Can BrowserStack be the World’s Software Tester?

Last fortnight, BrowserStack raised $200M, at a valuation of $4Bn, making it India’s most valuable SaaS startup. 

Exploring the Internet

In 2011, Ritesh Arora and Nakul Aggarwal were brainstorming in a Mumbai coffee shop.

The duo ran a business to help businesses on something called the Internet. Ritesh and Nakul’s Internet access put them in less than 10% of the Indian population. 

The internet itself was in its early days.

At the time, an internet user had very limited options for devices or browsers. Chrome and Firefox were in their infancy. The big daddy was Microsoft’s Internet Explorer. 

Bundled with Windows, IE’s market share was a whopping 41.1% globally and almost 60% in India. India was still operating on 3G, the average internet speed in 2010 was just 0.8Mbps.

Ritesh and Nakul, though, wanted to leverage this to expand their consulting business.

To do so, they decided to start a website to scout for projects. One of the to-be co-founders for Browserstack reverse-engineered a WordPress template.

The website was up and running in two days. The next task was to test the website across browsers and devices.

As they went through the motions, they realized it would take at least 4 days to download virtual machines, complete set-up and debug their code.

All of this while working on Internet Explorer with 3G internet. 

In the ‘Aha’ moment that all entrepreneurs wait for, Ritesh and Nakul hit upon a realisation. In all that they worked on, testing was the biggest challenge.

It would turn out to be a bigger challenge than even they thought. 

Not All Heroes Wear NetScapes

Ritesh and Nakul weren’t new to making a global impact with their technological acuity. 

Even as they continued functioning out of a coffee shop, they never thought of “going global” as an explicit to-do bullet on their agenda.

Right from their college days, Ritesh and Nakul used to think and act global. They would read all the blogs, forums, and interact with the global developer community. 

They contributed to the open source community, and were among the first few adopters of Ruby on Rails in India. Their Ruby on Rails blog became one of the most popular Ruby blogs on the internet.

Trailblazers since college, Ritesh and Nakul went through more than 70 research papers on ML/AI. They exhausted all there were at the time to put together a sentiment analysis algorithm in college.

They marched in with the same conviction and perseverance that they showed as third year undergraduates, to found BrowserStack.

Building for builders appeared as a monstrous challenge, but even so, it had a certain charm.

The beauty lay in the fact that the boundaries, literal as well as philosophical, all vanished, as the product went wherever the developers sat.

They were seated worldwide.

It was known that developers wouldn’t settle for the minimum viable product. They were often the first ones to leave harsh feedback on public forums if given the chance. 

But Ritesh and Nakul knew what the litmus test for their product would be.

They had to build a product that they themselves as developers would thoroughly love, and enjoy using. With their strong background in tech-focussed consulting, Nakul and Ritesh were well equipped to get it right. 

This would become an inspiring tale and curious case of what’s now widely called product-let growth.

To start out with this product-led growth, they would begin with what now is the first step in the “SaaS Playbook”

A Market Safari

Ritesh and Nakul did exactly what many now-successful ‘developer-first’ companies like Stripe and Twilio used for product-market fit testing. 

They went where the customers were.

The team found that on Twitter, Stackoverflow and Reddit, more than a 100 people were complaining to Microsoft because of the difficulty in using IE to test their websites.

There was a clear problem in the market and the founding team decided to tackle it.

The first iteration of BrowserStack was a simple tool to give developers the ability to test their website by building a virtual environment on a browser.

As a sign of the times, the browser their virtual environment used was Internet Explorer. From facing problems with IE to becoming resellers of IE, they found their product idea.

Since Indians loved using free products (they still do), the beta version was launched in 2011, free to use. However, the Internet blurs all the boundaries and by being truly developer-first at heart, BrowserStack was always meant to be a global product. 

The plaudits came from far and wide. John Resig, the founder of jquery, tweeted about liking the product/ too to his 200k+ followers. 

