← back to rankingEXPE · Expedia Group, Inc.
Travel Services · mkt cap $25.9B · calls: Q1 FY2026 vs Q4 FY2025
48.0 conviction · conf-adj 45
conf 3/10 partial
enthusiasm:27.0 · trend:8 · quantifies:5 · impact:0 · under_radar:5 · credibility:0 · business_impact:8 · disruption:-14 · commitment:6 · confirmation:3
Enthusiasm latest 9 / prev 8 (rising)
Expedia's AI thesis is that AI improves conversion through personalization, expands supply through faster onboarding, cuts service friction and costs, and opens new acquisition channels through answer-engine and chatbot integrations. Credibility rose in Q1 FY2026 because management moved from broad strategic claims to concrete operating metrics, including AI-powered service share, onboarding-time reduction, and marketing value. The main caveat is that traffic and bookings from AI-driven channels remain small.
GROUNDED NEXT-FY IMPACT vs CONSENSUS
Grounded on actual base — revenue $14.7B · net income $1.3B · net margin 8.8% · diluted EPS 9.81
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: % · next-FY EPS uplift: 6.47% · vs analysts: inline · priced in: medium (model's call-read: high; verdict above is the hard-data one used for ranking) · confidence: 3/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|
traveler interaction data for ranking/recommendation AI
engagement · soft | hundreds of millions (interactions) | Operational data-scale moat; no conversion-rate, take-rate, or revenue-per-interaction disclosed to convert to $ against $14.733B revenue. | | |
AI supply advantage from properties and exclusives
engagement · soft | 3.7M properties; 800k exclusive (21.6%) | 800k/3.7M = 21.6% exclusive supply breadth; no revenue-per-property, booking conversion, or take-rate uplift disclosed to anchor a $ impact. | | |
AI-enabled lodging property count growth
engagement · soft | 10% (~370k incremental properties on 3.7M base) | Real %; 10%*3.7M=370k properties, but property count != revenue and no revenue-per-property disclosed to translate into a rev line. Topline enabler, unconvertible. | | |
service interactions and self-service scale
productivity · soft | >250M interactions, >50% self-serve (>125M) | Cost-side efficiency; no cost-per-contact or service-opex base disclosed to size $ savings. | | |
AI-powered self-service interactions
productivity · soft | >30% (>37.5M AI-powered self-service) | % of an unpriced interaction pool; no human-contact cost avoided or $ base to compute savings. | | |
AI conversation summaries language coverage
productivity · soft | 30+ languages | Qualitative service-productivity benefit; no interaction volume by language, labor cost, or $ anchor. | | |
new agent onboarding time reduction
productivity · soft | ~60% faster onboarding | Training-time %, but no baseline onboarding hours, new-agent count, or wage cost disclosed; cannot size $ saving. | | |
AI-enabled marketing value
productivity | hundreds of millions of dollars | Only dollar-anchored claim. X: $300M low end but 'realized' run-rate already in FY25 base, so count only ~$125M incremental next-FY portion -> $99M after-tax (21%) -> 5.0% adj EPS. Y: $200M conservative lower bound full -> $158M after-tax -> 7.94% adj EPS. Both defensible; used average 6.47%. Revenue effect ~0 (efficiency, not new sales). | 0 | 6.47 |
AI-driven property onboarding speed improvement
productivity · soft | 70% faster | Cycle-time gain (1.7x throughput) enabling 10% supply growth; no baseline onboarding cost or revenue/property conversion disclosed. Topline enabler, unconvertible. | | |
Assumptions: All claims adopter-side; no supplier revenue. EPS % computed against the adjusted/non-GAAP consensus earnings base (FY25 consensus NI ~$1,989M / EPS $15.36), NOT GAAP ($1,294M / $9.81), per earnings-basis rule. Tax rate 21%; savings flow at 79% after-tax. Incremental margin on any new revenue would be current net margin 8.78% (none sized — no revenue claim anchored). Marketing-value claim is the only dollar-anchored item; X sized only the incremental next-FY portion of a $300M realized run-rate (5.0%), Y took a $200M lower bound at full value (7.94%) — both defensible, so the consensus uses their average (6.47%).
