Corporate Travel Manager Intelligence Brief

Executive Summary

The Phocuswright Conference 2025 arrived at a decisive inflection point for managed corporate travel. The dominant signal across nearly every session was not that AI is coming — it is already here, already deployed, and already demonstrating measurable ROI in the back office. Phocuswright's own August 2025 survey of travel executives found 83% are actively using AI, with roughly 75% deploying it for internal operations and 50% for consumer-facing products. For corporate travel managers and TMC executives, this is no longer a future-state planning exercise. Peer organizations are moving, and the change management question is no longer whether to adopt AI-native tools but how quickly to scale them.

The landmark GBT-SAP Concur partnership dominated the business travel session and reverberated throughout the conference. Panelists framed it not as a market disruption but as a consolidation move with broadly positive implications for corporate buyers — particularly larger enterprises already straddling both platforms. However, mid-market TMCs face real structural pressure, and buyers should expect a renegotiation window to open as the integration matures. Caroline Strachan of Festive Road offered buyers the most direct counsel of the conference: "just breathe" and use the partnership announcement as an opportunity to revisit SLAs, platform commitments, and service-level baselines before the combined entity establishes new defaults.

The distribution landscape is fracturing in ways that directly complicate travel program management. Google AI Mode's announcement of direct agentic hotel booking partnerships with Booking.com, Expedia, IHG, Marriott, and Wyndham signals that travelers increasingly have paths to transact outside approved booking tools. AI models embedded in browsers, messaging platforms, and consumer apps create ambient booking capability that corporate policy enforcement frameworks — built around OBT channels and approval flows — were not designed to capture. The traveler experience imperative is now inseparable from policy compliance strategy: programs that fail on experience will lose travelers to AI-powered out-of-program alternatives.

The most actionable finding from Phocuswright 2025 for this stakeholder group is the convergence of NDC complexity and post-booking AI automation. NDC implementation has pushed call center handling times up 20–30% (per Acai Travel's Riccardo Vittoria), while experienced human agents are retiring faster than new ones are trained. AI-powered post-booking servicing platforms — Acai Travel, Wenrix's DeepFlow, and Amadeus-backed solutions — are demonstrating 60–93% automation rates on exchanges, refunds, disruptions, and schedule changes. For TMC executives, this is the clearest near-term ROI story in the conference. For corporate travel managers evaluating TMC partners, it becomes a vendor assessment criterion: which TMCs are deploying these tools, and does their performance data reflect it?

Key Findings

1. The GBT-Concur Partnership: Strategic Rationale and Buyer Implications

The GBT-SAP Concur partnership was positioned by Evan Konwiser (Chief Product & Strategy Officer, American Express Global Business Travel) as a direct response to years of accumulated customer feedback: enterprise clients want Concur's platform to work in ways that actually meet traveler needs, and GBT is now positioned to collaboratively build more integrated solutions. The partnership's logic is structural — Konwiser described the two companies as sharing tens of thousands of customers, making the overlap a natural consolidation rather than a hostile disruption.

Caroline Strachan (CEO, Festive Road), speaking from a pure buyer consultancy perspective, observed that mid-market TMCs have more cause for concern than enterprise buyers. Her advice to travel managers was calibrated: the deal is a natural fit, and buyers who were already on both platforms should expect net positive outcomes over a multi-year horizon. The critical near-term window is contract renegotiation. As the combined entity establishes its go-to-market posture, buyers who proactively revisit SLAs, integration roadmaps, and service commitments before the integration hardens will be in a stronger negotiating position than those who wait.

Steve Singh (Executive Chairman and CEO, Spotnana), representing a competing TMC infrastructure provider, offered a useful framing for evaluating the deal: "partnerships are an output of strategy, not the strategy itself." For travel managers, this implies the right evaluation criterion is the underlying product and service integration quality — not the announcement itself.

2. Post-Booking AI: The Clearest Near-Term ROI for TMCs

Two Innovation Launch competitors — Acai Travel (People's Choice Award winner) and Wenrix (DeepFlow, runner-up Travel Innovator of the Year) — presented compelling, data-backed cases for AI automation in the post-booking servicing stack, and both are already deployed within TMC environments.

