Authors: Bojan Stojkovski and Bogdan Iordache
From an early age, Pavel Dolezal was immersed in the world of entrepreneurship. His uncle, a craftsman during the communist era, sold handmade wooden toys to tourists at Charles Bridge. This inspired Pavel to quickly develop his own sales skills, exploring the world of sales even at the tender age of ten.
In 2014, Pavel became a partner at Keboola, then a fresh startup developing an agile data management and cloud transformation platform. In late 2023, Keboola secured $32 million in a Series A funding round, marking the largest such investment in Central Europe that year. Pavel’s hands-on, founder-led sales efforts have been crucial in driving Keboola’s growth and global reach.
Before Keboola, Pavel was an accomplished entrepreneur – he founded Atlas Portal, the largest portal network in CEE, and Ataxo, one of the EU’s largest digital agencies in the performing media space, which serviced 15,000 customers with nearly 200 customer support agents.
With Atlas and Ataxo, Pavel rode the wave of digitalisation in CEE, supporting companies to integrate digital tools, channels and practices in their work. After selling Ataxo in 2009, and investing as an angel in a few companies, Pavel was ready for the next step.
Not Always Good To Be Ahead
“Although everything we did back then eventually worked out, it took years for the results to show.”
Pavel’s next project, Keboola, faced the challenge of launching years ahead of the moment the market was ready for them.
Keboola’s vision was to integrate multiple sources of unmatched data that companies store in separate applications, and deploy analysis and decision-making tools on top of it. This would be most useful in a context where a company would have multiple data sources, and multiple client-facing employees – such as a bank, for example.
“My co-founders had a consulting business, which led them to identify the problem we aimed to solve. Meanwhile, I brought my expertise as the original founder and CEO of previous businesses. Together, we organically stumbled upon the concept for Keboola. It wasn’t the result of isolated brainstorming in a glass-walled office, but rather a collaborative effort grounded in real-world challenges,” Pavel notes.
As traditional business intelligence (BI) tools proved insufficient; the team sought to build a robust data platform capable of handling complex analytics at a reasonable cost. Petr (one of the co-founders of Keboola), recommended for his swift dashboard delivery using an internal tool, brought his expertise in system administration and early AWS adoption, setting the startup on a path toward scalable data infrastructure, Pavel recalls.
“This was a pivotal architectural decision. The core idea was to anticipate numerous use cases and a growing user base, especially among non-technical users, with the expectation that demand for such solutions would increase across a wide range of clients,” he points out.
Moving from data to insights, and from insights to actions and automation, posed a unique challenge: it couldn’t solely be a task for tech experts, as tech teams often lack deep business understanding.
“To truly run the entire process pipeline and automate it fully, thereby activating the data, we anticipated a shift around 2017-2018. We estimated that people would start solving these challenges then, but it ultimately took three more years and the impact of COVID-19 to drive this change,” Pavel explains.
Going the Bootstrapped Route
“Value is added by gathering, cleaning, connecting, and transforming the data, then presenting it to business users.”
Keboola was bootstrapped for most of its existence—from 2014 to 2022. During this period, they focused on identifying valuable use-cases and building a complete product rather than just a step in a data pipeline. However, investors thought they lacked focus.
“They typically focused on connector technologies, whereas we wanted to develop the platform further ourselves. As a result, we chose to continue building the platform with our small team, enabling us to achieve more without external funding,” Pavel adds.
One of the early customers was a company with 400 salespeople, many of whom were using different systems due to various company acquisitions. As they sold Yellow Pages services, issues such as double-selling to customers or overlooking service renewals emerged.
To address these challenges, they aimed to consolidate data from production servers and software-generated files into a unified dashboard for better oversight and analysis.
Community Building and Hackathons To Discover Use-Cases and Leads
“We did large hackathons to experiment with new data and technologies, allowing us to play with ideas and test our hypotheses.”
Before any direct sales, Keboola started organising hackathons to get product feedback, discover use-cases, and engage with potential customers or partners.