Top line: No AI claim carries an anchored incremental-revenue figure. The strongest topline items — 3.7M/800k-exclusive (21.6%) supply, 10% AI-driven property growth (~370k properties), 70% faster onboarding, hundreds of millions of traveler interactions — are real engagement/supply enablers but have no disclosed revenue-per-property or take-rate to convert into a revenue line, so est_rev_uplift_pct is null (soft, not zero-impact). These trends already feed the consensus FY26 revenue path ($14,733M -> $16,009M); AI's marginal contribution above that is not separately quantifiable.
Bottom line: The only dollar-anchored claim is 'hundreds of millions of dollars in realized marketing value through productivity/automation.' Sizing splits: counting only the incremental next-FY portion of a $300M realized run-rate gives ~5.0% adj EPS, while a $200M conservative full lower bound gives ~7.94%; consensus uses the ~6.5% average. Combined with 'actions will favor margins' and self-service automation (>30% AI-powered, ~60% faster onboarding), AI is a margin/efficiency story, not a revenue story.
Consensus already models EPS up +28.1% ($15.36 -> $19.68) on only +8.7% revenue and NI +24.5% ($1,989M -> $2,476M) FY25->FY26 — heavy margin expansion baked in. A ~6.5% incremental AI EPS uplift sits comfortably inside that 28% modeled growth, and management's own 'savings...favor margins' framing is exactly what that consensus margin expansion represents. AI efficiency gains appear embedded, not additive; nothing in the disclosures points clearly above the consensus trajectory.
MODEL CONSENSUS (impact)
partial
Agree on verdicts (inline/high, adopter, no anchored revenue); differ only on marketing-value EPS sizing, resolved by averaging the two defensible methods.
Conflicts reconciled
- marketing eps_uplift_pct: X=5.0 (incremental portion of $300M run-rate) vs Y=7.94 ($200M full lower bound) -> used 6.47 (average; both defensible)
- est_eps_uplift_pct: X=5.0 vs Y=7.94 -> used 6.47 (average)
- soft-claim rev_uplift_pct: X=null vs Y=0 -> used null (soft claims null per schema)
- confidence: X=3 vs Y=4 -> used 3 (lower on conflict)
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | – | – |
| EPS uplift % | 5.0 | 7.94 |
| Priced in | high | high |
| vs analysts | inline | inline |
| Confidence | 3 | 4 |
| Top line | No AI claim carries an anchored incremental-revenue figure. The strongest topline items — 3.7M/800k-exclusive supply, 10% AI-driven property growth, 70% faster onboarding, hundreds-of-millions of traveler interactions — are real engagement/supply enablers but have no disclosed revenue-per-property or take-rate to convert into a revenue line, so est_rev_uplift_pct is null (soft, not zero-impact). These trends already feed the consensus +8.7% FY26 revenue path ($14,733M -> $16,009M); AI's marginal contribution above that is not separately quantifiable from the disclosures. | No hard revenue uplift is calculable. The strongest topline evidence is operational: 370,000 implied incremental properties from 10% growth on a 3.7M base and 21.62% exclusive supply, but Expedia did not disclose revenue/property, booking conversion, or take-rate uplift. |
| Bottom line | The only dollar-anchored claim is 'hundreds of millions of dollars in realized marketing value through productivity/automation.' At $300M, after-tax (21%) = $237M = 11.9% of adjusted NI ($1,989M) on a gross basis — but it is described as already-realized run-rate, so the incremental next-FY lift is the growth in that value, assumed ~$125M -> $99M after-tax -> ~5.0% incremental adjusted EPS. Combined with explicit 'actions will favor margins' and self-service automation (>30% AI-powered, 60% faster onboarding), AI is a margin/efficiency story, not a revenue story. | The only monetizable claim is AI-enabled marketing value. Using the conservative $200M lower bound, after-tax value is $158M, equal to 7.94% of the adjusted 2025 consensus earnings base. Service automation is material operationally at >37.5M AI-powered self-service interactions, but cost/contact is undisclosed. |
| Reasoning | Consensus already models EPS up +28.1% (15.36 -> 19.68) on only +8.7% revenue and NI +24.5% (1,989M -> 2,476M) FY25->FY26 — i.e. heavy margin expansion baked in. A ~5% incremental AI EPS uplift sits comfortably inside that 28% modeled growth, and management's own 'savings...favor margins' framing is exactly what that consensus margin expansion represents. The AI efficiency gains appear embedded, not additive. Nothing in the disclosures points clearly above the consensus trajectory. | Consensus already models 2026 revenue of $16.009B, up 9.57% from 2025 consensus revenue of $14.611B and 8.66% above the supplied actual revenue base of $14.733B. Consensus EPS rises from $15.36218 to $19.67948, up 28.10%. The hard AI EPS math of 7.94% fits inside that expected EPS growth and there is no separately quantified revenue uplift above consensus. |
Rows highlighted where the two models disagreed.
