Q&A with Bart Bazinski, COO & Co-Founder of SentiOne

3609 0

We’ve all been there — clicking swiftly to close the chatbot popup or getting an auto-generated email barely related to our claim. There’s nothing more infuriating or defeating. But what if there was another way? What if bots actually got the job done? And better than humans?

“We’ve developed technology that provides state-of-the-art customer service automation based on social listening and data analytics,” says Bart Bazinski, COO and co-founder of the AI-powered startup SentiOne.

“Our natural language understanding engine, combined with our simple interface, allows companies to build superior conversational chatbots without IT team involvement,” he adds.

SentiOne’s multi-channel solution is now relied on by over 350 brands in more than thirty markets around the globe, including Procter&Gamble, McDonald’s, Nike, Unilever, Beiersdorf, Tesco, Starcom, Mindshare, Mediacom and Saatchi&Saatchi, among others. Last year, the company pushed past $2.1M in revenue and is currently on target to hit at least $3.5M, in spite of COVID-19.

We recently Zoomed with Bart to hear how he scaled the company to a team of 80, how he approaches sales and customer success, what hiring for a culture fit should look like, and his thoughts on the future of Polish software houses.

Bart Bazinski with co-founders Kamil Bargiel (c) and Michał Brzezicki (l) in 2017

How did you make your way into the world of startups?

When I was doing my engineering degree, I figured that the best way to learn is to start my own company. That turned out very well, and the first company was a software house. After a while, I realized that it doesn’t make much sense for us to continue as a software house because developing software for others doesn’t scale up. We then decided to launch our first product with SentiOne, and that’s worked well enough for us to grow the company to 80 people now.

What’s the difference between working for a startup and a software house?

It’s a totally different world. In a software house, the most important thing is to deliver on time with good quality. When you’re building your own product, the quality has to be excellent because that is the only way you can win in today’s software market.

How would you describe the growth of SentiOne?

In 2011, we raised $200,000 of pre-seed capital to get the company off the ground, and then in 2016 we closed a $3.5M seed round. We started with a social listening platform — basically we were analyzing what’s happening on the internet for our clients. This product gave us $1M in ARR [annual recurring revenue], and we managed to retain our large enterprise clients for the long term.

But then we realized that the overall churn was way too big to scale up. We were losing clients as fast as we were gaining new ones. Typically you would think that this company is not doing that great, and that we should just try to get as much money back to our investors as we can.

Fortunately, we had – and still have – a great technical team full of very talented people. Back then, we also already had a lot of experience with natural language processing. We reached out to our clients to see what solutions they were looking for, and it turned out they needed customer service automation right away.

We managed to find a problem in the market that could be solved by our AI engineers. Our clients were so excited by our customer service powered by AI that they wanted to buy it before we even built an MVP [minimum viable product]. So we had to build it and commercialize it really fast.

How did you build such good relationships with your clients?

You need to have customer service and sales teams that will stay with your company long term and build those relationships with the clients.

Most startups build scalable products and put a lot of effort into creating self-service for the clients, but I see a lot of added value from hours of presentations and meetings with them. It helps us build relationships and learn what technology they might need.

— Bart Bazinski

How do you prepare an enterprise client for innovative solutions?

There is a lot of knowledge transferred whenever we prepare enterprise clients for the implementation of our solutions. We invest a lot of time and resources in delivering training and technical workshops to our clients in order to convince them to use a specific technology. In fact, not all of them are ready for solutions like ours. That’s why we work together, help them learn, and wait until they’re ready to undertake a bigger project with us.

Usually, when the right time comes, we already have the right relationship with them. Our business is scalable in that sense. We prepare materials, train ourselves, and get more effective with every new client by building those relationships.

It also very much depends on the management — some enterprises look into the future, try new solutions, and develop a culture that enables people to experiment. This is usually the type of client we want to work with, where management is actually building a culture of innovation and not just saying that they do.

What happens after the sale?

There’s a lot of communication and very close feedback loops. Often there’s a dedicated team working on the customer side of the project. A few months after the sale, there is a lot of communication and improvement. Afterwards, if everything works correctly, there’s less work. We’re very happy with our current enterprise clients. They’ve signed on for the long term, with contracts ranging from three to five years.

Bart Bazinski (l) and Tomasz Tomaszewski at a conference to discuss future of chatbots and AI in customer service in financial industry (2019)

How often do you discuss client cases on a management level?

