First-Mover Risks: How This Founder Bootstrapped On-Device Speech Recognition

You’ve probably read about the first-mover advantage: the benefits of being first to bring a product or service to market. In business, entrepreneurs often see it as an achievement and how to secure an insurmountable head start.1 But what about first-mover disadvantage

For example, a first mover must invest heavily in educating and persuading consumers to try something new.2 You need to explain why someone needs your product, what problem it solves, and how they should use it. Meanwhile, your competitors-in-waiting watch your every move, learning from your mistakes and copying your successes. 

Ognjen Todic (who goes by Ogi) founded Keen Research to bring on-device speech recognition to the mainstream market, which no one had done yet. Now with a team of four, the company develops software tools for speech recognition on mobile devices and custom hardware platforms, powering apps in education, entertainment, gaming, and medicine. 

As a bootstrapped founder, Ogi had to build his product, acquire customers, and start generating revenue without a financial runway. How did he do it as the first to market?

When a Friend’s Joke Inspires Your Next Startup

Ogi, originally from Serbia, came to the United States in the nineties. He got a job at a startup while studying for his master’s degree in electrical engineering. In the early days of his career, he worked in software development and speech recognition, fulfilling R&D (research and development) roles at both startups and large companies. 

In 2004, he left his full-time job to start Keen Research. Back then, Keen Research helped other businesses develop software products. It was also his first crack at sales, marketing, operations, and so on. Ogi then spun off a music search company called Tunezee that synchronized song lyrics. He ended up selling that startup to Shazam to power its synchronized lyrics feature. 

In 2016, Ogi returned to Keen Research with a different focus: on-device speech recognition. He recounts, “It all began six years ago when a friend jokingly complained to me about the repetitive chore of helping his daughter practice her multiplication exercises. ‘Why isn’t there an app for that?’ he asked. I began toying with prototypes for a voice-enabled solution that quizzed a learner on multiplication tasks using a mobile phone or tablet.” 

The challenge was that the speech recognition in the app would need to run locally on the device to be responsive. After a couple of months of intense work, Ogi launched the multiplication app, Multiply With Me, in the App Store. 

While a fun project, the founder thought this small segment of the educational technology (edtech) industry wouldn’t turn into a sustainable business. That’s when the realization hit him: Developers likely faced the same challenge, yet there was no good solution on the market for on-device speech recognition. So, Ogi decided to build his own.

Surviving Your First Year 

It took Ogi a year to develop the first version of his product: KeenASR SDK, a software development kit for on-device speech recognition that would allow other app developers to voice-enable their mobile apps. But the founder knew there was a risk in investing all his time and money into developing an untested product.

He says, “I wasn’t worried about the technical side of the business, I knew I would be able to create it. The real question was: Do people want and need this? Are they willing to pay for it? One way to get the answer is by asking people. But the only real validation is when somebody writes you a check. Until that happens, you don’t know whether your product and business are viable.”  

Ogi made sure he had a Plan B. “I knew I could always find a soft landing in the industry if my startup didn’t work out. I could work for a big company and help them develop on-device speech recognition. Early on, I pitched KeenASR SDK to some large companies interested in voice-enabling their apps and they asked me to come work for them. I wasn’t looking for a job or a big salary, because I wanted to build this and make a business out of it – but it was good to know that I had that safety net,” he says. 

That first year was rough according to Ogi. He already had a family with two kids and, alongside his efforts on Keen Research, didn’t have another job to supplement his income. The founder did a couple of consulting gigs but mainly had to dip into his savings to keep the business going. 

Ogi reminded himself of why the world needed on-device speech recognition: It allowed voice applications to run offline in weaker networks, protected privacy regulations (like the COPPA and GDPR) by not sending personal data from the device, and offered app developers more affordable and scalable apps that don’t require backend speech processing. 

With the early beta releases, enough people were interested in the product to keep Ogi developing it. Keen Research morphed from a software development shop into an on-device automatic speech recognition SDK provider. 

The Challenges of Being a First-Mover 

As a solo founder, Ogi wore all the possible hats for Keen Research, including R&D, business development, sales, marketing, and customer support. It’s not that he didn’t consider partnering with someone, he just didn’t find the right cofounder. 

