Chapter 15: Raising the Ambition Level

Competing with giants isn’t for the faint of heart. Dive into Signavio’s entry into the Process Mining arena and the lessons learned in disrupting the status quo.

www.linkedin.com/in/gerodecker/

Transcript

“Gero, the journey so far has been extraordinary. How about finding someone who wants to acquire Signavio?” Matthias Allgaier asked me.

I wasn’t interested. Signavio was such a great company, and I was enjoying the process of taking it further. Why sell?

Matthias had a point though. We operated in a market that was not growing all that much. Process modeling was a niche and customers only paid a limited amount of money for process modeling software. Gartner called it “Enterprise Business Process Analysis” (EBPA) and Software AG was still bigger than we were. IBM was in the mix, too. We were the only young and cool company in our category. 

Software AG struggled to bring their products to the cloud. Even worse, their profitability mandate kept them from investing enough and pursuing the market opportunity with speed. For IBM, this was just a small side business, and they probably didn’t even notice that they had a product in this category. All the other competitors were small, stagnant, and unremarkable.

LeanIX was a company that reminded me of us. I met Andre Christ shortly after he founded it in 2012, and we quickly hit it off. His company shared our principles - cloud-native, collaboration-centric, easy to use, designed for widespread adoption - and, like us, they operated in an equally unsexy market. In their case, it was called Enterprise Architecture (EA).

While the EBPA and EA markets were similar in terms of size and growth, there was one important difference: Gartner continued to publish a Magic Quadrant for EA, but they had stopped doing so for EBPA years earlier. That wasn’t a great sign.

Of course, we could continue to grow within the EBPA space for a few more years, but our category didn’t spark much excitement. The stars of the process world were our adjacent markets: Robotic Process Automation (RPA) and Process Mining.

The RPA category produced companies that shot straight through the roof. UiPath, one of the fastest growing software companies of all times, was the poster child and it seemed unstoppable in 2019.

The first time I met the UiPath team, their revenue was less than $500,000 dollars. Just a few years later, they were pulling in several hundred million dollars, and their CEO was flying around the globe in a private jet. Their software was surprisingly simple, and I often joked that UiPath was the only company to reach $100 million dollars in revenue with a product built by just three people.

RPA was awful technology. As a software engineer, I was trained to build clean architectures, robust software, and API-based integrations. RPA was the exact opposite. It was basically a script that emulated human click sequences on the screen. Analysts called it “digital hands.”

On the positive side, RPA allowed for fast automation of tasks by non-technical people, without requiring any API access or technical data mapping. On the negative side, this was the least robust way of building software, as it could break every time the UI of the underlying application went through an update.

It was a miracle to me that RPA created so much hype. Anyways, it was here now, and it helped the whole process space appear a little sexier. We benefited ourselves from RPA, as customers needed to understand the end-to-end process context for RPA scripts and often pursued process simplification along with their RPA work.

In the Process Mining space, Celonis was the clear leader. They had started a couple of years after us and were the first ones to provide a product that could be deployed in a scalable fashion. Their marketing was phenomenal, and they captured a lot of attention as Germany’s hottest new startup.

There were other Process Mining players, too, but they were all much smaller. ProcessGold from the Netherlands had turned their Business Intelligence (BI) tool into a Process Mining product. PAFnow leveraged the Microsoft BI platform and put Process Mining capabilities on top as a thin layer. Apromore put the latest algorithms from academia to practice. And there were a handful more, like Lana Labs from Berlin, Minit from Slovakia, myInvenio from Italy. They were all tiny companies chasing Celonis from behind.

We had brought our own Process Mining product to market, but our revenue was still miniscule. The year prior, we had to refactor the product from the ground up. Making it work for a customer was a pain, and we suffered from project to project. We tried to bring consulting partners on board to help with data integration and project delivery, but most of them were already married to Celonis.

At least, we had made the decision to skip On-Premise and build our Process Mining offering for the cloud from day one. This allowed our developers to be much more productive and take advantage of cloud scalability from the start. We could iterate faster and quickly address issues whenever something went wrong with the software. Even our cloud provider, AWS, was impressed by what we built and invited us to talk about it at their conferences.

Without Sven Wagner-Boysen, one of our best developers, our Process Mining product would have been long dead. Sven was everything all at once - key architect, main engineer, presales consultant, seller, and implementation consultant. Sven was our rockstar.

Unfortunately, Sven alone wasn’t enough to compete with Celonis’ army.

Celonis had done one thing that propelled them to the top: they partnered very closely with SAP. Celonis became the killer app on top of the HANA database, which SAP heavily pushed into the market. SAP brought Celonis into many of their big customers and Celonis managed to scale their company fast enough to deliver.