Riccardo Vittoria (CEO, Acai Travel) quantified the problem precisely: NDC implementation has pushed call center handling times up 20–30%, experienced agents are retiring faster than new ones can be trained, and the travel industry collectively spends $200 billion annually managing post-booking complexity — changes, cancellations, disruptions, and schedule changes. Acai's three-agent AI team (supervisor, front-office, back-office) deployed inside TMC call centers has demonstrated 60% reductions in average handling time alongside 20–30 basis-point improvements in customer satisfaction (CSAT). Since its seed round, Acai has grown 50% per month in the two most recent months, reaching $2M ARR with sales cycles under three months and customer acquisition cost below $10,000.

Amir Balaish (CEO, Wenrix) approached the same problem from the execution angle rather than the conversation angle. His core finding: the industry has automated roughly 40% of simple in-flight ticket servicing tasks, but the remaining 60% — the complex edge cases — consumes 82% of agent time. DeepFlow, trained on over 50 billion real-world servicing data points, has achieved 93% automation at one travel company and over 90% accuracy on refunds and exchanges at CWT specifically. Integration time is cited at one month.

Robert Buckman (SVP Solution Consulting Americas, Amadeus) corroborated the post-booking thesis from the GDS perspective, identifying travel disruption management as the next major frontier likely to be meaningfully solved by agentic AI within a few years — and pointing to Acai Travel (a minority Amadeus investment) as the signal investment.

For TMC executives: this is the most evidence-rich, near-term AI deployment case at the conference. For corporate travel managers evaluating TMC partners: begin asking specifically which post-booking AI tools are deployed, and request handling time and CSAT benchmarks as standard evaluation criteria.

3. Corporate Travel Policy Is Broken — and AI Will Expose the Cracks

Caroline Strachan introduced the "Disney effect" framing — corporate travel has two customers (the company paying and the traveler booking), analogous to Disney marketing simultaneously to parent and child. Her structural diagnosis was pointed: corporate travel policy is largely a "red herring." Policies filled with the word "must" are routinely ignored because enforcement mechanisms do not exist at scale.

Evan Konwiser reframed policy not as a compliance instrument but as "the manifestation of the company's culture into what you can buy as a traveler." The Venn diagram of what travelers want and what companies will pay for, he argued, has more overlap than most programs assume. AI-driven integrated solutions will surface that common ground more effectively than policy documents ever have.

The "be where I am" principle articulated by Strachan is directly actionable: travelers want to interact from within the tools they already use — Slack, Microsoft Teams, WhatsApp — rather than switching to a separate OBT. Amadeus's Microsoft Teams booking capability was cited as a proof-of-concept. Steve Singh projected that within three to five years, an AI will know the traveler, their family, their employer, and every city they visit — delivering predictive and proactive support as an "agentic travel agent for life." The implication for policy architecture is significant: rigid OBT-centric compliance frameworks will erode as AI-native booking interfaces multiply.

Strachan also described a change management framework for buyers adopting AI innovation — a spectrum from "traditional ways of working" to "transformational ways of working" with a "comfort zone of change" in between. She cited a pharmaceutical client that mandates AI training for all employees with consequences for non-compliance. Konwiser observed that consumer use of agentic AI for daily life is organically training the entire ecosystem, accelerating corporate risk tolerance faster than any internal program can.

4. Agentic Commerce Is Moving Faster Than NDC Compliance

The payments panel ("AI Agents, Real Transactions") revealed that the infrastructure for AI-initiated corporate travel transactions is further developed than most travel managers realize. Visa's Trusted Agent Protocol (TAP) and Stripe's Agent Commerce Protocol (ACP, co-developed with OpenAI) both went to production in late 2025 and are already handling transactions. Both protocols include shared payment tokens with spend limits, time limits, and revocation capabilities — the core controls that corporate payment programs require.

Jennifer Watkins (ARC) noted that agentic commerce was absent from airline distribution conference agendas in spring 2025 but dominated every panel by late 2025. ARC processes approximately 20% of transactions via NDC today; order-based processing (the next step toward full modern retailing) could be supported by airlines in 2026. The look-to-book ratio explosion under NDC — further amplified by agentic AI generating logarithmically more search queries — was identified by Sabre CEO Kurt Ekert as a systemic cost challenge. Ekert's counter is intelligent caching at petabyte scale, which he positioned as a capability only a few platforms in the world can provide — meaning TMC and GDS selection now has a caching competency dimension.

For corporate travel managers: lodge virtual card programs, Approved Travel Authority frameworks, and payment controls designed around human-initiated bookings will need to be reviewed against agentic scenarios. The protocols exist; the question is whether your TMC and payment provider have integrated them.

