Lenskart’s AI-Based Face Measurement: Turning Fit Uncertainty into a Proprietary Data Asset
Industry & Competitive Context
India’s eyewear market is large, growing and fragmented. Lenskart’s February 2026 shareholder letter puts the market at roughly ₹79,000 crore. The same letter says some estimates put the organised share near 23%, while the company argues that players with real technology, systems and scale account for closer to 10%. RedSeer projections cited in prospectus coverage point to about 13% annual growth, reaching roughly ₹1,483 billion by FY2030, with prescription eyeglasses holding around 73% of the market. Per RedSeer, as reported by PTI, Lenskart was India’s largest organised retailer of prescription eyeglasses by B2C volume in FY25.
The competitive relevance of face measurement lies in the problem it is presented as solving. Lenskart’s Google Play listing describes an AI tool that uses a selfie to recommend eyewear suited to the user’s face shape and to indicate the size that fits. The capability therefore addresses two questions at once: what fits, and what looks good.
Brand Situation Prior to Campaign
Lenskart was founded in 2008, began operating online in 2010 and opened its first retail store in New Delhi in 2013. The face measurement feature sits within a company that began online-first and later built a physical network. By Q1 FY26 that network had 2,806 stores, including 2,137 in India. In FY25 the company reported revenue of ₹6,652 crore and a net profit of ₹297.3 crore, after a ₹10.15 crore loss in FY24. Prospectus coverage also credits a technology team of more than 500 people with building and maintaining the core platforms.
The Google Play release text introduces Face Analysis as an AI tool that works from a selfie, and presents it alongside the launch of Blu zero-power eyeglasses for people working from home. That text carries no date.
Strategic Objective
Lenskart has not published a standalone objective for the face measurement feature in the sources reviewed. The objective has to be read from the company’s documented statements, so what follows is an interpretive synthesis rather than a stated goal.
The February 2026 shareholder letter states that AI allows the company to create more markets at lower cost, and lists a set of data feedback loops, including one in which every face scan improves the recommendation engine. The letter also says the company has built proprietary models for face shape analysis and frame recommendation, among other AI capabilities, as it moves from being an eyewear company toward an AI-powered eyewear and data company.
Read together, these statements point to three layers of intent. At the customer level, the feature gives a fit and style answer. At the channel level, it supports a more self-service store journey. At the enterprise level, it feeds a data asset the company describes as compounding. Only the company’s own descriptions support this reading, and no document explicitly ranks these layers.
Campaign Architecture & Execution
Digital layer. The app offers Face Analysis alongside a 3D Try On feature, through which users can virtually try eyeglasses and sunglasses at home. Twimbit, a research firm, describes a three-step flow (take a selfie, run face analysis, view recommended frames) and reports a later feature that combines facial analysis with customer behaviour data to suggest frames matching a user’s personality. That personality feature is documented only through this secondary source in the material reviewed.
Physical layer, international. At its Middle East entry, Lenskart said its flagship store would feature a 3D Try-On machine with virtual face-mapping that suggests suitable frames, and that the service was also available on its website and app. The company described its augmented reality technology as indigenously developed. At its sixth UAE store, in Dubai Mall, the machine was positioned as suggesting frames from a range of more than 7,000 options.
Physical layer, India. The Q2 FY26 shareholder letter points to an AI-enabled virtual stylist in stores as a driver of high-intent in-store purchases. The Q3 FY26 letter reports that since January, store visitors complete facial scans that create customer profiles, with more than 100,000 scans captured daily. According to the letter, a customer entering a store or opening the app is recognised by face shape and preferences, and the system is rolling out across all geographies. The same letter lists face-recognition self-check-in, live across India, and AI-based frame recommendation among its steps toward a self-service model.
The documented sequence runs from an app-based tool, to an in-store machine abroad, to store-wide scanning at scale in India. Over that sequence the capability moves from an optional feature to an identity and recommendation layer shared across channels.
Positioning & Consumer Insight
The documented positioning is functional. The app listing promises the right size and eyewear that will look good, from a single selfie. The sources reviewed contain no advertising creative or brand narrative built around the feature, so its positioning is carried mainly through product descriptions and in-store tools.
