving (complex) | 5-10% | Low-Medium — limited judgement | Very high — lateral thinking, advocacy |
| Emotional support (anxiety, excitement) | 5-10% | Low — can't feel | Very high — genuine empathy |
Key insight: AI can handle 40-55% of what agents currently do (research, comparison, administration). But the remaining 45-60% — needs analysis, creative recommendation, relationship building, complex problem solving, and emotional support — represents the highest-value work that differentiates great agents from mediocre ones.
The Evidence: What's Actually Happening
Agent Numbers
| Market | Agent Count 2019 | Agent Count 2024 | Change | Context |
|---|---|---|---|---|
| UK | ~37,000 | ~34,000 | -8% | Pandemic exit, not AI replacement |
| US | ~74,000 | ~71,000 | -4% | Bureau of Labor Statistics |
| Australia | ~15,000 | ~13,500 | -10% | Market consolidation |
The decline in agent numbers is driven primarily by pandemic exits and business model consolidation — not AI automation. In fact, the most technology-forward agencies are growing, not shrinking.
What Agencies Say
ABTA industry surveys show:
| Statement | % of Agencies Agreeing |
|---|---|
| "AI will help us serve customers better" | 72% |
| "AI will help us be more efficient" | 68% |
| "AI is a threat to our business model" | 22% |
| "We plan to reduce headcount due to AI" | 8% |
| "We plan to invest in AI tools in the next 12 months" | 55% |
The majority of agency leaders see AI as a tool, not a threat.
What Consumers Say
| Statement | % of Consumers Agreeing |
|---|---|
| "I would book a complex holiday entirely through AI" | 15% |
| "I would use AI for research but book through a human" | 42% |
| "I prefer human advice for important/expensive trips" | 68% |
| "AI can't understand what I really want from a holiday" | 58% |
| "I trust a human travel agent more than AI recommendations" | 62% |
Source: Phocuswright Consumer Travel Survey
What AI Will Replace
Some agent tasks and roles will be automated. Being honest about this is essential:
Tasks That AI Will Absorb
| Task | Timeline | AI Solution |
|---|---|---|
| Basic price checking | Already happening | Automated comparison engines |
| Simple booking administration | 1-2 years | Process automation |
| Routine destination information | Already happening | AI chatbots, knowledge bases |
| Standard itinerary creation | 1-3 years | AI itinerary generators |
| Booking confirmation and reminders | Already happening | Automated communications |
| Basic customer queries | Already happening | Self-service and chatbots |
Roles Most at Risk
| Agent Profile | Risk Level | Why |
|---|---|---|
| Order-taker (processes requests without adding value) | High | AI can take orders more efficiently |
| Price-focused (competes on finding cheapest option) | High | AI price comparison is instant |
| Information-only (provides facts without insight) | High | AI has better information access |
| Low-complexity specialist (simple beach holidays) | Medium | Commoditised products are easiest to automate |
What AI Cannot Replace
The Human Advantage
| Human Capability | Why AI Can't Replicate It | Example |
|---|---|---|
| Genuine empathy | AI simulates, not feels | Calming a nervous first-time flyer while recommending the right trip |
| Creative problem-solving | AI follows patterns; humans innovate | "The resort is fully booked, but I know the GM — let me make a call" |
| Trust and accountability | AI has no personal stake | "I've been there myself. I wouldn't recommend it for children under 5" |
| Reading between the lines | AI processes words; humans understand meaning | Recognising that "we want somewhere quiet" means "we're going through a difficult time" |
| Advocacy under pressure | AI follows rules; humans fight for customers | Advocating for a customer during a dispute with a supplier |
| Relationship memory | AI stores data; humans remember context | "Last time you mentioned your daughter loves marine biology — there's a new coral conservation programme" |
Roles That Become More Valuable
| Agent Profile | Why AI Makes Them More Valuable |
|---|---|
| Trusted advisor (deep relationships, life-stage knowledge) | AI handles admin, freeing time for relationship work |
| Specialist expert (deep destination/product knowledge) | AI can't replicate authentic experience-based expertise |