5. The NDC Adoption Gap Is a Real Program Risk

Kurt Ekert (Sabre CEO) was unusually candid on NDC timeline reality: the IATA 2030 offer-and-order mandate is unlikely to be met on schedule. He described the transition as "open heart surgery" for airlines, noting that most top 20–30 carriers have not committed to a formal order transition timeline. NDC adoption among brick-and-mortar and TMC agencies — as opposed to OTAs — remains slow due to the attendant workflow and back-office integration changes required.

The practical consequence for managed travel programs is a bifurcated content world for an extended period. Travelers may encounter richer, more personalized NDC content in direct and OTA channels that does not surface through TMC OBTs still running on Edifact-first infrastructure. Ekert characterized this as an exponential growth opportunity once the industry begins collaborating more effectively — but "collaborating more effectively" is doing heavy lifting in that sentence.

BoomPop (Travel Innovator of the Year, Healey Cypher presenting) highlighted the group and corporate event planning dimension of this content gap. Hotels cannot be booked for groups of more than 10 rooms online, forcing planners into RFP processes that yield less than 1% conversion. BoomPop's AI-native platform, built on MCP architecture, already processes approximately $100 million in annual event spend with 450 brands on platform. For travel managers overseeing MICE and events alongside transient, this represents a category where AI-native tools are already further along than the main transient booking stack.

6. The New Travel Seller Landscape Threatens Program Compliance

Chase Travel's acquisition-led vertical integration strategy and Hopper Technology Solutions' B2B pivot (now representing over 90% of Hopper's revenue) represent a structural shift in who is building and deploying travel booking capability. Chase Travel targets $12–15 billion in annual gross bookings, growing at approximately 20% CAGR. iSeatz powers the American Express Travel platform and estimates the US financial institution travel marketplace at $45–50 billion in gross bookings — a fraction of total market but growing rapidly.

The compliance implication: high-spending employees who hold premium cards with Chase, Amex, or fintech travel platforms face booking alternatives that offer compelling rewards, financing, and experience-layer perks outside any TMC program. Dakota Smith (President, HTS/Hopper) described Hopper's agentic AI assistant handling approximately 3 million live customer service conversations annually — a servicing capability that matches or exceeds many mid-market TMC operations.

For travel managers: the approved supplier list and OBT mandate face pressure not just from AI chat tools but from financial institution travel platforms that corporate travelers already trust and use for leisure. Evaluating how your card program interacts with your TMC arrangement — and whether corporate card rewards can be directed toward in-program behavior — becomes a more urgent priority.

7. AI Adoption Readiness: The Real Benchmark Is Internal Productivity, Not Booking Chat

Wall Street analysts on the "Street Talk" panel (Mark Mahaney/Evercore, Lloyd Walmsley/Mizuho, Emma Taylor/Barclays) offered a grounding perspective: the most measurable near-term AI returns across the travel sector are in call center automation, software product development velocity, fintech, and advertising optimization — not in AI-powered booking interfaces. Emma Taylor specifically identified back-end AI ROI as the clearest near-term investment thesis for travel companies.

This was corroborated across the conference. Sabre reported 15% developer productivity improvement using Google Gemini and Vertex AI. Skyscanner's Bryan Batista reported 20%+ engineering productivity gains. Agoda's Omri Morgenshtern noted that 95% of developers across Southeast Asia and India are already using AI internally — near-universal back-office adoption. American Airlines' Neil Geurin (VP Global Sales) confirmed the airline is focusing personalization efforts on the post-booking, pre-travel window, where latency constraints on its 40 million customer records are manageable.

For TMC executives: the first AI ROI story to bring to your CFO is not the chatbot in the booking flow. It is agent handling time reduction, servicing automation rates, and software development velocity — all of which have existing benchmarks and can be measured against baseline within 90 days of deployment.

8. Traveler Experience Is the New Compliance Lever

Phocuswright's own research on the "indulgent explorer" traveler segment — those with high per-person, per-day spend, averaging 4.7 trips per year — found that six of the seven top research and planning resources for this high-value cohort are offline: calls to suppliers, travel advisors, and print media. OTAs appear mid-list. This cohort is twice as likely to book packages, driven by trust, white-glove service, and the sense that high spend levels feel too significant to handle online.

For corporate travel programs managing senior executives and high-frequency road warriors, this maps directly to a service gap. The travelers whose compliance matters most — and whose out-of-policy spend is most costly — are precisely those most likely to self-service through trust-based channels outside the OBT. Steve Singh's benchmark of 95% accuracy on the first recommendation (versus the current state where he estimates most search delivers 95% irrelevant content) sets the bar for what AI-powered personalization in managed travel needs to achieve before it becomes a genuine compliance tool.