QUANTIFICATIONS
traveler interaction data used for AI models: hundreds of millions (Q1 FY2026, topline)
“We're using data from hundreds of millions of travelers' interactions from shopping to reviews, servicing and more, to continuously improve our ranking and recommendation models.”
properties and exclusive properties at scale for AI supply advantage: nearly 3.7 million properties; 800,000 exclusive (Q1 FY2026, topline)
“We work with nearly 3.7 million properties, of which 800,000 are exclusive to us.”
lodging property count growth after AI-enabled onboarding: 10% (last quarter, topline)
“AI enables us to onboard partners faster. And last quarter, we grew lodging property count by 10% with our fastest growth outside the U.S.”
service interactions handled: more than 250 million service interactions a year (Q1 FY2026, bottomline)
“We handle more than 250 million service interactions a year with over half resolved through self-service.”
AI-powered self-service interactions: More than 30% (Q1 FY2026, bottomline)
“More than 30% of those are powered by AI and that number keeps increasing.”
AI conversation summaries language coverage: over 30 languages (Q1 FY2026, bottomline)
“We're automating conversation summaries in over 30 languages, enabling seamless handoffs with context across our global workforce and reducing new agent onboarding time by about 60%.”
new agent onboarding time reduction: about 60% (Q1 FY2026, bottomline)
“We're automating conversation summaries in over 30 languages, enabling seamless handoffs with context across our global workforce and reducing new agent onboarding time by about 60%.”
AI-enabled marketing value: hundreds of millions of dollars (Q1 FY2026, both)
“Overall, AI-enabled tools are driving hundreds of millions of dollars in realized marketing value through greater productivity and workflow automation.”
AI-driven property onboarding speed improvement: 70% faster (Q4 FY2025, topline)
“And in fact, some of the work we did on AI has sped up the time it takes to onboard properties. It's 70% faster than it was before.”
PAST (realized)
- In the first quarter, this translated into higher conversion at Vrbo and record attach rates on Expedia.
- As flights were canceled across the Middle East, AI helped us handle surging volumes at speed, allowing our human agents to focus on more complex time-sensitive issues.
- On Brand Expedia, for example, our refined recommendation models drove our best fourth quarter attach rates ever.
- This is already delivering tangible benefits.
- Our product and tech teams are using AI to design and build products, improving quality while shortening cycle time.
- Our supply teams are leveraging AI to speed up inventory onboarding teams, and our service team is using AI to resolve traveler issues faster and more effectively.
- And in fact, some of the work we did on AI has sped up the time it takes to onboard properties. It's 70% faster than it was before.
CURRENT (now)
- We're using data from hundreds of millions of travelers' interactions from shopping to reviews, servicing and more, to continuously improve our ranking and recommendation models.
- Our 2 most widely adopted features are our servicing agent and AI-powered filters and travelers who use AI filters return more often and convert at higher rates.
- More than 30% of those are powered by AI and that number keeps increasing.
- Answer Engine Optimization is now our fastest-growing channel.
- Overall, AI-enabled tools are driving hundreds of millions of dollars in realized marketing value through greater productivity and workflow automation.
- We're using AI to deliver more personalized experiences across all our brands.
- We're working with all the major platforms to capture traveler demand, ensuring our brands show up prominently in gen AI searches and function effectively with agentic browsers.
- At the same time, we're deploying AI internally to give our teams superpowers and make our offerings to travelers and partners even more competitive.
FORWARD (guidance)
- As we look ahead, we'll continue advancing our strategic priorities, thoughtfully deploying AI and driving value for all our stakeholders.
- I expect token cost to go up.
- We're going to be adding back skills that we need to optimize our AI plans.
- I expect we're going to continue to experiment over the quarters to come.
- Looking ahead, we're well positioned to build on our momentum as we execute our strategy and capitalize on the opportunities created by AI.
- And what the team is working on, and we'll have more to share later this year, is how we can use natural language and sort of AI to allow people to go from the trip planning all the way into the booking.
