India’s Unlikely SaaS Success Story Begins Its Second Act With AI

It was 2011. A candidate had just cleared his interview at Wingify. The process was supposed to end the usual way—with an offer letter. 

Instead, Sparsh Gupta, the Co-Founder and CEO, found himself making an unusual pitch. Not to the candidate, but to his father.

“I had to speak to his dad, convincing him why his son should not join TCS or any other service company–but join us,” Gupta says.

“This was a point where building a product company in India was not a cool thing. It wasn’t celebrated. It wasn’t something people were aware of.”

Wingify, however, was doing just that—building a product in India, pricing it in dollars, and selling it to customers in the United States and Europe. 

It followed an unconventional path for an Indian tech company in the early 2010s. It remained bootstrapped for over 15 years. 

By the time Wingify was acquired by Everstone Capital for $200 million in 2025, it was profitable and had crossed $50 million in annual recurring revenue. 

Around this period, Co-Founder Paras Chopra, who had started the company in 2009, exited with 90% of his stake, while Gupta, who had joined in 2011, continued as CEO.

The biggest change in my life after Wingify exit is that I've started taking Uber Premier instead of Uber Go :)

— Paras Chopra (@paraschopra) February 6, 2025

The company has since acquired French-based A/B testing leader AB Tasty, creating a combined entity with over $100 million in annual recurring revenue and a presence across 11 global offices. 

It also acquired Blitzllama, a Y Combinator-backed AI-driven user insights startup. And in the era where generative AI is often seen as a threat to SaaS, Wingify is entering one of its strongest phases ever. 

Why AI Offers Wingify More Promise than Ever 

The current wave of AI has shifted the software bottleneck from creation to decision-making. While generating copy, layouts, images, and flows is no longer difficult, determining which of those variations actually drives outcomes has become harder.

“You still have to figure out what the right image is to show to the right person. It should not just be good, it should convert to metrics,” Gupta says.

Wingify’s core product, VWO, is built around this problem. It enables teams to run controlled experiments, measure user responses, and make decisions based on observed behaviour. Different versions are shown to different segments, and performance is evaluated against defined metrics.

Gupta argues that the abundance of AI-generated variations introduces inconsistency without a system to evaluate performance. “LLMs can guide, but they do not have access to your business data to decide what actually works,” he says.

To address this, Wingify has embedded variation creation within its platform to reduce the time between idea and test, while retaining experimentation as the evaluation mechanism.

“The throughput of A/B testing has gone up because the friction of creating variations is gone.”

VWO integrates generative AI across the experimentation lifecycle through its Copilot layer. It analyses behavioural data, such as heatmaps and session recordings, to suggest hypotheses, generate variations, and recommend audience segments. It can also detect friction points and propose changes. 

The integration of Blitzllama enables analysis of qualitative feedback from surveys and support conversations, identifying sentiment and themes at scale.

Together, this allows teams to connect what users do with why they do it, shifting optimisation from manual workflows to a continuous and AI-assisted system. Although Wingify began operating in the early 2010s and now functions in an entirely different technological landscape, its underlying philosophy has remained unchanged.

<img src="https://public.flourish.studio/visualisation/28222961/thumbnail" width="100%" alt="visualization" />

Building the Experimentation Layer

Wingify was built at a time when the internet was just reaching the masses, and businesses were focused on acquiring users rather than understanding them.

“The money was completely going towards attracting more and more visitors,” Gupta says.

Marketing teams were optimised for traffic, not behaviour. As search advertising scaled, companies could bring in users—but struggled to understand what they did once they arrived, or why they failed to convert.

Google attempted to address this with Website Optimiser, linking ad spend to on-site experimentation. “They did a fantastic job at educating the market, but did not fulfil the needs completely,” Gupta adds.

Wingify approached the problem differently, making experimentation the core product rather than an extension of advertising. 

VWO allowed teams to create variations, run experiments, and measure outcomes without engineering overhead, while tying results to behavioural signals such as clicks, scroll depth, and navigation paths.

At the time, the market offered two extremes. Google’s tool was free but not intuitive. Enterprise tools like Test and Target were expensive and complex. Wingify positioned itself between the two, with a focus on usability and affordability.

“VWO succeeded because for marketing teams, it cut effort and time by 10x for an existing behaviour versus their status quo,” wrote Chopra in archives of the company’s journey that he has been publishing since 2013.

In one such account, he described launching Wingify on HackerNews to gather feedback. Users pointed out that the product was difficult to understand, that the technical language needed simplification, and that one user took 14 page views and nearly 30 minutes to find the product tour. The team used this feedback to iterate.

“Don’t code in isolation,” wrote Chopra in these archives. “Involve potential users from day one, they know their pains.”

Chopra also documented early revenue milestones.

“I hit that milestone in the first week of launch. The first month's revenue was $4,000, and the second month it shot to $9,000. Third month, [it was] $15,000. By the 18th month, Wingify made a total of $1 million in revenue.”

Selling globally from India came with its share of friction. For over a year, the company listed a London address while operating out of Delhi. This reflected concerns about market perception.

Gupta noted that exposure to entrepreneurship was limited at the time. Most talent opted for higher education or large companies, and access to capital was constrained.

He also pushed back on the idea that building products was harder then, considering access to knowledge and information is far easier today. 

“It’s funny because I can go back to the 1600s, 1800s or during the time when the bulb was invented—access to information was much more difficult,” he says. He recalls reading IBM research papers to learn software development. “It was fun in its own way.”

In a recent podcast episode on The Neon Show, Chopra said he relied on extensive written documentation to align the team. This allowed him to communicate expectations clearly and shift roles without being trapped in meetings.

He also argued against building in markets with no competitors. 

The presence of companies like Google validated demand for A/B testing, meaning he did not need to convince the market of its importance. He only needed to build a better product.

In the early stages, the team spent little on aesthetics. They used a basic $5 template theme for user interface design, drawing on existing inspiration while focusing on usability, onboarding, and solving real pains by building “one feature done well”.

Chopra also advocated for going beyond customer support. He described the role as a “customer happiness engineer”, where the goal was not only to resolve issues but to delight users. It resembled what is now described as a forward-deployed engineer, long before the concept became popular.

From 2010 to 2013, the company received three acquisition offers and was approached by more than 30 venture capital firms. It scaled to handle over 7,000 requests per second and grew to 2,500 paying business customers across more than 70 countries, including Microsoft, AMD, Disney, GE, and MakeMyTrip.

“For the last 15 years, we’ve built Wingify in an economically conscious manner. You can raise a lot of capital that can fund your way to growth, but we were always looking at our EBITDAs,” Gupta says.

Chopra has described the company’s guiding principle in these archives as simple mathematics. The focus was to earn more than they spend and not optimise for valuation.

With capital from Everstone in 2025, Wingify is now entering a new phase.

“When private equity comes into the picture, the stereotype is that they build efficiencies, cut costs, and make changes in the system.”

“There is absolutely not even a single change in our leadership or strategy or product direction,” Gupta adds.