Weekly Digest
— in which the Indian OTA battle is quietly bifurcating into companies with contractual …
Editor's note
"You just don't get stuck anymore," Naval said about building with agents. He's right — I don't get stuck on execution the way I used to. What I do get stuck on is harder: whether what I'm building is the right thing, whether the plan holds up, whether I've thought past the first week. The tools removed one kind of friction and made another kind more visible. The projects that worked are the ones where I treated that second friction as the real work, not a prerequisite to skip. — Paarth
The Brief
Two things happened this week that appear unrelated but point at the same thing. MakeMyTrip launched guaranteed early check-in and late check-out across 10,000+ properties — a monetised ancillary layer built on pre-negotiated supply contracts. IndiGo announced Vision 2030 with a fleet doubling and a widebody international push. The common thread is contractual depth: the companies moving fastest right now are those whose supplier relationships give them the right to build product features, not just display inventory. MMT's check-in feature cannot be replicated without the underlying hotel contracts; IndiGo's Stretch and transit hub story requires OTA co-marketing agreements to surface properly. What has shifted this week is the nature of the competitive gap — it is no longer primarily a UX or AI gap, it is a supply-contract gap, and UI investments made without the underlying contractual architecture to support them will produce features that look equivalent but convert worse. The question the team should be sitting with is not what features to build, but what supply agreements need to be in place before those features are worth building.
Our read
The Indian OTA battle is quietly bifurcating into companies with contractual supply depth — MMT on hotels, Ixigo on rail and bus — and those without it; Cleartrip has roughly one quarter before MMT's early check-in feature is read as the new category baseline by hotel partners and users alike, at which point the ask shifts from 'can we build this' to 'can we get hotels to agree to this', a much longer cycle.
Continuing pattern
For two consecutive weeks, the signals have pointed at the same underlying dynamic: Ixigo building vertically integrated supply infrastructure (Brevistay, Vestra.AI) while MMT converts its supply depth into new ancillary product categories. Last week the pattern was visible in hotel inventory architecture; this week it is confirmed in hotel contract structure — MMT's check-in slot feature is only possible because the supply agreements already existed. The pattern is strengthening: what began as a hypothesis about Ixigo's accommodation ambitions now has a clearer competitive shape, with MMT demonstrating that supply-contract depth is the actual moat, not the feature itself.
By the numbers
The share of MakeMyTrip's new check-in slot bookings that are for early check-in — a single number that reveals the scale of overnight-arrival travel in India and validates why this ancillary category will become a baseline expectation, not a premium differentiator, within 12 months.
Mint ↗Skift · 8 Jun 2026
IndiGo's Vision 2030 targets a fleet of 550+ aircraft, 200 million annual passengers, and 300 billion ASKs — roughly doubling current scale — with international capacity rising from around 30% to 40% of total, powered by A321XLR and A350 widebody inductions. The airline is also premiumising via expanded Stretch business-class cabins and positioning India as a transit hub to divert connecting traffic away from Gulf and Southeast Asian hubs.
Rajesh Magow frames MakeMyTrip as having already resolved the AI threat question — positioning its proprietary GenAI assistant Myra and OpenAI partnership as proof that discovery-layer disruption has been absorbed — while the India IPO remains structurally complex due to dual-listing restrictions and Mauritius incorporation, with the company running two parallel listed entities as a transitional structure.
Industry lens
Any OTA without a named, differentiated AI product faces a growing perception deficit as MMT consolidates the 'AI-native OTA' story ahead of a domestic listing that will generate sustained Indian media and investor coverage.
MakeMyTrip has launched confirmed early check-in and late check-out booking slots across 10,000+ domestic properties and 1,000 international locations, available in 3-, 6-, or 9-hour windows purchasable at reservation time — claiming an Indian OTA first. The platform uses flight, train, and bus data to surface the relevant time window automatically, and early adoption data shows 77% of slot bookings are for early check-in, reflecting the scale of overnight-arrival travel in India.
Industry lens
OTAs with deep direct-contracted hotel supply have a structural advantage in launching ancillary time-slot products; platforms reliant on GDS or aggregated inventory will find this feature architecturally difficult to guarantee at booking time, creating a supply-quality differentiation that is invisible to users until they actually travel.
Railway Minister Ashwini Vaishnaw announced that a new IRCTC website will go live by July 15, 2026, in direct response to student complaints about CAPTCHA failures, payment timeouts, and booking errors during Tatkal windows. Simultaneously, Aadhaar-based OTP authentication becomes mandatory for all online Tatkal bookings from July 1, with a new website interface to follow on July 15 — the two changes arriving in sequence, not together.