- While we'll be using much of the savings to strategically rehire in key areas like AI and machine learning, these type of actions will favor margins as well.
TRACK RECORD — PROMISE vs DELIVERY
—/100 (no quantified promises) no-quantified-promises 6 calls reviewed
Across these six calls Expedia discusses AI extensively (recommendation/ranking models, AI service agents, GenAI search/AEO, developer-assistant productivity) and later discloses several quantified AI-related results (cycle times down >20%, agent onboarding down ~60%, sites 30% faster), but never issued a forward-looking AI target carrying both a number AND a timeframe/milestone, so there is no auditable quantified AI promise to judge delivery against.
PRICED-IN (REFINED)
MEDIUMEst. revisions rising · Fwd P/E 14.7 · EV/Sales 1.5x
AI claim maps to Lodging, Advertising and Media Customers, Air
Estimate momentum is rising: analyst mix has shifted modestly toward more buys and fewer holds, price targets are higher over the last month versus quarter and year, and consensus revenue/EPS growth is being baked into forward years. Valuation is not especially rich for a mature travel platform at 14.7x forward P/E and 1.5x EV/Sales, so the rising AI/growth expectations are only partly priced in rather than fully reflected. AI upside would most plausibly flow through Lodging, Advertising and Media Customers, and Air via personalization, conversion, marketing yield, and operating efficiency.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
7Q4 FY20248Q1 FY20259Q2 FY20259Q3 FY20259Q4 FY202510Q1 FY2026
AI enthusiasm across 6 calls — trend ↗ rising
AI progressed from broad accelerator theme to quantified driver of conversion, service automation, marketing value, supply scale, and new traffic channels.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
8/10 qualitative impact material near-term · mixed evidence
Where AI matters: personalization, conversion, supply onboarding, service automation, marketing productivity
AI is already embedded in core travel shopping, ranking, filters, service agents, partner onboarding, and marketing workflows, with disclosed signs of higher conversion, record attach rates, 10% lodging-property growth, 30%+ AI-powered self-service, 60% faster agent onboarding, and hundreds of millions of dollars of marketing value. The upside is material but not yet transformational because incremental revenue lift is not separately quantified and AI-originated traffic remains small.
Caveats: AI channels remain small and may not scale into repeat direct customers; Answer engines could disintermediate Expedia brands and reduce organic/search traffic; Hotels and airlines may use AI to strengthen direct booking channels; Token and AI talent costs could offset productivity gains
AI DISRUPTION / CANNIBALIZATION RISK two-sided · 6/10
AI answer engines and travel agents can compress Expedia's role as a discovery and metasearch-like acquisition layer, shifting traffic control toward Google/OpenAI/Claude/Gemini or direct supplier channels and pressuring paid marketing economics. The model is still partly durable because booking, payments, inventory breadth, servicing, loyalty, and supplier relationships remain hard to fully automate or replace.
OPTIONS / MARKET STRUCTURE
option liquidity: good
proxy inputs — dollar-ADV $424M · beta 1.3 · px $226.37
source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.
CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders selling, institutions adding, management language 8/10 committed.
INSIDERS selling 1 open-market sell(s) vs 0 buy(s) — net distribution
INSTITUTIONS (13F) adding as of 2026-03-31: 142 new / 204 closed positions; 543 increased / 413 reduced; institutional ownership -3.25pp; -66 net 13F holders
MGMT LANGUAGE 8/10 committed Highly owned, metric-rich AI discussion with realized impacts, though indirect-channel growth is framed as small and longer-term.
commit “More than 30% of those are powered by AI and that number keeps increasing.”
commit “reducing new agent onboarding time by about 60%.”
commit “AI-enabled tools are driving hundreds of millions of dollars in realized marketing value through greater productivity and workflow automation.”
VERBATIM AI QUOTES
“Simply put, AI reinforces our core advantages and amplifies our execution against our priorities.”
— Ariane Gorin, Q1 FY2026
“First, AI enables better personalization at scale in our products. We're using data from hundreds of millions of travelers' interactions from shopping to reviews, servicing and more, to continuously improve our ranking and recommendation models.”
— Ariane Gorin, Q1 FY2026
“In the first quarter, this translated into higher conversion at Vrbo and record attach rates on Expedia.”
