Indians Are Beginning to Search for Travel in the Language They Speak: MakeMyTrip

Data from 20 lakh voice conversations on Myra, MakeMyTrip’s Gen-AI Trip Planning Assistant, indicates that voice queries are beginning to reflect a richer, more contextual form of travel intent, with early trends showing significantly higher adoption across Indian languages and more conversational search behaviour compared to text.

 Gurugram, March 9, 2026 – MakeMyTrip, India’s leading online travel company, today shared observations from Myra, its Gen-AI powered Trip Planning Assistant, pointing to an emerging behavioural trend in how Indian travellers are beginning to interact via voice. While Myra’s user base is still growing and now delivers more than 50,000 conversations daily, the initial data suggests that voice is beginning to enable a more expressive, contextual, and linguistically inclusive form of travel discovery, one that is different from how users search via text.

 How Voice and Text Queries Are Beginning to Diverge

Even in early usage patterns, the contrast between how users type and speak their travel intent is visible. Majority of text searches are in the neighbourhood of 3-4 words, compressed and keyword-driven, such as “Goa hotels cheap” or “Delhi Mumbai flight.” Voice queries are beginning to look considerably different. Nearly 23% of voice queries exceed 11 words, compared to just 7% in text, as users tend to naturally articulate destination proximity, amenities, budget, group size, and dates within a single spoken interaction. For instance, “Show me affordable hotels in North Goa near the beach with a pool” or “2 adults and one kid, 3 nights from 14th January, budget under ₹15,000 per night.”

 Across several query categories, early data points to voice being used noticeably higher than text. Date-specific queries show the most pronounced difference, at 3.3x higher on voice, with users naturally saying “26th December to 29th” or “next Friday to Sunday” rather than typing compressed date formats.

 Informational queries, where users are looking for guidance or explanation around certain processes or services, are indexing 2.7x higher on voice, suggesting that users are beginning to turn to conversational assistance for questions beyond transactional search. Examples include, in Hindi, "सऊदी का वीज़ा कैसे मिलेगा?" (How to get a Saudi visa); in Kannada, "ಜಪಾನ್‌ಗೆ ಪ್ರಯಾಣಕ್ಕೆ ಹೋಗುವ ಮೊದಲು ಯಾವ ದಾಖಲೆಗಳನ್ನು ಸಿದ್ಧಪಡಿಸಬೇಕು?” (What documents should be prepared before going on a Japan trip?); in Marathi, "कोल्हापूर ते दिल्ली जाण्यासाठी कोणती ट्रेन आहे व किती दिवस लागतात आणि ती कोणकोणत्या वारी असते?" (Which train is available to travel from Kolhapur to Delhi, how many days does it take, and on what days of the week does it run?); in Malayalam, എനിക്ക് ആനയിൽ നിന്ന് കൊച്ചിയിലേക്കുള്ള തത്കാല ടിക്കറ്റ് തിങ്കളാഴ്ച ഉള്ള ടിക്കറ്റ് ആണ് വേണ്ടത്. അതിന് വേണ്ടി ഞാൻ എന്താണ് ചെയ്യേണ്ടത് (I need Tatkal ticket from Anai to Kochi for Monday. What should I do for that?); and in Tamil, "வீல் சேர் எப்படி புக் பண்ணனும்னு சொல்லு?" (Tell me how to book a wheelchair?).

 Location-specific queries account for 25.1% of all voice searches and index 1.5x higher on voice than text, with users naturally expressing proximity in phrases such as “near beach,” “walking distance from Golden Temple etc. Similar patterns are visible across languages. For instance, in Bengali, "হ্যালো মাইরা এটা এই হোটেলটা গুয়াহাটি রেল স্টেশন থেকে কতদূর" (How far is this hotel from Guwahati railway station); and in Telugu, "గోవా బాగా దగ్గరలో తక్కువ రేట్లు మంచి రూమ్స్ కావాలి" (Need good rooms at low rates very close to Goa).

 Voice Is Opening Up Travel Search Across India

One of the more telling early signals in the data is how voice is beginning to enable users to search in the language they naturally think and speak. While English dominates text searches, voice interactions are considerably more linguistically diverse with English representing far lower % of voice queries. The gap points to something simple: users who are most comfortable in their own language are gravitating towards voice. For example, for Malayalam, voice trumps text by 46 to 1; for Tamil this ratio is 36:1 and Telugu 32:1.

 For users who found typing a barrier, the shift is less about technology and more about expression. Voice lets them describe exactly what they want, in their own words, without compression. Code-mixed users, who blend Hindi and English, are among the most expressive in early observations, averaging 10.5 words per query. A search like “Manali mein 3 nights ke liye hotel chahiye with mountain view and breakfast” captures nuance and preference in a single breath, something that text search has rarely accommodated.

“What we are beginning to see through Myra is encouraging. Voice is starting to give a new set of users, those who are most comfortable in their own language, a more natural way to search and plan travel. For someone in Kochi or Coimbatore who thinks in Malayalam or Tamil, being able to simply speak their requirements, rather than type them in English, changes the experience meaningfully. It is still early, but these initial signals point to voice having the potential to make travel planning more inclusive and accessible across India,”

Rajesh Magow, Co-founder and Group CEO, MakeMyTrip.

 Premium Travellers and Multi-Constraint Queries

Early data also points to premium and elite traveller segments using much longer sentences while using voice on Myra, often combining star category, amenities, group size, and budget into a single spoken request. For example: “5-star villa in North Goa with private pool, 6 bedrooms, for 8 adults under ₹50K per night.” While these highly layered, multi-constraint queries represent a smaller share of total searches, they offer an early indication of how voice may be better suited to capturing complex travel intent than keyword-based text search.

 MakeMyTrip has been deeply invested in AI and machine learning for several years, embedding intelligence across the travel lifecycle.  From inspiration and discovery to search, booking and post-sales support, AI is integrated at every stage. These proprietary models, built on large language architectures and rich travel-intent data, power Myra. The AI Trip planning Assistant now facilitates over 50,000 conversations daily across multiple languages, including Bengali, Hindi, Kannada, Malayalam, Marathi, Tamil, Telugu and English. More than 45% of queries on Myra are coming from Tier-2 and smaller cities.