Industry lens
Any OTA surfacing IRCTC rail inventory faces a compliance deadline that is also a UX crisis for users — platforms that pre-communicate the Aadhaar linking requirement and guide users through it before July 1 will reduce booking abandonment and build credibility with a segment that plans around train availability, not flight availability.
Ixigo's AbhiBus unit is actively courting travellers priced out of air travel by high fares, launching product features targeting new bus riders and positioning intercity bus as a beneficiary of the current airfare environment. This builds on ixigo's April 2026 integration of AbhiBus into ChatGPT alongside its flights and ConfirmTkt rail products, and its Q1 FY26 bus GTV growing 81% YoY — the same rate as flights.
Industry lens
OTAs with a single-mode or air-dominant book of business face structural customer leakage during sustained high-airfare periods; platforms with owned multi-modal inventory across bus and rail are the only ones whose CAC yields a traveller who stays regardless of which mode wins on a given route.
Air India is piloting meal-opt-out 'lite fares' on short domestic routes including Delhi–Amritsar, Delhi–Chandigarh, and Bengaluru segments, with potential base fare savings of up to ₹250 per passenger. The move is driven by fuel cost pressure and a domestic market share that has slipped below 25% against IndiGo's near-65% dominance — and mirrors unbundling already executed by Air India Express and now extending to the full-service mainline brand.
Why it matters
Fare unbundling at the full-service carrier level forces OTA search and results UX to handle a new ancillary dimension — meal add-on — on Air India inventory alongside the existing IndiGo buy-on-board model, and OTAs that can't display and upsell meal options at search will show artificially higher Air India base fares relative to direct booking.
Noida International Airport (IATA: DXN), India's second NCR airport located in Jewar, begins commercial operations on June 15 with IndiGo as inaugural carrier on five domestic routes; Akasa Air joins June 16 on two routes; Air India Express has withdrawn entirely due to cost-cutting pressures, leaving the airport's first phase as a two-carrier operation with a domestic UDF of ₹490 — roughly 4x IGI's ₹129.
Industry lens
OTAs without a multi-airport display logic for Delhi NCR (IGI vs DXN) will surface incomplete or misleading fare comparisons for a 12 MPPA catchment from June 15 — the longer the fix is deferred, the more search trust erodes for NCR-origin customers.
Indian Railways is replacing its 1986-era Passenger Reservation System with a new AI-powered platform launching in phased migration from August 2026, capable of handling over five lakh ticket transactions per minute. The new system integrates waitlist confirmation prediction at 94% accuracy (up from 53%), is closely tied to the RailOne app which already processes 9.29 lakh daily bookings, and continues the IRCTC authentication hardening trajectory including Aadhaar verification for Tatkal and general reserved windows.
Why it matters
The PRS replacement is not just a capacity upgrade — it is a platform transition that resets the API and authentication surface OTAs depend on for rail inventory, and any OTA integration built on legacy PRS assumptions will require re-certification with CRIS before August or face booking failure on migrated train inventory.
Skyscanner CEO Bryan Batista is revisiting agentic booking as a mechanism to reduce friction at the handoff point between metasearch and OTA partners — an experiment the industry tried and abandoned in the early 2020s that Batista believes AI agents can now make viable. The company has the largest global flight search market share, is expanding into B2B via the Longtail acquisition, and is actively investing in India through partnerships with airlines and OTAs including MakeMyTrip, with a localised Hindi-language platform already live.
Why it matters
Skyscanner's India push via MakeMyTrip means the metasearch layer in India now has a well-resourced global actor deepening its integration with one of India's largest OTAs — if agentic booking removes the OTA redirect step, inventory relationships and direct API depth determine who surfaces in agent-mediated results, not brand.
John Cutler argues that organisations mandating AI adoption are systematically destroying the psychological conditions — agency, trust, and dignity — required for teams to actually innovate with AI. Drawing on Bandura's social cognitive theory, he distinguishes between individual agency (how a person adapts to AI on their own terms) and organisational affordances for agency (the environment that enables or suppresses it), arguing that most corporate AI mandates conflate the two and produce compliance theatre rather than genuine capability.
Why it matters
Top-down AI adoption mandates without squad-level experimentation space will produce vanity metrics — AI usage rates, prompts per sprint — that mask whether the tooling is actually improving decision quality or product outcomes, creating a false confidence that compounds when leadership reads adoption rates as capability.
