Do your customers want what your selling? It's a question to which any business will, sooner or later, find the answer. We sat down with Bartosz Baziński - founder of successful social listening and automation tech startup SentiOne and laureate of the IT Olympics - to explore how, in his experience, new startup founders can make sure they achieve produce market fit.
First, thanks for meeting with us Bartosz. For those among our readers who may not have come across your work before, can you please give them a quick introduction to yourself and your work at SentiOne?
Thanks for having me! I’m the COO and co-founder of SentiOne - AI startup that helps big B2C clients automate their customer service. Our solutions range from online listening and fully-integrated customer contact centres to new-generation conversational bots. We have been operating for almost 10 years, in 26 markets and our main clients are marketing agencies, big FMCG brands, telco and retail banks.
Great! Thank you, I’m sure people will find that helpful! As many will know, SentiOne is huge in the automation space, focusing on social listening and AI customer service. Could you explain to our readers why SentiOne is unique in this space, particularly on customer service automation, which is a type of service many will have seen before?
You are right - businesses have been trying to automate some processes of their customer services for many years. The key to success is technology. We have decided to focus on conversational AI - currently the most advanced form of chatbots - and created a dedicated AI Team of deep learning engineers and built our proprietary intent recognition algorithm. As a result, our chatbot technology scores 98% accuracy in intent recognition - that is the highest result globally. So I can proudly say that our chatbots are the most accurate in the world.
What does that mean? SentiOne chatbots are now able to easily understand customer motivations and intent – almost as well as humans! That is why our bots can be easily placed alongside human customer service agents and successfully take over the most repetitive and mundane customer requests.
As we’ve discussed over at Develocraft.com in quite a lot of detail, product-market fit is critical to the success or failure of any digital product. What steps did you take to make sure your products - particularly the customer service automation, or ‘chat bot’, had strong market fit?
I hope this answer will satisfy you and your readers - we simply talked to our clients. You’ve got to understand your clients’ needs to build a product that will meet those needs. For SentiOne it was quite easy, actually, as our most loyal clients contributed to product development along the way. We have been offering a multichannel customer service platform - one tool where customer service teams could reply to messages coming from various channels (Messenger, Twitter, online form, etc.). One day our clients hinted to us ‘Hey, it is all great, the platform works fine, but we have received many more customer messages than we anticipated. It would be good if we could automate it somehow’. So we started developing small automation features: templated answers suggested by AI algorithms, automated ticket routing, and finally virtual assistants for replying to the simplest messages. Your customer, their needs and problems are the key to your success! After all, they are the ones willing to pay for your solution. After a relatively long time of operating (for a startup!) our product market fit is a result of product development based on clients’ feedback.
Interesting! Did you encounter any unexpected difficulties when it came to researching your product-market fit? Were there any challenges in finding potential users to test with or interview, for example?
That is a very good question. As we have been on the market for a while, we did not have problems finding clients to show our MVP or conduct an interview. However, once we started analysing all the answers from their interviews we realised most of them contradicted each other! Bank A would be looking for solution X, but Bank B, a similar company, would choose a completely different solution. We quickly realised that the market of customer service automation is so complex and challenging that if you were to listen to every single feature request you would end up building custom solutions for every single client. That’s not scalable at all. So we had to ignore some of the answers, accept the fact that we couldn’t build a chatbot suited for every process (yet!)! To satisfy our clients and their users, the solution was to narrow down our market - to be able to fit.
Did you learn anything while working to make sure the chatbot was meeting a real appetite in the market that would inform the development of your next product?
At this stage I feel like our chatbot offering requires further development so I’m not thinking about future products yet. It was only after working on chatbots and completing the first pilots and enterprise implementations that we realised how big this market is. There are so many industries we want to enter, so many new processes to automate!
You’re quite a technical guy - if you had to give some advice to a non-technical founder of a startup what would it be?
Get yourself a technical co-founder! There is probably no better way if you want to develop a tech startup. Combining the business and tech skill sets gives you the highest chance of success. Someone technical by your side will ensure your product is unique, scalable, and will be able to calculate what [digital product] is profitable and what is an absolute nightmare to develop.
Have you been surprised with the success of SentiOne, and what do you think underpins that success?
Success is relative - when you see how many good and innovative startups have failed - yes we may feel successful. But then, we look at our well-funded competition - and we feel how much work is still ahead of us. What definitely helps is the uneven number of co-founders - you always need to get a majority to agree! That leads to very interesting discussions and arguments. Another thing - get someone onboard who is good at recruitment very early on. Recruiting talented and motivated people that are smarter than you is definitely a huge catalyst.
It’s a difficult time for many startups. The massive economic contraction caused by COVID-19 has caused real difficulties for many, such as decreased revenue and difficult decisions such as having to lay-off staff. What advice would you give to startups trying to get ahead in this environment?
These are tough decisions and everyone should make their own call. Look at your financial spreadsheet and decide what you can cut. Our aim was to keep 100% of our staff at any cost. So very early on we had to cut down all the nice benefits like gym memberships, language courses, and so on. We even tried to renegotiate our office lease.
As this is an interview about market fit, do you anticipate that the post-COVID world will bring significant long term shifts in the market that startups and companies with existing products will have to consider carefully to make sure they remain aligned with the needs of users?
I think this crisis showed us very early that if you don’t solve a significant problem for your client it will be difficult for you to stay afloat. When developing a service or product you should always provide one of these three benefits; you save time, you save them money, or you save them a headache. Be a ‘must have’, not a ‘nice to have’. I think the post-COVID world will validate the offerings adding real value and show which were just technological gimmicks.
And finally, if you were starting a new tech company today, what industry would you focus on? Would you go with an area of economy where startups have previously enjoyed great success in disrupting the status quo, like transport or would you target somewhere new? And would you go for B2C or B2B?
The question is, how much money would I have at the start? B2C requires a lot of funding and in B2B it’s a bit easier to find your niche. In all fairness, I feel like AI is still only getting started and I find finding new applications for it fascinating - especially in new industries. Medicine is a very intriguing field but you need a lot of funding. There is also the construction industry - how come it hasn't been disrupted in decades? We still build houses the same way. I would love to invent a system to speed this up.
A software engineer and entrepreneur, Bartosz is passionate about applying cutting-edge technology to the world of business. Co-founder and COO of a conversational AI startup - SentiOne. With 8M Euros in funding, 80+ top talents onboard and customers like Procter & Gamble, Beiersdorf, Huawei, Microsoft, PwC and others, SentiOne leads the market in AI-based customer management systems.
Hello! I'm the head of content at Develocraft. I'm also a startup guy (no joke)! I've worked with a lot of them and learned so much. I'm here to help you by sharing that knowledge. I'm always open to collaborations. Find me on LinkedIn.