We all work very close to the business side of the company, and are often involved in the sales process. This way, we know what the client wants. But then we also collect feedback in a more structured way.

How did Covid-19 affect your company?

Unfortunately, a few large customers dropped out because their business was hurt really badly. Despite that, our sales team still managed to meet their targets for the past three months. We’re working in the digital space and there’s even more interest in our solutions right now. One of our banking clients actually had a 50% increase in volumes in their remote customer service channels.

In general, it’s a systemic change because companies are becoming aware that their businesses are very reliant on people showing up at the office. When you have a large customer service team at your company, you start to think: how are we going to operate without these people in the office? Our solution is the best answer to this problem. 

— Bart Bazinski

When it comes to our company, our team was ready to work remotely even before the pandemic. We already had a lot of project management solutions in place. But then, we also got feedback from our teams that, although their efficiency is great, they’re getting slightly frustrated as it’s hard to separate work from home life. We’ve been organizing a lot of online catchups, virtual beers and other integration games, which turned out to be cool.

On top of that, against the pandemic current we managed to hire five people. As the demand for automation solutions grew, we had to expand our team of AI engineers and chatbot designers.There is a lot of talent available now, and the pandemic makes it easier for us to hire. And, generally people want to work at SentiOne because they know that AI, digital transformation and automation are the future.

What is Sentione’s culture and what type of people do you look to hire?

We look for people who would like to build meaningful connections with other employees and will feel a sense of responsibility for how this company will grow and develop. During the hiring process, we include many of their future teammates. We want to work with people who want to learn a lot from one another, and we want them to feel that they can develop their own careers at SentiOne.

Where are you in terms of scaling?

Last year, we made $2.1M in revenue, and this year we’ll hit at least $3.5M. Nevertheless, we still need to restructure our sales process and shorten our sales cycle. It typically takes us several months to close a deal now. We want to scale up our sales and marketing efforts, and the trick is we really need to know whether we are spending our sales budget in the right way. The crucial point is how to scale efficiently.

In enterprise sales, it’s very easy to burn a lot of money on events, marketing, proof of concept — and in the end, people will still not buy your solution or, worse, you will lose the client at some point in the procurement process. We’re working on how to segment our potential customers to identify the right ones. Then we’ll be able to predict much more efficiently how to allocate our resources.

What advice would you give to the companies that want to scale, especially those based in Poland and its neighboring countries?

Developing markets lack the experience on how to do business on a global scale. I think there is no other solution to this than just to build marketing teams outside of Poland. We have a lot of talent, great developers, and many product managers, but it’s hard to build efficient sales and marketing departments. The business development and marketing teams should be based in London, New York, or Berlin.

Bart Bazinski at the 2019 AI & Big Data Congress

What is your take on Polish software houses becoming startups by launching their own products?

There are success stories like Booksy and a few not-so-successful projects on the market. I think the big challenge awaits in the next 10–15 years when Polish software developers won’t be cheap anymore, and a lot of these software houses will have huge problems, such as how to stand out in a global market? Now they’re able to deliver high-quality, custom-made software development at a discount. In Silicon Valley, it’s very hard to find any software house, right?

As the employment costs in Poland grow, a lot of software houses will lose their competitive edge of providing good software developers at a reasonable price. They will need to think about how to transform into product-based companies and, in my opinion knowing many entrepreneurs personally, most of them will fail.

So where do Polish startups need the most help?

I think the biggest problem for Polish startups is to build global-thinking sales and marketing structures. You can find a lot of really good sales and marketing people in the US or the UK.

The worst way forward for the Polish startup ecosystem is for the companies to focus on what they’re doing now, which is building software. We need to try to build more international businesses where sales and marketing is done outside of Poland, and then transfer learned skills and knowledge back into the country.


Lightning Round

1. Last podcast you streamed: Tim Ferriss’s interview with Brad Feld on the art of unplugging
2. Must-have in your WFH space? Silence
3. Dream coffee chat: Melinda & Bill Gates
4. Your morning ritual: Coffee, quick scan of morning news, and a daily 9am call with our biggest chatbot client
5. Best business advice you’ve received: Have a laser-sharp focus*

*“Don’t spread too thin, don’t try to achieve too many things at once, don’t go into too many markets at once, don’t distract your team with too many features, products, etc. As an entrepreneur you may have too many ideas, but you need to stay selective and diligent.”


Learn More