Using Uber's uSpec as a case study, the piece argues that design systems are now consumed by AI agents as much as by human engineers — and that the quality of agent-generated code is determined not by documentation written for humans but by how precisely contracts are encoded in token naming, prop types, and component composition rules. A system with semantic token naming (e.g. `color-feedback-critical` vs. `blue-600`) and typed prop constraints generates accurate agent output; one relying on Storybook prose generates plausible-but-wrong output that compiles.
Industry lens
Product teams in travel that are accelerating front-end velocity with agentic coding tools are doing so on the assumption that their component libraries are machine-readable; teams whose systems were built for human implementation will not discover the gap until agent-generated components start diverging from design intent in production.
Mark Pincus, whose Zynga achieved 8 of 10 major launch hits reaching over a billion players, presents a 'Proven, Better, New' framework for consumer product success: start with what is already proven to work, make it meaningfully better until 10 of 10 people say they would use it, then layer in one genuinely new element. The episode, tied to his forthcoming book, also surfaces the counterintuitive principle that instincts are right 95% of the time but specific ideas are wrong 75% of the time — and that the path to ambitious outcomes runs through deliberate constraint, not expansive ideation.
Industry lens
Travel platforms shipping AI-native booking interfaces before their baseline search and checkout flows meet a 10-of-10 preference threshold among existing users are building on an unstable foundation — the 'Proven, Better, New' sequencing suggests that agentic travel interfaces will only convert if the underlying inventory, pricing transparency, and booking reliability are already demonstrably superior.
Addy Osmani argues that individual prompt-crafting is being replaced by 'loop engineering' — designing automated systems with five components (scheduled automations, worktrees for parallel agents, skills files encoding project knowledge, tool connectors, and sub-agent verification) plus persistent memory that run coding agents continuously without human intervention. Boris Cherny, head of Claude Code at Anthropic, is cited as evidence this is now operational practice, not theory.
Why it matters
When coding loops replace human-in-the-loop prompting, the quality ceiling shifts entirely to the 'skills' files — structured documentation of project conventions — meaning teams without mature, machine-readable component specs and design system contracts will produce higher-variance, lower-quality output at greater token cost than teams that have invested in documentation infrastructure.
John Maeda's June 11 post extends his 2026 Design in Tech Report thesis — that design is shifting from UX (user experience) to AX (agentic experience) — with a direct framing question: what happens to product design when the user can bypass 90% of a carefully crafted interface? AX redefines the designer's role from arranging screens to specifying intent, flow, and boundaries for AI-driven systems that act on the user's behalf, with the interface becoming a low-frequency fallback rather than the primary interaction surface.
Industry lens
Travel platforms whose product value is concentrated in search and booking interface design face the most acute AX disruption; platforms that own proprietary inventory, loyalty data, or post-booking service relationships have something an agent can surface as differentiated — pure front-end OTAs without those assets become invisible in an agent-mediated funnel.
David Hoang argues design leadership is entering a structural compression driven by two simultaneous forces: experienced design leaders retiring faster than they can be replaced, and the AI-native generation not yet ready for senior seats — creating a two-year window before a seismic talent gap becomes visible at the exec level. He maps three tectonic stages: initial uplift (now), sustained thrust where AI-native leaders get installed at challenger companies first, and a new mountain range where multiple design exec archetypes emerge.
Industry lens
Travel platforms that treat design leadership hiring as a continuity appointment — replacing like-for-like on the previous era's archetype — will find themselves in Hoang's 'leaders become irrelevant' category by 2028, when AI-native exec norms become legible and the gap from incumbent orgs becomes permanent.
Signals Worth Keeping
“MakeMyTrip is still weighing next steps for a domestic Indian IPO and thinks it has already won the AI debate.”
“As of June 2026, Air India's domestic market share has faced downward pressure, dipping below the 25% threshold, while arch-rival IndiGo continues to solidify its dominance with a market share nearing 65%.”
“IndiGo will begin commercial operations on the opening day of June 15, becoming the first airline to operate flights from Noida International Airport.”
“With 53 per cent of domestic travellers arriving before 9 AM and 54 per departing after 3 PM, the feature lets travellers confirm an early or late room slot at the time of booking.”
“Aadhaar-based OTP authentication will be mandatory for online Tatkal bookings from the 15th of July this year.”
“Your system is already making promises. The question is whether those promises are written down clearly enough that something other than you can read them and get the right answer. The difference shows up in the output. It always has, and it's just easier to see now.”