— Ariane Gorin, Q1 FY2026
“Our 2 most widely adopted features are our servicing agent and AI-powered filters and travelers who use AI filters return more often and convert at higher rates.”
— Ariane Gorin, Q1 FY2026
“Second, AI is a great tool to strengthen our supply advantage given the massive scale we operate at.”
— Ariane Gorin, Q1 FY2026
“AI enables us to onboard partners faster.”
— Ariane Gorin, Q1 FY2026
“Third, AI improves our post-booking experience.”
— Ariane Gorin, Q1 FY2026
“More than 30% of those are powered by AI and that number keeps increasing.”
— Ariane Gorin, Q1 FY2026
“Overall, AI-enabled tools are driving hundreds of millions of dollars in realized marketing value through greater productivity and workflow automation.”
— Ariane Gorin, Q1 FY2026
“We're using AI to deliver more personalized experiences across all our brands.”
— Ariane Gorin, Q4 FY2025
“On Brand Expedia, for example, our refined recommendation models drove our best fourth quarter attach rates ever.”
— Ariane Gorin, Q4 FY2025
“Finally, as gen AI changes how travelers do trip discovery, it opens up new growth opportunities for us.”
— Ariane Gorin, Q4 FY2025
“At the same time, we're deploying AI internally to give our teams superpowers and make our offerings to travelers and partners even more competitive.”
— Ariane Gorin, Q4 FY2025
“Our product and tech teams are using AI to design and build products, improving quality while shortening cycle time.”
— Ariane Gorin, Q4 FY2025
“Our supply teams are leveraging AI to speed up inventory onboarding teams, and our service team is using AI to resolve traveler issues faster and more effectively.”
— Ariane Gorin, Q4 FY2025
ANALYST QUESTIONS ON AI
Q (Q1 FY2026, Jed Kelly): Great. And just as a follow-up, just on the cost for the back half, is there anything we should be thinking about AI investments around higher token cost, token usage? Anything around there that could inflate the cost?
A: Definitely, we're seeing the use of AI and the associated costs, broadly speaking, increasing. I expect token cost to go up. We're going to be adding back skills that we need to optimize our AI plans. And I think that will put some upward pressure on costs but I think the teams have done a good job continuing to take costs out. So it's kind of a mix of productivity to pay for additional investments in AI.
Q (Q1 FY2026, Deepak Mathivanan): Ariane, I realize that the traffic from AI channels is still very small. I think you noted new users and conversion are all very good. But can you expand on the efforts to capture this traffic and potentially make them into repeat customers?
A: So with each of these AI channels, whether it's Gemini, whether it's ChatGPT, Claude, it's about making sure that our brands show up well there and then making sure that where we can, we then get traffic and we can land into our own brands.
Q (Q1 FY2026, Alex Brignall): Could you talk a little bit about that? We've heard the big hotels try to talk up their own native apps. But it feels like we're going towards a very similar place. I think Hilton alluded to this, a similar place to where we -- where Google has landed after a long time of experimentation. So could you talk about how you feel about that as the direction and how that would affect your business?
A: I think the decision and what we're seeing of a pullback in checkout reinforces our view that AI can be a powerful discovery layer but the actual booking and servicing is best handled by really a trusted, scaled provider.
Q (Q4 FY2025, Mark Stephen Mahaney): Ariane, could you just talk about the product or the features that you would want to try to roll out or are rolling out in order to really enhance the travel planning process on Expedia?
A: Right now, we've got an agent, sort of the AI agent in Hotels.com. What works the best is actually the point solutions like AI filters or property Q&A.
Q (Q4 FY2025, Deepak Mathivanan): Ariane, can you talk a little bit more about the product development efforts on the AI experiences side? Are you approaching it -- generally using the current LLM architecture and your cloud partners? Or do you think you need to fundamentally build new AI capabilities specific to travel, maybe with Expedia data in a unique way?
A: So in the product, I think of AI in a couple of ways. One is just in the existing flows, how do we use AI to make a better traveler experience. So that is personalization. It's better recommendations. It's better ranking models. It's more personalized content.
Q (Q4 FY2025, Lee Horowitz): I guess how are you thinking about the potential urgency to invest more aggressively into loyalty in your B2C business as some of these general purpose chatbots take on more of the customer relationship and the travel funnel?
A: Look, we always feel a sense of urgency to make sure that we're delivering more value and more trust to our travelers.