Strategic Implications

For corporate travel managers:

The GBT-Concur partnership creates a renegotiation window. Whether you are on both platforms, one, or neither, the market shift creates leverage for buyers to request clearer integration roadmaps, richer data reporting, and explicit SLA commitments before the combined entity normalizes its pricing and service structure. Use the next 12 months.

Policy architecture built around OBT mandates and approval flows is not keeping pace with ambient booking capability. Travelers already have access to AI-powered booking through browsers (Gemini in Chrome), messaging platforms, and consumer apps. The response is not stricter mandates — it is closing the experience gap so that the approved channel is also the preferred channel. "Know me, be where I am, tell me only what I need to know" (Strachan's traveler need formulation) is the product specification for your OBT evaluation criteria.

NDC content gaps represent a near-term traveler satisfaction risk. If richer, more personalized airline content is visible in direct channels but absent from your OBT, frequent travelers notice. Begin requiring explicit NDC content roadmaps from OBT and GDS providers.

AI-generated group and event planning tools have outpaced transient booking tools in maturity. BoomPop's $100M in managed event spend with a 4.9/5 rating and NPS of 85 is a benchmark that most MICE programs cannot match through traditional RFP-and-spreadsheet processes. Evaluate AI-native event management platforms as an immediate, low-risk entry point for AI adoption in travel management.

For TMC executives:

Post-booking servicing automation is the most evidence-backed AI deployment case in the market. Acai Travel (60% handling time reduction, 20–30bps CSAT improvement) and DeepFlow/Wenrix (93% automation, 90%+ refund/exchange accuracy at CWT) provide the benchmark data for business cases. Sales cycles are short (under three months), integration is fast (one month for DeepFlow), and cost comparison against manual handling is straightforward. This is the first AI capability to bring to RFP responses.

The competitive pressure from AI-native disruptors is asymmetric: incumbents face the highest switching friction from corporate buyers, but that friction is shrinking. The "AI-only OTA" scenario presented by Mike Coletta (Phocuswright Research) — 10-second refunds, zero fees, automatic disruption management, 5% commission — was considered possible within 10 years by a majority of conference attendees. TMC executives who frame this as a distant threat are misreading the velocity.

Traveler data graph construction is your most durable moat. Steve Singh's projection that truly personalized recommendations require a comprehensive data graph around each traveler — employer, family context, city preferences, behavioral history — and that 95% accuracy on the first recommendation is the bar, defines the capability that will differentiate TMCs from AI-powered alternatives. The question is not whether to build this; it is whether to acquire, partner, or build.

Action Items

Priority 1 — Immediate (0–90 days):

1. Initiate GBT-Concur contract review. Before the integration hardens, request explicit integration roadmaps, SLA commitments, and data portability provisions. Regardless of your current platform mix, the renegotiation window is open now.

1. Pilot a post-booking AI servicing platform. Issue an RFI to Acai Travel, Wenrix/DeepFlow, and your primary TMC partner's own AI capabilities. Request handling time baselines, automation rate benchmarks, and CSAT impact data. Sales cycles are under 90 days; a pilot can be running before Q3.

1. Audit your OBT for NDC content gaps. Request a written roadmap from your OBT and GDS provider covering NDC content parity. Identify specific airline content (fare families, ancillaries, seat upgrades) that your travelers can see in direct channels but cannot access through the OBT.

1. Evaluate AI-native event management. Request a demo from BoomPop or comparable platforms. Measure against your current MICE RFP conversion rate and planner time-per-event. This is a standalone, low-risk AI adoption entry point that does not require changes to your core TMC arrangement.

Priority 2 — Near-term (90–180 days):

1. Review corporate card and payment controls for agentic scenarios. Work with your payment provider to understand how virtual card programs and spend controls interact with Visa TAP and Stripe ACP. Define what "approved agentic booking" looks like for your program before it happens without a policy.

1. Run a traveler experience audit against financial institution travel platforms. Identify which premium corporate cards your highest-spending travelers hold and audit what Chase Travel, Amex Travel, and comparable platforms offer that your approved program does not. Close the gap on at least one differentiator (refund speed, itinerary change ease, mobile experience).

1. Define your AI adoption change management framework. Model Strachan's "comfort zone of change" spectrum. Identify one segment of your traveler population (e.g., tech-sector teams or frequent international travelers) as an AI pilot cohort and instrument them with different booking tools or policy nudges before rolling out broader changes.