Ogi says, “In my close circle, there was nobody who was the right fit to launch this specific company with. But instead of waiting until I found a cofounder, I turned my attention to building the product. I still put feelers out and told myself that if somebody came along, that would be great, but I didn’t want that to be a barrier to developing the product and growing the company.”

The founder applied that same principle to funding. While he spoke to a couple of angel investors early on, Ogi didn’t want fundraising to stand in the way of his product. Instead of focusing on raising money – which can be a full-time job – he built the product first and would consider funding later if his business succeeded.  

It took Keen Research about a year and a half to acquire its first customer – and it took another couple of months to secure the second one. Voice interfaces and speech recognition weren’t mainstream yet. While big companies now have voice search services (think Amazon’s Alexa and Apple’s Siri), smaller companies building voice-powered products were, and still are to a large extent, in their infancy.

This meant that acquiring new customers was difficult. It required helping prospects understand the capabilities of the technology and helping with integration and product design. After the first few customers, certain industry verticals started to show potential. So, the founder focused his outbound sales outreach and marketing efforts on those verticals. However, it took about four years before Keen Research started doing any serious marketing.

Along the way, Ogi realized that not every inbound email needs to be treated with the same effort. In the early days, he would (rightfully) get excited about any sort of inbound inquiry. He would spend a lot of time on the phone with a prospect, trying to understand their needs and whether KeenASR SDK was the right solution for them. But it became evident that certain leads were unlikely to generate any revenue. Being bootstrapped, Ogi was forced to focus only on opportunities that could bring in money. So, he developed ways to filter out and prioritize inbound leads.

He says, “You also want to be careful about getting stuck in something that seems like a huge opportunity. It’s great if you secure a contract with a huge company, but watch out for the red flags. If they require and expect a lot of customization and upfront work for free, it might not be worth the effort.”

Keen Research rarely commits to free work, aside from helping prospects with the initial integration and evaluation. And even then, the expectation is that customers do this (in the context of their product) with Keen Research providing assistance where needed. Ogi learned to ensure that a company is serious about becoming a customer, putting some skin in the game, and not just taking advantage of a small startup.

As consumer demand grew, Keen Research started getting customers in the edtech and kidtech space. The founder developed acoustic models optimized for children’s voices and features focused on automated oral reading instruction and assessment to cater to that industry. When Covid hit, Keen Research garnered even more interest from educational companies. Digital education grew rapidly with children studying from home in virtual classrooms. 

For example, PBS Kids uses KeenASR SDK for on-device speech recognition to voice-enable a robot in The Cat in The Hat Invents game. Nickelodeon uses the SDK to create interactive videos in its Noggin app. Kids can talk to a Paw Patrol dog and interact with the narrative to learn social and emotional skills.

But Keen Research doesn’t only focus on edtech, its tools can also be used in a variety of industries and use cases. For example, enterprises can use it to increase the productivity, safety, and efficiency of their employees. In the entertainment industry, companies can use it to create fun and engaging experiences within games and educational apps. Developers can use it to extend virtual and augmented reality apps with voice interaction. 

Ogi believes there’s a big future ahead for on-device speech recognition tools. But that doesn’t mean the path going forward will be easy. The founder says, “The entrepreneur’s journey is like a rollercoaster ride. I’ve learned not to get too excited when things are going great, nor to get depressed when things aren’t. You have to become good at modulating the ups and downs if you want to survive entrepreneurship.”


1The Half-Truth of First-Mover Advantage

2 First-Mover Disadvantage

Want to share your bootstrapping story with the world? Enter your details in this form and we’ll be in touch.

Andrew Gazdecki
Andrew Gazdecki
Andrew is an award-winning serial entrepreneur with three exits. He’s the founder and CEO of MicroAcquire, the world’s most founder-friendly startup marketplace, and its rebellious child, Bootstrappers, which gives voice to the entrepreneurial underdog. When not building businesses, he writes for Forbes, Entrepreneur, and now, Bootstrappers.

Stay in the Loop

Get our weekly newsletter wrap-up with all the latest news in the Bootstrappers community.

Latest stories

You might also like...