“We organized three-day hackathons with 500 participants, inviting experts from IBM, Microsoft, and other new companies. The goal was for people to experiment with tools, data, collaborate, and learn together. It was a significant event for us, marking a major opportunity,” Pavel says.
Hackathons were essentially a way to discover and accelerate the learning process. As he explains, it mirrored an idea from South Korea where practising Zen in a group helps individuals help each other, similar to the parable of peeling potatoes together in a bowl, where they peel off one another effortlessly.
Relying on feedback from one customer alone would make their progress slow, but a network with multiple nodes would increase connectivity and accelerate iteration.
In their case, they focused on hackathons for two main reasons: to create a platform for professionals in data science, analytics, and business intelligence to learn and share, and to address talent shortages through initiatives like #DataGirls.
Pavel describes that organizing a single hackathon took them up to six months.
“To host a great data hackathon, it can take up to six months. You need to find the right datasets, get the right technology, and make sure everything is well-prepared for a smooth event,” he explains, adding that the period that followed saw them hosting such events with hundreds of participants.
Soon after, they replicated this model in the UK. “Our initial thinking was,” Pavel recalls, “we couldn’t afford to hire ten salespeople in the UK in 2017. So, how could we do things differently? Why not replicate our community-driven approach? People in the data field are eager to learn; everyone wants to be a data scientist. Why not cater to that?”
They began by focusing on what the community wanted: compelling content, hands-on learning with new tools, networking opportunities, and job prospects.
“We delved into understanding their companies and the daily challenges they faced, naturally drawing them to us through mutual interest,” he recalls.
The result was a community of 4,000 data enthusiasts in London and nearly a year of monthly events. Their next plan was to expand into the US.
However, leaving London for the US turned out to be a mistake for Keboola. The community and lead generation process in London were just starting. And as they learned the hard way, it would take Keboola years to restart their community work there.
The Importance of Design Partners
“In each iteration, there was a significant customer who would become our new design partner.”
As a bootstrapped startup, discovering the right partners was also a crucial step. With limited financial resources, the company couldn’t rely on extensive funding to drive development. Instead, the founders drew on their diverse backgrounds and experiences.
The community development brought in a surge in inbound interest. A dedicated team member handled these inbound leads, while Pavel focused on enterprise sales.
“As a founder, I took a hands-on approach, personally targeting ten key companies. I worked on building pipelines and leveraged my network to make meaningful connections. This focused strategy spanned two years and resulted in securing several of these companies as clients,” Pavel points out.
Guided by their hypotheses and market mapping, Keboola targeted potential customers in the key sectors where their solution could have a significant impact.
“We needed to sustain the company because we were bootstrapped. We made some sales, and then we recognized our potential in retail banks, so I actively pursued opportunities in that sector,” Pavel explains.
Eventually, one of those design partners also became their Series A investor.
“We were fortunate to partner with Tomas Cupr,” Pavel explains. “He previously utilized Keboola extensively at his first company, Slevomat, which was the largest Groupon clone at the time. Tomas relied on Keboola for all business decisions and process automations. Together, we co-developed solutions and recognized the uniqueness of our partnership. Many years later, he became our Series A investor and joined our board.”
During the time they were developing the platform, Keboola also partnered with big data companies GoodData and later Looker, handling challenging customer cases. These collaborations taught them to avoid focusing only on easy tasks that could be easily commoditized.
Around 2020, Keboola had its first major enterprise implementation at Erste Bank. The bank saw the potential in Keboola’s offerings but also identified areas needing improvement. Despite Keboola not being fully market-ready, Erste Bank believed in the value Keboola could provide.
“While we believed we were market-ready, their feedback made us reconsider. And we believed in the significant value we offered and wanted their input to shape the solution. This experience taught us the importance of having supportive design partners willing to collaborate and provide guidance. Such partnerships are invaluable for startups seeking to refine their offerings, something which is also highly appreciated by enterprises,” Pavel reflects.
Enterprise Sales Expansion Through Top-Down Sales and Bottom-up Adoption
“We found the vertical where we believed we could deliver significant value.”
Before the pandemic, Keboola decided to target the US market, seeing it as the largest opportunity. Pavel relocated to Chicago in 2018 and 2019, spending 60% of his time there, and beginning the process of assembling a sales team to establish our presence and operations in the region.
“When I made the decision to move to the US, I had to bring along our best resources, and the network we had built in the UK, just wasn’t strong enough to withhold this pressure. In Chicago, recognizing the much larger market, we initially abandoned the community approach, believing it wouldn’t scale. In hindsight, this was a mistake. We decided to hire six junior salespeople for telesales, as we couldn’t afford a manager. However, as soon as we started, we realized this was a misstep,” Pavel says.
Junior sales people delivered junior results and, when COVID-19 hit, Keboola had to restructure its operations as they couldn’t sustain the entire team during the pandemic. What seemed to be working for Keboola at that time though and hadn’t worked out in Chicago – was their focused enterprise sales in Europe.
They decided to hire their first senior enterprise salesperson, who is still among the top three salespeople in the company.
“We hired him because of his expertise in financial services, a vertical where we saw potential to deliver significant value. With this focus in mind, we decided to expand our presence in the US by introducing a pay-as-you-go model,” Pavel recalls.
Before the Keboola Freemium, starting with Keboola required contacting someone to grant access. The new product took six months of preparation, but when finally launched it gave them the ability to expand their user base.
“The key to entering the modern enterprise market is to have both approaches: a top-down strategy with dedicated salespeople, and a bottom-up approach with a freemium version and community engagement,” Pavel asserts.
From Founder-Led Sales to a Predictable Sales Function
“Figuring out the sales process has been one of the most complex things I have done.”
What he also learned during the process is that one of the most important challenges for every founder doing sales is establishing a repeatable sales process.
“Navigating the transition from founder-led sales to a repeatable sales cycle in today’s landscape is a challenge. This is my fourth sales department I’ve built over my career, and it’s probably the most challenging, and perhaps I’ve relied on founder-led sales for a long time,” he says, adding that now, he is facing another particular challenge.
“When I meet a customer, I see many ways we can help them – which is great for upselling but tough for making the first sale,” Pavel adds.
Thus, he also learned that working with the salespeople is fundamentally different from working with engineers.
“Our original onboarding process was lengthy, involving a checklist and 25 different files to learn. As it turns out, this approach suited technical roles better. You provide them with a repository, show them where to start and where to go next, and they progress independently, with a test at the end to gauge understanding,” Pavel explains.
Salespeople, on the other hand, require a different approach; they can feel overwhelmed and prefer concise guidance.
“They want to know their Ideal Customer Profile (ICP), how to build their prospect list, and what value propositions to offer, preferably in a succinct half-page format. This doesn’t mean they don’t need the detailed information eventually, but they need to start quickly and achieve early success. Once they experience success, more information can be introduced gradually,” he says.
Art or Craft?
“Sales is both human interaction and math – where activities should lead to measurable conversion rates through defined steps.”
For Pavel, sales is fundamentally a fusion of art and science. It involves working with people, which requires trust and understanding. While there’s a mathematical aspect to it – where activities ideally lead to measurable conversion rate – it’s not as simple as expecting people to buy from you immediately.
“Customers need to trust that you understand their problems, even if they may not openly express their concerns at first. Building this trust involves designing a process that is measurable and allows for comparison, akin to basic arithmetic where two plus two equals four,” he explains.
However, there is also the element of craft, as sales involves working with people and knowing when to push and when to step back. It’s a delicate balance because if you push too hard, you risk overwhelming them, yet if you don’t push enough, they may not take action, despite knowing it’s beneficial.
“Even experienced salespeople have moments of hesitation, fearing rejection. Knowing when to stop pushing is crucial to avoid pushing them to a breaking point where their anxieties escalate and affect their decision-making. People in sales are resilient but also fragile, so to know when to push, and when to be more hands-off – that’s a craft that not everyone can master.” Pavel concludes.