The solution was working for people, and BrowserStack caught on in the developer community like wildfire. 

So much so that they grew purely on word-of-mouth, an almost extinct form of growth in the world of influencer marketing and 10x growth hacks. 

Their thought process was simple. They wanted to let their product do the talking. 

Within a month there were 10K beta users. Some of them were asking for a paid version of the product, surprised that they were getting this level of functionality for free.

But it wasn’t a product in Ritesh and Nakul’s eyes yet, simply a tool to solve their own problems and help the community. 

But they were clear that if they wanted the business to grow, they had to solve a problem no one else was solving. 

The free nature of the product was only step 1 in discovering the depth of the problem.

These aren’t the Androids you’re Looking For

BrowserStack’s free product received unprecedented interest from the community. 

They went on to charge $19 per month as a part of a pay as you go model (also explaining how they’ve been profitable since day 1).

Within half a year, BrowserStack grew to making $20,000 dollars a month, and realised that they were hitting their limits because of the pay as you go model.

They needed to get creative with their pricing / business model to find a way to meet their customers’ needs in the most efficient and collaborative way.

After some brainstorming, Ritesh and Nakul decided to go with a tiered subscription model. They started off with their first product, Live Testing, and offered 5 subscription tiers to their customers.

Proximity to the dev community meant that feedback would reach them ultra fast.

They only had to look into their support tickets to realise all that was wrong with their offerings, or what was confusing their customers.

The feedback was suggesting that their users were confused by the number of options presented, while picking a plan for a team.

Many users would often share the same log-in credentials, nullifying the need for a multi-license subscription. 

Upon the launch of Automate and Screenshot & Responsive, BrowserStack now presented 15 unique pricing plans, leaving their customers even more befuddled than the last time.

Trying to choose the right offering of each product for all the many members of one’s team was a dizzying experience.

BrowserStack quickly rectified this by simplifying their pricing models, now offering 3 plans offering a broad set of features in each of them.

The first was Live, which offered unlimited access to manual testing, Automate, which offered All Live features and Automate features, and a third, customisable plan for enterprises.

The customisable Enterprise plan proved to be their show-stopper, with it playing a pivotal role in the onboarding of the many companies that partnered with BrowserStack as their chosen testing platform, over the years.

BrowserStack was scaling, and nobody could stop them. 

Founders usually would go to VCs at this time, having developed early product market fit. But the founders’ experience with prior failed ventures made them realize the importance of being capital conscious. They decided to bootstrap this one for the near term. 

That near term would last not for one or two years, but for far longer than they’d thought. 

Keeping Tabs on Chrome

Within 4 years of bootstrapping, the company had reached a revenue of $20M, all completely funded by their profits. 

As Browserstack entered 2016, it was a particularly exciting time to be a part of the testing ecosystem.

Change was the only constant in the realm of software development. Digital transformation was occurring across sectors. 

A staggering number of players began requiring early and continuous testing to yield high quality innovative products and improve end user satisfaction.

Previously a back-end function within the product development lifecycle, software testing began taking the centre stage

Testing was no longer done in silos. It was instead becoming a necessity at every stage of product development.

Companies were adopting the agile and CI/CD methodologies for a variety of reasons, among which were accelerated software delivery, the enhanced ability to manage changing priorities, and efficient management of distributed teams; with the latter positioned to play a hand in the future of work envisioned by many.

The share of companies leveraging the agile methodology was expected to go from 40% to beyond 75% in 5 years. The addressable market for software testing, expected to grow at a CAGR of 8% to USD 76 Billion in the same time frame.

It was a massive opportunity. 

Rapidly evolving business models and computing technologies, coupled with glaring redundancies in pre-existing SDLC methods were forcing shifts in testing approaches. There was a constant re-evaluation of the sophistication of testing tools and platforms.

For BrowserStack, the key components contributing to the problem statement were also dynamic in nature. As the company was entirely bootstrapped, they could move quickly and nimbly.