1. Request AI-powered disruption management benchmarks from your TMC. Using the Acai/DeepFlow benchmarks as reference points, ask your TMC to report its current average handling time for disruption scenarios, its servicing automation rate for exchanges and refunds, and its roadmap for AI-augmented servicing. Make this a standing quarterly metric.

Priority 3 — Strategic (180+ days):

1. Develop a traveler data strategy. Identify the data inputs needed to approach Singh's "95% accuracy on the first recommendation" benchmark. This requires combining corporate HR/travel profile data, behavioral booking data, and loyalty program data. Define who owns this effort internally and which vendor partners can help build the data graph.

1. Prepare a multi-channel booking policy for the agentic era. Begin drafting policy language that addresses AI-initiated bookings, out-of-channel transactions via financial institution travel platforms, and group bookings via AI-native platforms. The legal and compliance frameworks for agentic corporate travel do not yet exist; the organizations that draft them first will have first-mover advantage in implementation.

Sessions to Watch

1. “Business Travel's Next Move” — Evan Konwiser (GBT), Steve Singh (Spotnana), Caroline Strachan (Festive Road). The most directly relevant session for this audience: covers the GBT-Concur partnership rationale, corporate policy reform, the "Disney effect" traveler experience framework, and the 3–5 year timeline for agentic travel management. Konwiser and Strachan's debate on policy philosophy alone is worth the watch.

1. Acai Travel — Innovation: Launch 2025 (People's Choice Award) — Riccardo Vittoria. The clearest business case for post-booking AI automation you will find in the conference. Data-backed: 60% handling time reduction, $2M ARR at 50% monthly growth. Watch alongside DeepFlow/Wenrix for a complete picture of the post-booking AI competitive landscape.

1. Wenrix DeepFlow — Innovation: Launch 2025 (Runner-Up) — Amir Balaish. The execution-layer counterpart to Acai's front-office story. Critically, DeepFlow is already deployed at CWT, which provides a direct peer benchmark. The 93% automation rate and one-month integration timeline are the key numbers to take into TMC conversations.

1. "AI, Innovation and the Next Era of Distribution" (Amadeus) — Robert Buckman. Essential context on how a major GDS and travel technology provider is thinking about AI infrastructure, MCP protocol standards, and the disruption management opportunity. The Amadeus-Anthropic MCP alignment signal matters for understanding the technology direction of large platform vendors.

1. "Competing in Travel's Next Era" (Sabre CEO Kurt Ekert) — Ekert's candor on NDC timeline reality ("IATA 2030 mandate unlikely to be met"), the look-to-book ratio explosion under agentic AI, and the Concierge IQ product announcement are all directly relevant to TMC infrastructure decisions. The framing of agentic AI as "a new distribution channel analogous to how OTAs emerged 30 years ago" is the strategic lens for a 10-year planning horizon.

1. “AI Agents, Real Transactions — The New Travel Economy” — Clara Liang (Stripe), Gloria Colgan (Visa), Jennifer Watkins (ARC). For travel managers and TMC executives who need to understand what agentic payments infrastructure actually looks like, how controls work, and what the B2B use case timeline is. The protocols (ACP, TAP) are in production now. This session provides the vocabulary for conversations with your payment provider.

1. “The Industry Says BoomPop Is the Travel Innovator of the Year” — Healey Cypher. The most immediately deployable AI solution for managed travel programs. Group and event planning represents $7,500+ corporate offsites planned daily, all currently done via spreadsheet and manual RFP. BoomPop's MCP-native architecture, $100M in managed event spend, and NPS of 85 make this the lowest-risk AI entry point for travel managers who need an internal win.

1. "From Discovery to Departure — Designing Effortless Experiences" (Sutherland roundtable) — Neil Geurin (American Airlines), Jeffrey Wright (Allianz Partners). American Airlines' frank acknowledgment that it has 40 million customer records but cannot parse them within the sub-second window consumers will tolerate — and its pivot to focusing personalization on the post-booking window — maps directly to where TMCs should concentrate their near-term AI investment. Wright's framing that empathy and follow-through in service delivery (cashless medical care versus reimbursement) can be more loyalty-generating than any algorithmic offer optimization is a useful corrective to over-indexing on booking AI.

*This brief was produced from session content recorded at The Phocuswright Conference 2025, San Diego. All data points, quotes, and projections are drawn directly from speaker presentations and Phocuswright research as delivered at the conference. Briefing prepared March 2026.*