InMobi’s Glance reinvents itself as an AI-first shopping platform, expands to iOS with Glance AI

InMobi Group’s subsidiary, Glance, a screen lock discovery platform, has set out to transform itself into a provider of an AI-driven personalized shopping experience app. From its earlier

Android-only access, the company is launching the Glance AI app for iOS users to expand its base, while simultaneously morphing its existing Android version of Glance to integrate the

proprietary AI architecture.

According to the company, the team has been working on the product for the past 18 months, and the beta version was made available for US users earlier this year, reaching 1 million users

and demonstrating impressive retention rates. For iPhone users, downloading the app will be the only way to access it, whereas on newer GenAI-enabled Android devices, the new version of

Glance comes preloaded. Currently, the company has integrated nearly 400 D2C fashion brands in the US, while in India, it has partnered with Myntra and other D2C fashion brands. With an

initial focus on fashion brands, the company plans to extend this AI-personalized shopping experience to other categories such as beauty, pet care, accessories, and home furnishings in the

coming days.

What is new with Glance AI?

In its new form, Glance allows users to sign in with either Google or Apple email IDs, followed by a selfie and a set of Q&As around body type, ethnicity, style preferences, among others.

The AI model takes around three minutes at the backend to tune itself to users’ preferences and curate the shopping experience. Once done, the app displays styles using the photo to show

users how they would look and redirects users to the shopping sites of integrated third-party apps/sellers if they wish to buy any of the looks generated by AI-personalized styling. Speaking

to Fortune India, Naveen Tewari, CEO & Founder, InMobi & Glance, said the idea was to look at the world as a marketplace, with every brand and product being a potential client for the

company. “Think of it like a search engine. Any website can choose to be visible or not. Similarly, brands can choose to be discoverable.”

Explaining the non-discriminatory nature of the AI model, which shows products based on the user's profile and intent, he added, “The system reads product information and sends traffic back

to the brand or seller. We're not trying to retain that traffic or own the transaction. We are enabling better discovery.” Glance makes money from the actual sale of a product, and the

company expects to be on 100 million devices, both phones and televisions, by the end of this year.

InMobi says that to bring out Glance AI, learnings from past versions have played a significant role, driving insights to double down on commerce. Trained on global datasets—not just from

Glance but also from users of Roposo (a video shopping app)—feedback from the beta also played an important role in fine-tuning the product, according to the company. While not giving out an

exact number, Tewari said the company has spent hundreds of millions on developing this technology and has entered into a nearly billion-dollar long-term deal with Google (also one of its

leading investors) to scale its AI infrastructure using DeepMind and Gemini.

Explaining how AI is leveraged across the group, Naveen says that once the AI platform picks up a set of user-tailored products, those become eligible for bidding, allowing brands to bid for

higher placement within that chosen set. “However, bidding is only permitted once the model has already selected the right products for the user, and no brand can override this logic by

paying to promote an irrelevant product,” he adds.At the same time, Roposo—which is much smaller in scale and an India-focused video e-commerce platform—will also see Glance’s intelligence

being shared, but will continue as a direct seller.

Even as tech pundits predict zero-cost AI soon becoming a reality, which Naveen also seconds, he says it will undoubtedly bring down costs for companies. However, in the AI app race, he

says, “The winner will be the one with the self-learning model. This self-learning will improve core intelligence rapidly—intelligence about the economy, your transactions, your

preferences—so much that no one else will be able to catch up. The real barrier to entry won’t be cost but intelligence.”