The company’s own survey data complicates any assumption that fit uncertainty was the dominant barrier. The letter says the main reason customers give for not buying from Lenskart is the absence of a nearby store. Face measurement is therefore one lever among several, and the company’s own materials weight proximity heavily. The company also states that most customers arrive through word of mouth and that its marketing favours pull over push.
Media & Channel Strategy
The feature is delivered through owned channels: mobile apps, the website and physical stores. An IPO note summarising the prospectus reports that frame recommendation and virtual try-on with size are the most used features on the company’s mobile applications. On the Q2 FY26 earnings call, the CEO placed AI virtual try-on within the technology-led omnichannel pillar of the business, and said AI powers frame recommendations when customers try on frames.
On marketing spend, the company reports that marketing fell from roughly 10% of revenue in FY23 to under 6% in Q3 FY26, and describes its focus as storytelling and consumer education rather than transactional advertising. The letter does not attribute this change to face measurement, so the two should not be linked.
Business & Brand Outcomes
Usage scale. Prospectus coverage reports that in FY25 Lenskart conducted 38.59 million virtual try-ons and 37.87 million face and frame size measurements for customers in India through its mobile applications. The IPO note attributes the 38.59 million try-ons to applications and websites combined. The two sources describe the channel slightly differently, so the underlying prospectus should be checked before the channel is stated definitively. These are counts of events, not unique users.
Store-level scanning. The company reports more than 100,000 face scans per day following the in-store rollout.
Digital influence. The CEO said on the Q2 FY26 call that 45% of Indian sales are digitally influenced. The letter’s glossary defines this as the share of India revenue from customers who engaged digitally through organic searches, social media or other online channels in the 90 days before purchase. The definition is broad and does not isolate face measurement or virtual try-on.
Customer experience. Net Promoter Score reached 80.9 in Q3 FY26, up from 78.5 in FY25. The letter credits AI-enabled personalization, but also cites better eye-test quality, faster returns and warranty handling, shorter waits, and store pick-up for online orders. The score therefore cannot be assigned to face measurement.
Financial context. The same letter reports Q3 FY26 revenue growth of 37.4% year on year and EBITDA margin expansion from 14.5% to 20.0%. These results are reported for the whole business and are not attributed to face measurement.
Strategic Implications
The capability is positioned as infrastructure, not a feature. The documented trajectory runs from an app tool, to in-store machines, to store-wide facial profiling, and the company describes feedback loops in which each scan improves recommendations. It also states that models trained on millions of Indian customers now serve operations in other countries. Strategically, this treats customer face data as a reusable asset that carries across markets.
Disclosure outpaces attribution. The most verifiable evidence is usage volume. Causal evidence is absent: there is no public conversion data, and the broad “digitally influenced” metric cannot separate face measurement from other digital touchpoints. Large usage counts show adoption, not incremental sales. The company itself notes that some disclosed information is sensitive and that future disclosure may vary.
Data governance is the least documented dimension. Scanning more than 100,000 faces a day in stores is a significant volume of biometric-adjacent data collection. The sources reviewed do not describe consent mechanisms, retention, access controls or regulatory treatment.
Proximity and fit are complementary. Given the survey finding on store proximity, face measurement works alongside the store network and remote services rather than substituting for them.
Discussion Questions
Lenskart reports 38.59 million virtual try-ons and 37.87 million measurements in FY25 but no conversion data. Design an attribution approach that would isolate the incremental contribution of face measurement to sales without relying on data Lenskart has not disclosed.
The company describes face scans as feeding a compounding recommendation engine. Evaluate whether such a data loop constitutes a durable competitive advantage, and identify what would erode it.
Lenskart reports 45% digitally influenced sales but defines the term broadly. What are the risks to managerial decision-making when a headline metric cannot be tied to a specific investment?
In-store facial profiling at scale creates value and governance exposure at the same time. How should a retailer weigh these, and what disclosures would build customer trust without undermining competitive position?
Lenskart describes its AR capability as indigenously developed rather than licensed. Discuss the build-versus-buy decision for customer-facing AI, and the conditions under which proprietary development justifies its cost.
Comments