By 2018, BrowserStack was cruising at a $50MM revenue run-rate, still bootstrapped and profitable. 

The founders realized there was a stark need to have a partner who would understand and coach them to think globally.

Due to its growth and profitability, Browserstack would raise a giant $50M Series A. The rocketship was now readying itself to become the global leader for cloud software testing.

Between the time of Browserstack’s founding and Series A, the browser landscape had completely changed. 

Site at the Opera

The erstwhile monopoly Internet Explorer had gone from the most widely used web browser to an oft-unused tray icon on one’s task bar. 

Google Chrome was now a behemoth, Safari a revived old favorite, and a dozen new browsers were coming up real quick.

This was just the web browsers. Apps had exploded on the scene, driven by the rapid adoption of the mobile phone. 

The team, as Ritesh recollects, put together the beta version of their web browser testing offering within months, but building their mobile app testing environment took them almost three years.

Cellular subscriptions had reached almost 8.5 billion in 2018 as compared to 738 million in the year 2000, with a penetration rate of almost 97%.

Mobile apps were burgeoning in number, with levels as high as 2.1M for Android based and 2M for iOS based apps reported in the 3rd quarter 2018.

The industry was estimated to generate close to USD 200 billion in 2 years, which made the value attached to testing surge as well. 

Mobile testing emerged as one of the key areas that would aid companies in improving the end-user experience.

As a direct result, more than 92% of enterprises announced plans to focus on mobile testing, opening up yet another market opportunity in testing

Soon, ML/AI were changing every aspect and every field that technology had ever touched, and software testing was not going to be left behind.

BrowserStack’s Automate was positioned to operate at the bleeding edge, thanks to the founders’ zeal.

Positioned as a B2D company, BrowserStack’s founders initially went with their own intuition, figuring out how much they would pay for such a product.

But now that they had scaled, they needed to identify what made Browserstack different from the hundreds of startups sprouting up.

They didn’t need to look very far. 

We Didn’t Start the Firefox

Startup moats usually come from the founders’ visions

Bezos wanted Amazon to keep the priorities of their customers at the forefront of their innovation and operations. Customer centricity is now what differentiates Amazon from the hundreds of other e-commerce companies.

Browserstack founders’ thinking from day 1 was to solve a very painful, real problem, for a large market, with a product-first strategy. Ideally, the users may be living with the problem on a daily basis, but they would be totally unaware of that.

The growth thesis for BrowserStack early on was wonderfully simple – it would be community driven. 

When the target segment starts using their product, they should feel like it has changed their lives. That’s when people start talking. 

Heavy traction would generate word-of-mouth publicity. Once network effects kicked in, the entire developer community would know about such a product being given out for free.

This did the trick for BrowserStack. They spent a sum total of 0 marketing dollars for the first eight years in their journey. Before long, a Postman-esque success story would be in the making. 

This eventually led to two important pillars on which BrowserStack would lay its foundations. 

A strong symbiotic community doling out feedback necessary to improve the product, and a positive cashflow.

The idea of building a strong community sparked from the observation that the founders found it very difficult to hire for their bootstrapped, profitable, venture in the initial days.

The reason – candidates found it amusing to work for a company that operated out of Starbucks. 

Yet, when they tried out their first office space, it was just the two of them, sitting across each other in pin-drop silence.

That would exactly be the feeling the users of BrowserStack would have, if there was no vibrant community. This idea snow-balled over the next ten years to build up to what now was a differentiating competitive moat for BrowserStack.

The community moat would also be a source of inspiration for products. 

Yahoo Let the Dogs Out?

Browserstack’s evolution over the years was a direct implementation of the demands of the developers.

The first paying product that BrowserStack had launched was when many of the 10K beta users of the product started asking about a paid version. The second and third products too were a result of scanning through the support tickets to identify what the main reasons behind those tickets were. 

All of these products were built keeping in mind the mission of BrowserStack – empowering developers to build amazing experiences. 

With the acquisition of Percy in July 2020, a San-Francisco based startup specialising in automated visual testing for applications, BrowserStack enabled developers to not only test their code for accurate behavior but also for visual correctness.

Visual testing was the automated process of testing and reviewing visual changes on the website. Percy captured screenshots and compared them against the baseline, highlighting visual changes. This enabled the users to detect even the most elusive visual change.

The acquisition was well thought out.

Percy already had a history of working with Google, Canva and Shopify, helping them gain the confidence to build visually stunning applications.

BrowserStack’s market was expanding, and that meant even more investor interest. Among numerous problems that BrowserStack would face, capital was never one of them.

As they said, focus on building a sound profitable business, and the VCs would find you. The company had the happy problem of being chased by investors both from India as well as Silicon Valley.

By July 2021, BrowserStack had tripled its employees to 750+, had 50K+ paying customers across 135+ countries and 4MM+ developers on its platform. 

It also had established presence in seven countries and opened ten new data centers, taking its total count of data centers upto 15.

But the name of the game was growth, and after a pandemic that only expedited adoption of technology. 

Like its approach to the first round, it was time for BrowserStack to look for a partner who had a reputation of blitzscaling profitable Tech Cos globally.

This philosophy couldn’t have been truer when BrowserStack raised a $200MM in Series B funding at a $4B valuation in July 2021. The lead investor in the round brought with them the know-how to create giants in the technology business

The company was now India’s largest SaaS startup. 

Its revenue run rate was estimated to be $150M+, almost 8x from their first round. The founders’ unorthodox approach to raising money came from this unorthodox approach to building business. 

Due to the business’ strength, the founders gave up only 10% of the company in the first round, and 5% in the second round. This implied that the founders owned more than 80% of a monster $4Bn business – a founder wealth of $3Bn+.

Most Indian unicorn founders owned low single digits, making this an incredible feat. 

BrowserStack would now have a warchest and strategic guidance to do acquisitions and expand its product offerings.

Ask Your Developer

If software is eating the world, then BrowserStack wants to keep its appetite going.

Even though BrowserStack’s claim to fame was as a platform for cross-browser testing, evolving to include mobile application testing, BrowserStack’s ultimate vision is to become the testing infrastructure for the internet. 

The complexity of testing in the DevOps cycle will continue growing exponentially, as more ways open up allowing consumers to interact with the internet.

New classes of devices, such as Smart TVs and wearables, mobile devices and new OS versions will require a testing and iteration platform. 

Testing processes as well would get more nuanced because an amazing UX now includes testing for accessibility, visual correctness, as well as security.

BrowserStack aims to decode these layers of complexity to provide a platform that takes care of all the fragmented pieces of testing developers dread. 

A large number of companies are looking at cloud solutions to replace their on-premise infrastructure. They are going through rapid digital adoption and are moving key systems, tools and processes to the cloud. 

All of this will act as tailwinds for the growth in the use of BrowserStack’s products.

Looking forward, people will spend more of their time digitally. Companies would want to tap that attention by cutting down the time spent in production and development.

Ship products faster to satiate the ever-growing demand for new products. 

The bar on speed and quality would be raised higher, which would require more sophisticated tools for flawless testing. 

Testing itself would need to accommodate for the increasing number of edge cases involving access, security and usability, across a multitude of devices expanding beyond the traditional desktop browser. 

Development teams would have to manage and mitigate all these challenges.

BrowserStack has patiently evolved to be that overnight success lessening the “load” of all the developers, giving them instant access to the widest range of device configurations to test their applications. 

BrowserStack is undoubtedly the boon that developers across the world did not think they needed, but now can’t live without.

By Bhoomika, Mehak, Omkar, Raj, Shelley, Shiraz and Aviral

guest
0 Comments
Inline Feedbacks
View all comments