| Complex trip designer (multi-destination, special requirements) | Complex logistics need human creativity and judgement |
| Luxury/high-value specialist (concierge-level service) | High-spending clients want human attention and advocacy |
| Problem resolver (crisis management, complaint handling) | Emotional situations require human empathy and authority |
The Empowerment Model
The most likely future isn't replacement but augmentation — AI handles the mundane so agents can focus on the valuable:
Before AI: The Agent's Day
| Time | Activity | Value to Customer |
|---|---|---|
| 9:00-10:00 | Checking emails, admin | Low |
| 10:00-11:00 | Researching destinations for enquiry | Medium |
| 11:00-11:30 | Price comparison across suppliers | Low |
| 11:30-12:30 | Customer consultation | High |
| 12:30-1:30 | Lunch and catching up on supplier updates | None |
| 1:30-3:00 | Processing bookings and amendments | Low |
| 3:00-4:00 | Following up on quotes | Medium |
| 4:00-5:00 | More admin, end of day tasks | Low |
High-value customer time: ~2 hours (25%)
After AI: The Agent's Day
| Time | Activity | Value to Customer |
|---|---|---|
| 9:00-9:30 | Review AI-prepared briefings for today's appointments | Medium |
| 9:30-10:30 | Customer consultation (AI has pre-researched options) | High |
| 10:30-11:00 | AI coaching review from yesterday's calls | Medium (future value) |
| 11:00-12:00 | Customer consultation | High |
| 12:00-12:30 | Lunch | None |
| 12:30-1:00 | AI roleplay practice for afternoon specialist appointment | Medium (future value) |
| 1:00-2:00 | Complex multi-destination design consultation | Very High |
| 2:00-2:30 | Review AI-flagged opportunities (customers due for re-contact) | Medium |
| 2:30-3:30 | Customer follow-ups (AI has prepared personalised proposals) | High |
| 3:30-4:00 | Product training on new supplier | Medium |
| 4:00-5:00 | Relationship calls to existing customers | High |
High-value customer time: ~5 hours (63%)
The agent's role doesn't shrink — it transforms from 25% high-value work to over 60%.
What Agents Should Do Now
Invest in Skills AI Can't Replicate
| Skill | How to Develop It |
|---|---|
| Consultative selling | AI coaching on needs analysis and recommendation |
| Emotional intelligence | Roleplay practice with diverse customer scenarios |
| Specialist knowledge | Destination and product training for depth, not breadth |
| Relationship building | More time with fewer, higher-value clients |
| Complex problem-solving | Experience from handling difficult situations |
| Storytelling | Practice describing experiences, not listing features |
Embrace AI as a Tool
| AI Tool | What It Does for Agents |
|---|---|
| AI training | Learn new products faster and more effectively |
| AI coaching | Get personalised feedback on selling technique |
| AI roleplay | Practise difficult conversations safely |
| AI research | Get destination information compiled instantly |
| AI content | Generate personalised proposals and itineraries quickly |
What Agency Leaders Should Do
| Action | Purpose | Timeline |
|---|---|---|
| Implement AI enablement platform | Equip agents with AI-powered training and coaching | Now |
| Redefine agent KPIs | Measure value-added, not transaction volume | This quarter |
| Invest in specialist development | Build expertise AI can't replicate | Ongoing |
| Redesign agent workflows | Identify tasks to automate vs amplify | This quarter |
| Communicate the vision | Help agents see AI as an ally, not a threat | Immediately |
The Bottom Line
AI will replace travel agents who act like search engines — researching, comparing, and processing. It won't replace travel agents who act like trusted advisors — understanding, recommending, empathising, and advocating.
The travel agents of 2030 will be fewer in number but individually more valuable, more skilled, and more effective. They'll handle more complex work, build deeper relationships, and deliver higher-value service — because AI handles everything else.
The agents who invest in the skills AI can't replicate — and embrace AI for everything it can do — will thrive. Those who try to compete with AI at information processing and price comparison will find it an unwinnable race.
Equip your agents for the AI era with TravAI →
This article is part of our Travel Industry Trends series. Related reading:
