Le Travenues Technology Limited (IXIGO) Earnings Call Transcript & Summary

Ladies and gentlemen, good day, and welcome to Le Travenues Technology, also known as ixigo Q1 '26 and '27 Earnings Conference Call hosted by DAM Capital. This conference call may contain forward-looking statements about the company, which are based on beliefs, opinions and expectations of the company as on date of this call. These statements are not the guarantee of future performance and involve risks and uncertainties that are difficult to predict. [Operator Instructions] Please note that this conference is being recorded. I now hand the conference over to Mr. Anmol Garg from DAM Capital. Thank you, and over to you.

Thank you, Dan. Good evening, everyone. On behalf of DAM Capital, I welcome you all to ixigo's 1Q FY '27 Earnings Call. We have with us Mr. Aloke Bajpai, Chairman, MD and Group CEO of the company; Mr. Rajnish Kumar, Director and the Group Co-CEO of the company; Mr. Saurabh Devendra Singh, Group CFO of the company. Now before I hand over the call to Aloke, Saurabh and Rajnish, I would like to highlight the safe harbor statement on the second slide of the presentation, and it is assumed to be read and understood. I now hand over the call to you, Aloke. Thank you, and over to you, Aloke.

Thank you. Good evening, everyone, and thank you for joining us for ixigo Q1 FY '27 Earnings Call. If I had to describe this quarter in one sentence, I would say that it demonstrated the value of having more than one engine of growth. The external environment was not particularly cooperative. The Iran conflict affected not just international flights, but also domestic ones through the second order impact of oil prices on the airfares. Aviation capacity remained constrained and airfares were at the highest levels we've seen. The train ticketing ecosystem continued to operate under policy constraints. Yet against this backdrop, our diversified platform continued to grow, our scaled businesses remain profitable, and we continued gaining share in all the categories we operate in. We finished the quarter with an all-time high GTV of INR 5,524 crores, up 19% Y-o-Y, revenue from operations of INR 356.75 crores, up 13% Y-o-Y and profit after tax of INR 34.24 crores, up 81% Y-o-Y. Our adjusted EBITDA came in at INR 29.24 crores, lower by 7% than the same quarter last year. This quarter also illustrates an important point about how we manage ixigo. We do not run each quarter to maximize the reported margin of that quarter. We try to maximize the long-term value of the platform while operating within disciplined financial guardrails. Our more mature businesses, across trains, flights and buses continue to generate operating leverage. We are choosing to reinvest a part of that into two areas where we see significant long-term opportunity, hotels and AI. That investment is visible in our current cost base. It is also intentional. Let me begin with buses, which continues to be the strongest part of our portfolio. Our bus business is now our largest vertical by contribution margin and continues to grow substantially faster than the broader market. Some of the category tailwinds are structural, continued highway development, growth in bus capacity and travelers increasingly using buses when confirmed train inventory is unavailable. Higher airfares also made ground transport relatively more attractive. But the more important story is that AbhiBus is outperforming the category. Our year-on-year GDP growth came in at 39%, much higher than that of the market, and it even exceeded 60% in 17 of the states we operate in, including large markets such as Delhi, Odisha, Uttarakhand, West Bengal and Himachal Pradesh as well as in several states in the Northeast. This is not growth concentrated in one region or driven by one campaign. It reflects deeper supply, stronger local execution and better products. We recently expanded our government bus inventory through an integration with West Bengal State Transport network. We are also solving some of the most basic anxieties associated with bus travel through product innovation. Our Roadside Assistance Program, for instance, provides eligible travelers with a replacement taxi to their destination in the event of a bus breakdown or accident. It is now available across 20 states and covers approximately 95% of our bookings with an AC taxi typically showing up within 45 to 60 minutes of a breakdown even in the middle of a highway. These interventions may sound operational, but they have strategic value. When we reduce uncertainty for the traveler, we improve trust. When we improve trust, we improve conversion, repeat behavior and word of mouth. And when better products are combined with broader supply, the marketplace compounds most efficiently. We are also expanding beyond consumer apps. BusBase from AbhiBus enables offline and smaller travel agents to access and sell bus inventory and ancillary services through a single platform. It is already contributing its share of transactions while allowing us to address demand that was previously difficult to reach digitally. So although buses is already our largest business by contribution margin and the fastest-growing one, we believe there is still plenty of road ahead, no pun intended. Let me now turn to flights. The Aviation Market faced an unusually difficult operating environment during the quarter. The Iran conflict and its second order impact on oil prices contributed to capacity reductions and sharp fare inflation. Domestic average transaction values or fares on our platform increased 22% year-on-year and 13% sequentially. International average transaction values increased 38% year-on-year and nearly 30% sequentially. Industry passenger growth was negligible domestically and international market contracted after the conflict. Despite this, our flight segment grew 4% and flight GTV grew 27% Y-o-Y. We have continued to be the fastest-growing flight OTA in India and have continued gaining market share. However, the current environment affects an important part of our growth funnel. A significant number of our new to flying users come from the next billion user base we serve in trains and buses. When the price gap between a flight and a train or bus journey expands so sharply, these users are more likely to defer or reconsider their upgrade to air travel. That effect is especially visible in Tier 2 and 3 markets. The encouraging part is that these trends can reverse quickly when the war is over and the fares moderate or capacity returns. However, we don't see this pain easing in the near term given that JAS quarter capacity cuts by both Air India and IndiGo has led to a year-on-year decline in overall flown passenger segments in the market. Airline commentary also suggests that meaningful capacity restoration may begin only when the festive period starts in the third quarter. Therefore, we remain cautious about aviation passenger growth in the near term while remaining constructive about the medium-term opportunity. We are also evaluating adjacent demand pools. ixigo's flight business has historically been leisure-led, but we meaningfully see usage among merchants, traders, small business owners, unmanaged corporate travelers who use GST numbers on our platform already. And we also see that in our bus and train business. This gives us a credible path into SME and corporate travel through a dedicated business product at some point, which can diversify the demand base over time. So we will explore getting into this segment in subsequent quarters. Coming to trains with that, the volume pressure there remains a category issue. Changes to Tatkal access for OTAs last year, lower waitlist inventory and additional authentication requirements that were imposed in the last few quarters have constrained the broader online train ticketing ecosystem growth. We are not going to predict the timing of policy or platform changes that are outside our control. We are planning the business without assuming any immediate relief, but we are very hopeful that OTP-based authentication mechanism will be rolled out to OTAs to ease the drop-offs we experience today, and our discussions on that front have been constructive. What we can control though is our performance within the category. Our share of the OTA train market has increased from approximately 60% a few quarters ago to 63% this quarter. We've chosen to preserve the structural profitability of this business too, while continuing to improve the post-book experience, customer service and adjacent monetization. Food on trains crossed 17 lakh meals during the quarter. touching 20,000-plus on many days. Metro bookings are growing month-on-month, and we are now live in five cities. Today, I also have an announcement to make here. We have gone live with Bharat Darshan rail packages on our train apps, offering affordable, all-inclusive tourist train experiences that encompass the journey, hotel, sightseeing or darshan at destinations in partnership with IRCTC and its tour operator partners. We are already seeing encouraging cross-sell from train audiences into buses and hotels as well. For buses, our own channels are, in fact, the fastest-growing ones and 90% of our hotel bookings are currently coming from our own captive user base. With that, let me talk about hotels, our next growth engine. Hotels has been our fastest-growing line of business over the last few quarters, and this is an area where our investment intensity is increasing. During the first quarter, we put 0.5 million Heads on Beds across our hotel network. We now have direct partnerships with more than 10,000 hotels across nearly 700 towns, including thousands of independent properties, and we are signing up several thousand more every quarter. For the first couple of years, we deliberately chose to learn before attempting to scale. The budget hotel category has complex problems around reliability, consistency and the gap between what a customer sees online and what they experience after arrival, which we refer to what you see is what you get in the previous quarter's call. We now have much greater conviction about the problems we need to solve, the supply we need to build and the peace of mind products that can differentiate our offering. The acquisition of a 54.66% stake in Brevistay accelerates this journey. Brevistay adds direct hotel relationships, feet on the street supply capabilities and expertise in flexible stay inventory. ixigo brings technology, AI-led distribution, a large captive traveler base and our HELLO partner platform. Together, we believe this can help us build direct supply faster, improve on-ground customer experience, access double-digit gross take rates and scale more efficiently across both conventional and flexible space. Hotels will have a longer gestation period than our other categories. We are currently a small player in a large market. But over the next four to five years, our ambition is to become the #1 discovery and booking platform for India's mid-market and budget hotels. That ambition will require investment, patience and disciplined execution. We intend to bring all three. The second major area of investment is AI, and Rajnish will talk more about that in a bit. Some of the AI investments will obviously appear in tech costs before they appear in productivity, growth or profits. That is the nature of building ahead of any major technology transition. Every company eventually reaches a point where it must decide whether to optimize for this year's margin or for the long-term value of what it is building. We believe the returns available from incremental investment in areas which are strategic for the long term remain important. We will remain disciplined, but it would be premature to optimize ixigo solely for near-term profitability growth when the opportunity to strengthen our competitive position is still so large. Our objective remains straightforward, grow faster than the categories in which we operate, continue gaining market share, invest behind areas where we have conviction and build a company whose advantages are durable and stronger with scale. With that, I will hand over to my friend and Co-CEO, Rajnish, to talk about what we have been building across AI, product and growth initiatives.

Thanks, Aloke. Let me start by mentioning two verticals where we see AI creating significant impact in product acceleration and growth. So at Prawaas 5.0, we launched busGDS.ai, an AI-first operating system and global distribution platform for bus operators. The platform combines AI-based inventory management, dynamic pricing, route planning, automated customer support and connectivity to multiple distribution channels. It is designed to allow an operator to set up or migrate in as little as 30 minutes. The second is Hotels, where HELLO, our AI-first hotel partner Extranet, is becoming an increasingly important supply acquisition and engagement engine. We are onboarding several thousand properties every quarter through the platform and now even support a WhatsApp-based hotel onboarding flow. HELLO supports both conventional stays and flexible stay inventory, and we continue to add tools and AI-led insights for hotel partners to improve their revenue potential on our platform. Our work over the last two years has helped us develop a much sharper hypothesis around what Product Market Fit in hotel means. We are now testing this PMF across select properties and developing peace of mind products, specifically for the budget category, which will be rolled out over subsequent quarters this fiscal, and we'll talk more about it then. But if we zoom out for a minute, AI is now moving from a capability race to an application and economics race. For ixigo, the key question is no longer whether AI can transform travel. It is how we use AI to transform the customer experience, improve the way the organization operates and build these capabilities sustainably at scale and at the right cost. Ixigo NEXT is the consumer-facing expression of this strategy. With Ixigo NEXT, we have introduced a new benchmark for what an AI-native travel application should be capable of doing. Traditional travel applications require users to translate their intent into rigid inputs such as destinations, states, filters, fare classes and predefined categories. With TARA, customers can express their needs naturally, including needs that may be complex ambiguous or difficult to capture through search form. The experience is not limited to answering questions. TARA can understand, intend, compare alternatives, generate interfaces dynamically, navigate users to relevant sections of the application and progressively move them from discovery towards action. Over the course of this year, we plan to extend this Agentic experience across our platforms, powered by our own voice optimized small language models and a growing suite of AI agents designed for multiple travel use cases, complex workflows and diverse tool calling capabilities. The significance of Ixigo NEXT goes beyond a new interface. it is beginning to change the quality of information we see about the customer behavior. Historically, travel platforms understood users largely through click stream and transactional data. We could see what users search for, which filters they selected and what they eventually book. But we could not always understand what they were actually trying to accomplish. TARA is now giving us a much richer view of customer intent at the top of the funnel. Customers can tell us in their own words, the type of experience they want, the constraints they face, the people they are traveling with and the trade-offs that matter to them. As TARA begins to handle a wider range of travel needs, we are learning not only what customers ultimately book, but why they make certain choices, what information they need before deciding and where existing travel journeys continue to create friction. These conversations are becoming a new source of product intelligence and a very valuable one. They allow us to identify unmet needs, discover new use cases and understand customer behavior at a level that conventional search and booking data cannot fully capture. We expect these insights to directly shape the next generation of features, workflows and customer experiences that we build across the ixigo Group. Over time, this can create compounding advantage. The more customers interact with TARA, the better we understand their intent. The better we understand their intent, the more relevant and personalized the experience can become. As the experience improves, customers have more reasons to use the platform for a broader range of travel needs. This is a shift we are trying to drive through ixigo NEXT from customers searching through travel applications to having an intelligent travel companion and understand, guide and eventually act on their behalf. Building this capability requires investment today. That investment is currently visible through higher expenditure on engineering talent, AI infrastructure, model development, token usage and orchestration. Some of these costs are front loaded. The initial development of the platform, the creation of evaluation systems and the training of our own small language models can be more resource-intensive than the subsequent refinement and deployment. Other costs, particularly inference, token and orchestration costs will recur and may grow as usage expands. However, we do not evaluate AI expenditure simply as an additional technology cost. We evaluate it against the economic output it generates. A higher token bill is not necessarily negative, if it allows us to ship significantly more without increasing headcount proportionately, automate operational workflows, resolve more customer interactions at a lower unit cost or accelerate product experimentation or improve conversion, engagement and retention. The relevant measures are, therefore, not only absolute AI expenditure. They include output per employee, cost per task completed, cost per customer interaction, speed of product development, incremental revenue generated and technology costs relative to revenue. And this is why we never report numbers like percentage of code written by AI because they are almost irrelevant today. The cost impact also naturally appears before the full productivity and revenue benefits become visible. We are investing today in platform development models, infrastructure, evaluation systems and organizational adoption. The benefits compound as more engineers teams work through the customer journey begin using the same underlying capabilities. AI is different from a conventional software investment because the same underlying intelligence layer can create leverage across multiple functions. It can improve developer productivity, automate larger parts of operations, accelerate product launches, reduce dependence on certain third-party software tools and create entirely new customer experiences. We, therefore, expect the rate of increase in technology expenditure to moderate as some of the initial platform and model development investments reduce, while the productivity and revenue benefits continue to scale. From a longer-term perspective, our objective remains for technology costs as a percentage of revenue to become more efficient year-on-year, although individual quarters may continue to reflect investments in new capabilities. The economics of the underlying AI stack are also evolving rapidly. Over the last few years, most of the industry's attention has been focused on the exponential improvement in foundation model intelligence. However, on several mature benchmarks, including Software Engineering Benchmark, also known as the SWE benchmark, leading models. [Break] However, on several mature benchmarks, including software engineering benchmarks, leading models are beginning to cluster within a narrower performance range. This does not mean model progress has stopped. It means that incremental difference between the best available model and another capable model may be becoming less consequential than it was during the earlier phase of rapid improvement. When intelligence was improving exponentially, being six months behind the frontier could create a significant capability gap. As performance begins to converge, a lower cost or open source model does not always need to be the absolute best. It only needs to be sufficiently capable for the specific task. At the same time, differences in costs can remain enormous. As an illustration, an Agentic workload consuming approximately 10 million input totals and 1 million output totals can cost under $2 on one model and around $150 on another, depending on the model and caching assumptions. That is close to a 90x difference in cost, even though the difference in usable intelligence for a particular application may be considerably smaller. In this case, less than 25. This changes the economic question. The objective is no longer to use the most powerful or the most expensive model for every problem. It is to use the right model for the right task based on the required intelligence, accuracy, reliability, latency and cost. This is why we are increasingly investing in our own AI harnesses. By harness, we mean everything that surrounds the underlying model, proprietary context, long-term memory, access to tools and data, workflow orchestration, domain-specific guardrails, evaluation systems, feedback loops and intelligent routing between different models. The foundation model provides raw intelligence. The harness converts that intelligence into reliable and specialized product or workflow at the appropriate cost. Simple and repetitive tasks can be routed to smaller open source or domain-specific models, more expensive frontier models can be reserved for problems that require deeper reasoning. Our own small language models can improve the performance, latency and economics of specific travel and organization use cases. The system can then continuously evaluate outputs and optimize for quality, speed, reliability and cost. For a vertical platform like ixigo, we believe that this is where a significant part of the long-term differentiation will reside. The underlying foundation models may become widely available, but they do not possess our proprietary travel context, transaction history, user preference, operational workflows or the deep understanding of Indian travelers. When this context is combined with the appropriate model and orchestration layer, general purpose intelligence can become a highly specialized travel intelligence system. We are applying this architecture not only to ixigo NEXT, but also across customer service, engineering, analytics, internal knowledge systems and workflow automation. Our objective is not to deploy one universal model everywhere. It is to build specialized systems that combine the right intelligence context and rules for each problem. This transition is also changing what exceptional talent looks like, give two people access to the same AI tools, the same models and effectively unlimited tokens, and they can still produce dramatically different outcomes. One may build something functional, the other may build something exceptional. When execution becomes increasingly accessible, outcomes are differentiated by distinctly human qualities. We believe seven of these qualities will become especially important, judgment, taste, empathy, agency, resilience, curiosity and vision. AI can generate options, but judgment determines which direction to take, which trade-offs to accept and which output to project. AI can produce endless variation, but taste determines what is coherent, differentiated and worth shipping. Empathy enables people to understand the underlying customer problem rather than merely completing a stated task. Agency determines whether someone takes ownership, uses every available resource and drives a problem to an outcome without waiting for detailed destructions. Resilience matters because working with AI is rarely linear or a straight line. Models misunderstand approaches fail and teams often need to reframe a problem several times before finding the right solution. Curiosity determines how deeply someone explores the available possibility space. Ambition determines whether someone nearly completes the brief or uses AI to imagine something substantially better than what existed before. This is increasingly how we think about talent and hiring at ixigo. We're not really looking for people who can use AI to execute faster. We are looking for people who can make better decisions, understand travelers more deeply, challenge existing benchmarks and use AI to expand the boundaries of what the company can build. AI is not making talent less important. It is making the distinction between ordinary and exceptional talent much clearer. Our thesis is therefore simple. As foundational model intelligence becomes more accessible, sustainable differentiation will increasingly move towards proprietary context, domain-specific models, efficient orchestration, superior product experiences and the quality of human judgment behind them. We believe these are precisely the layers where ixigo can build a durable competitive advantage. ixigo NEXT the most visible expression of this strategy today, but the opportunity extends across our entire organization. As these investments mature, we expect them to translate into deeper customer understanding, more personalized and agent travel experiences, higher productivity, lower unit costs, faster innovation and new sources of long-term growth. And with this, I'll hand over to Saurabh.

Thanks, Rajnish. Hi, everyone. Unfortunately, the flu has caught up with me this week. So, if my voice sounds a little off, it's just me, not a reflection of how I perceive the quarter. Now let us walk through Q1 FY '27. All figures are in rupees crores, unless specified otherwise, and year-over-year comparisons are Q1 FY '27 against Q1 FY '26. Starting with the headline numbers. Gross Transaction Value or GTV reached INR 5,524.33 crores, a 19% increase over INR 4,644.66 crores in Q1 last year. Revenue from operations stood at INR 356.5 crores, up 13%. Contribution margin grew to INR 144.94 crores, up 13% with contribution margin percentage at 40.6% against 40.5% last year. Adjusted EBITDA, excluding other income and ESOP cost was INR 29.4 crores compared to INR 31.4 crores, a decline of 7% Profit After Tax came in at INR 34.2 crores. Turning over to business lines. I would like to highlight that despite the challenging external environment affecting some of our business lines, we have continued to gain market share across all. For trains, we booked INR 2.44 crores train segment, down 8%. GTV was INR 2,138.6 crores, up 4%. Revenue stood at INR 141.04 crores, up 9%. Contribution margin rose to INR 52.74 crores, up 29% with a contribution margin percentage improving from 32% to 37%. Trains contributed 36% of the group's contribution margin. In flights, we booked INR 0.29 crores flight segment up 4% with a GTV of INR 2,341.84rores, up 27%. The revenue was INR 10.56 crores. Contribution margin was INR 41.04 crores at a contribution margin percentage of 39% against 42% last year. Flights contributed 28% of group contribution margin. On buses, passenger segment grew 33% to INR 0.89 crores. GTV grew 39% INR 9.43 crores. Revenue was up 34% to INR 102.55 crores. Contribution margin grew 28% to INR 54.22 crores at a margin of 53%. Buses are now the largest contributor of contribution margin, accounting for 37% of the growth. Now even though I don't usually mention the other segment, it's worth highlighting this quarter as the hotel build-out sits alongside businesses such as Loop. Others had a contribution margin of negative INR 3.06 crores against a positive INR 1.86 crores last year. That is a meaningful part of where this year's investment in hotel is showing up above the contribution margin, and this is a choice that we are making with our eyes open. In terms of operating metrics, Monthly Active Users were 8.5 crores, Monthly Transacting Users were 0.42 crores with our MTU to MAU ratio up to nearly 5%. App downloads for the quarter were INR 3.27 crores and 92% of our customer queries were solved by AI chatbot this quarter, up from 88%. Lastly, there isn't much to call out in terms of one-offs this quarter. But since I cover this every quarter, I will continue to do so this time, too. The one-offs and callouts for Q1 FY '27 include share of loss from FreshBuds and Squak, which are associates of INR 3.98 crores. For Q1 FY '26, the one-offs and call out comprised of share of loss of Fresh, which is associate of INR 2.33 crores. In closing, I'll leave you with a quote of attributed to Henry Ford. My wife reminded me about this on a particularly difficult day. Whether you think you can or you can't, you're right. to, we think we can. It takes conviction and a little. I'll now hand it over to the moderator for Q&A.

[Operator Instructions] The first question comes from the line of Mr. Anmol Garg from DAM Capital.

A couple of things that I wanted to ask. One is that there is 80% sort of sequential jump in our advertisement spend, which has taken our overall margins lower. Just wanted to understand, is this onetime spend? Or is this an ongoing trend? And how should we look at the ad spend going ahead? Will it go back to that INR 18 crores, INR 19 crores kind of quarterly levels? Or is it like that INR 32 crores, INR 33 crores kind of level is something which will be going on? And in addition to that, if you can sort of also answer that how should we look at the overall margins of the company going ahead? That is my first question.

This is Aloke here. So, if you look at the brand marketing spend that we typically do, it is very seasonal, right? Like if you look at Q1, typically, it's a little heavier on marketing spends. This year, we had IPL, where we had confirmed ticket doing some activities. We had AbhiBus also upping the ante a little bit on the marketing side there. But it's also typically the peak season of the year where marketing spends are typically higher, right? Is this going to be the new level? No, because there could be quarters where we take it to these levels. There could be quarters where we choose not to play because they may not be either the right season or the strategy is to not utilize spends in that particular quarter, right? So, by way of guidance, I mean, we don't typically give guidance on the brand marketing spend. But what I can tell you is that if you look at the overall customer inducement cost that we have, right, which is brand performance and discounts put together, we try to stay in the 4 percentage of the GTV range, right, whenever we can. Now there could be quarters where we go above this, and there could be quarters where we stick to this range.

Right, Aloke. And so in addition to that, how should we think about margins going ahead? Just wanted to understand that since our contribution margin is somewhere kind of maintained, will there be a flow-through because of operating leverage in terms of the margin improvement in the coming quarters?

Yes, great question. So if you think about what's happening, right, Rajnish talked about a lot of investments we are making on the AI side in terms of model training, in terms of our own small language models, harnesses, et cetera, being built. And you will see that some of the tech expenditure going up reflects that. But some of that is obviously front-loaded, right? If you look at hotels, which I called out as our next growth engine, we have spent compared to the same quarter last year, if you look at, and the other line of business includes hotels for us. And if you see the contribution margin that, that line contributes versus last year, right, there is almost INR 5 crore delta on that, right, which Saurabh also talked about in his portion of the call. So you can think of it this way, right? We are using the operating leverage that our core business is generating to invest in AI and hotels growth, right? That's the way to read this quarter. And some of those investments could be more like front-loaded tech and AI investments. Some of them could be things that investments that we will probably continue because hotels is something we're still building out, right? So that's where the EBITDA margin answer comes to.

And just to add on to what Aloke is saying on your first question, which is remember that if you see last year, too, it was a similarish factor. on advertising and branding. So just have that as the base. So it is something which we do strategically. And the other part of it is as the higher take rate businesses become more dominant as a part of our portfolio, you'll see some shift there, but that naturally flows advantageously to the EBITDA.

Sure, just one more question. So in the flight segment, while the volume growth is understandable given that market has also been a little weak. However, our GTV growth was quite strong at 27%. So just wanted to understand has the pricing for us moved like 20% growth in our tickets? Because if we look at the other player in the segment there, it was the pricing growth was nearly about 10-odd percent. So what's driving almost like a 20% kind of growth in pricing for the tickets on our platform? And in addition to that, like you also indicated that there's some capacity reduction, which has happened by Indigo and Air India. So from that context, should we expect that sequentially, there could be a slight decline in going ahead in 2Q within the flight segment?

Yes. So, see, first of all, on the fair, right, in the narrative, I had called out that domestic average transaction value that we are seeing. Now average transaction value is a function of two things, which is the segment fair as well as the number of passengers who travel. But average transaction value that we have seen increased 22% year-on-year on domestic. And on international, we have seen an increase of 38% year-on-year. Now the increase for us versus someone else could be a function of mix of what kind of segment mix they have on international versus domestic as well. Secondly, it could also relate to whether there is more long haul or more short haul on the international side. So, there could be reasons why this could vary across OTAs. But just to set the context right, see, what has happened in the broader market is that if you look at DGCA data, right, there was a 2% sort of a growth year-on-year in passenger segments, right? Now on that kind of overall market growth, typically think of it as following, right? Like if you're an NB-U user and you have to pay X thousand extra to buy a flight ticket and suddenly now that price delta is almost double of that, it becomes very difficult for them to sort of pay up and move up the ladder during this environment, right? And we saw this in the past also that whenever such crazy fare hikes happen, right, the NB-U funnel gets affected. But as and when this subsides, the NB-U funnel is also the first to come back very quickly as we saw during COVID and other such moments where the fares start going down and you start seeing the upside start to come in. So we remain hopeful that this Iran crisis at some point will get fully solved. And of course, it's an on and off situation as we have seen. But assuming that it does and the oil prices also correct, some of the fares might normalize again. The bigger issue right now, at least in the JAS quarter is that both Air India and Indigo has taken a conservative stance in terms of supply, right? So they have cut back on routes. And obviously, the overall capacity, and you can track the DGCA numbers for July yourself, but you will see that the overall capacity has shrunk out there. And if the capacity is shrinking at that pace, it could be even depending on how August, September plays out, so I'll not make any forward-looking guesses around it. But depending on how much that shrinkage happens, it will impact all the players, right, not just us. The good thing is that we have managed to continue growing even in that environment, at least in the June quarter. I think in the quarter to come, again, it will depend on whether supply comes back. Right now, I can tell you that Air India, we have seen close to 20% cut. Indigo, we've seen like a 10-ish percent cut, right? So on the supply side, if that continues, it's going to be a very tough situation for everyone who's in the Air OTA business.

Our next question comes from the line of Pankaj Mehendiratta from BofA Securities.

I have two questions. Firstly, thanks for the disclosure on hotels. Can you talk a bit more about the 0.5 million number? What share of your hotel customers is coming from your existing funnel versus a stand-alone customer acquisition? And a related line on that is the investments that are being done out there on the contribution margin or below contribution line. How much of the investment is being absorbed across hotels and AI, if you could separate that number out there, please?

Sure. So Pankaj, I think I called out this number, but I'll repeat that. So see, if you see the hotels business, right, we are hitting where a point where we have critical mass of supply. We have a critical mass of passengers who are opting for this in our funnel. And though I called it out, I will tell you again that 90% of the bookings we are seeing are coming from users who are already registered on our platform, right, which means they were already our users and not people who came and registered for the explicit purpose of buying the hotel. So I think for us, like that's a very important metric because it means our dependency at this point on third-party funnels to aid this growth is quite low. not saying that we will not go and look at those avenues for growth later on. But at least at this point, we've managed to reach this 0.5 million ads on weeds numbers largely from our own funnel, right? And the repeats are also coming out nicely because now we have, remember, we have close to two years of data. And even in terms of scaled up data, we have three, four quarters of sort of growth where we have seen people come back and buy the second time, the third time through our platform. So we understand what the behavior looks like. Therefore, we're gaining more confidence on how this market share evolution will go from here. And we will not be shy, obviously, of investing in aiding this growth, at least for new bookers to experience hotels on our platform, if we need to incentivize those new bookers, we are doing that already. And also from a product experience point, I think I'll let Rajnish talk about that in a bit on what the plans are there. But we'll not be shy of investing in this category because clearly, we see that for the user base we have, which is more attuned to book mid-market and budget hotels, let's say, three star and below, I think we are destined to be the largest player in this space over the next four to five years, let's put it that way. And we'll not leave any stone unturned to get there. Over to Rajnish.

Yes. I mean, like we said, I think like we have been, I think we were asked this question about disclosure on what numbers we were doing, and I think we were trying to avoid that because we are really not confident about the product market fit like I talked about earlier, which we are fairly confident now. And so like Aloke said, this is the time where whenever we find that product market fit and we have a conviction, we don't shy away from investing. So that's definitely going to happen now going forward. And we also have like this understanding of exactly what those pain areas are. Like one of the key things was identifying exactly what kinds of peace of mind products that can actually work well within hotels as a category. I think this is something that really worked well for us on flight buses and train, but I think the same thing was very hard to crack on the budget category of hotels, right? So now we have cracked it. So you should expect that these kind of products and features will start coming out really soon in the coming quarters. And we've already kind of created an experience now with ixigo NEXT within hotels, which is basically so unique that I haven't really seen products of this nature even globally. So we are quite confident that we should be able to kind of scale this product simply because I think the product is sticky enough for people to kind of come back to it and the word of mouth to kick in as well.

Just to add on that, Pankaj, so that you don't panic in that sense that; look, it's ixigo. We have behaved like ixigo throughout. So whatever we do, as I could have taken a call going aggressive two years ago, whatever we do is based on what we get as a feedback on the product working or the conviction that we build on. So in that sense, everything is weighed on the conviction. We feel very strong conviction in areas, and we are getting deeper into that. It doesn't mean that if the world changes, we continue going aggressive for the sake of being aggressive. We have guardrails, and we plan to grow within that guardrail.

That exactly was my second question out there. Could you give a broad idea on what the guardrails or benchmark on operating leverage that you have, which you are willing to reinvest? I mean one way to think about it is currently, the EBITDA decline, let's say, on a sequential or a Y-o-Y basis is about 120-odd bps on your revenues. Is that broadly the sense because, I mean, that is something that you can probably invest on one quarter. But for a full year or, let's say, for next two years, if somebody were to think of it that way, what is that number that either on absolute terms or margin terms that you could broadly guide the Street out there?

No, I'll explain it in a very different way, which might not answer your question. Look, if I start giving you. Let me put it this way, and I said the same in probably two quarters ago or a year ago or probably more. I don't remember which quarter. But what I have said there is, look, in case what Rajnish said in case doesn't seem to be happening, the product which we build is great, we put it out. It is not working that much. You get a much higher margin in that sense. If what we are doing works, we will put capital behind that. So what it is, is very different, and I need to open up my next five years plan to you if I want to give you any other detail. And probably when I do it to you, I do it to the market. So can't do that. But what I'm just saying is, look, it works both ways. So if you see a higher margin, it means that some of our plan and the feedback that we are getting, we are not getting that stronger feedback. And I'll pass it on to Aloke to add on to that.

Yes. I just want to add one thing here that if you look at ixigo's entire history of the last 20 years, we have never bought market share to win. We've basically gone and built the best product experience, built the best customer experience. And we have only fueled that with marketing spends or customer acquisition spends once we hit the right product market metrics and the right retention metrics. That's how we build trains. That's how we build flights, that's how we build buses, and that's how we'll build hotels.

That's very assuring, and Saurabh, if I could just squeeze one more question out here. And from a broader perspective, if you look at the other large elicit player out there, they also seem to be going after the market share construct, especially when the market is soft out there. From that perspective, if you say that you are also gaining market share, essentially, who is it coming at the expense of, right? Are you getting more people from offline to online? Or are you squeezing the other broad players in the ecosystem? How should we think about the market construct, let's say, in this particular quarter or maybe towards the end of this year out there? And I will stop out there as my last question.

Yes, I'll answer this very differently, okay? India is a very unique market because if you look at categories like trains and flights, the penetration online of bookings is higher in these categories. And obviously, the take rates are lower. But the big opportunity lies in categories like buses and hotels where even though the take rate is high, contribution margins are good, the penetration is still low, right? I mean if you think about mid-market and budget, the penetration is 10% or probably sub that, even if you go to below two star, it will probably be sub-10% in many small towns. And when you are in a category where penetration is that low, even if you are the #1 player, the category does not exist yet for you, right? And you have to create the category. Now the advantage we have versus any other player who has tried to do this in the last 20 years, the only difference we have is we have more users at this point of time than any other player in the OTA space. So we are starting off with a huge advantage on day one. And that is why we called out the 90% coming from existing funnel number that we are not going to ride this by buying users from 10 other places, right? We are going to make sure the users who come on our platform end up buying hotels from us. And that's a problem that we have a better handle of as we speak today than we did two years ago when we were starting to build out hotels.

Our next question comes from the line of Swapnil Potdukhe with JM Financial.