I was naive to think we could easily carve out a role in the Process Mining space. The odds were stacked against us. It must have been the same combination of curiosity and determination that had driven us to enter and disrupt the EBPA market ten years earlier. I thought maybe we could do it again.

However, this time was different. Our main competitors were no longer slow-moving legacy players like Software AG and IBM. Instead, we faced venture capital-fueled startups like Celonis, who were pursuing market opportunities with incredible speed.

Even worse, we did not have a clear differentiation to leapfrog the competition this time. Back then, we were the first and only ones to go for cloud and collaboration. This time, whenever we aimed at something different - conformance checking, then cloud-native - Celonis just outpaced us. It was a painful uphill battle.

Still, every time I met Steffan, Matthias, and Leo, I highlighted our progress with Process Mining. I truly believed in our ability to build great products and in the potential for us as a company. I am surprised why they didn’t laugh at me, given the little we actually had to show at the time.

The only real difference between us and the other Mining players was that we could position the idea of a broader process transformation suite, that combined human-centric and data-driven approaches of process analysis on the one hand and bridged current state analysis with future state design on the other.

The suite positioning was a safe place for us and the Mining vendors couldn’t easily break into this. Replicating all of the EBPA capabilities we had would be an impossible exercise for them.

Our new CPO, Ron Agam, helped me organize my thoughts. He had developed the product strategy for many companies before and helped us go through the process as well.

For the first time, we ran a proper product strategy cycle, analyzing all our different options. Ron called it future-back and present-forward: Only if you start working on your 3-5 year ideas now, you can actually progress towards them. But any future vision can’t be fully detached from the present reality; you need to find the increments to your current situation that are actually feasible.

Ron knew exactly how hard it would be to succeed in the Process Mining space. He predicted a long, painful road ahead if we wanted to make it. At least, we had a good understanding now of which steps we needed to take for the coming months.

Matthias from Summit suggested a more drastic route. He recommended merging with Celonis before their valuation got even crazier. They had raised a Series B at a whopping $1 billion dollar valuation the year prior. Alternatively, we could move in with UiPath and continue under their banner.

I hated both of these ideas. I knew there wouldn’t be much left of Signavio after a short while, and we would only become a side note to the bigger story of the other company. Also, there was just a glaring mismatch in company culture between us and them.

In the meantime, many investors continued to reach out to us. They were impressed by our growth and assumed good exit routes, given all of the hotness of our adjacent market categories.

The most serious investor to chase us was Apax. They had just launched a new fund and were looking for their first investments. The fund managers were the perfect combination: Mark Beith had a strong private equity background, going deep on numbers, while Dan O’Keefe had more of a venture capital approach, believing in founder-led companies, vision, and momentum. Dan could dream, and Mark would bring him back to Earth.

When I first met Mark, I was deeply impressed by his understanding of our space. He must have done tons of analysis in the background - he knew all the players and market dynamics. He had significant insights into our competitors and how they positioned themselves. He also had a strong thesis on how our space would develop and through which angle we could win. He knew our market better than we knew it ourselves.

Apax really wanted to make an investment. We were flattered, but it was unclear how that actually fit into our trajectory just yet.

Summit was getting ready to cash in, while we, as founders, wanted to take Signavio to new heights. Summit wanted to sell, while we wanted to pursue an IPO. How could we solve this dilemma?

One day, Matthias told me about another portfolio company that had recently completed a refinancing round. A new investor had come in and bought out most of Summit’s shares. Summit had then taken a passive role, retaining just a small stake for additional upside. The new investor had reset the clock on the company, with new ambitions for future valuations and a fresh timeline to reach them.

Apax was a great proof point that there was indeed interest in the market for Signavio. Maybe they could be the ones to do a partial buyout of Summit and help us reach the next level.

I started heavily lobbying this idea with my co-founders and with Steffan and Matthias. For us founders, it meant committing to a fresh 5-7 year timeline and promising a future valuation that could exceed the €1 billion euro mark.

I approached my C-level colleagues and shared my plan. Mark Holenstein was skeptical, “Why was Summit interested in selling? Wouldn’t it make sense for us to sell as well then?”

Daniel was excited. “I love transactions!” he said, and he began preparing materials with his team immediately.

Summit’s condition was that we find an investor willing to come in at a valuation of at least €400 million euros and capable of buying at least two-thirds of Summit’s shares. Given Summit’s stake and the need for fresh money in the company, we were looking at a substantial financing round. Doing the math, this meant the new investor had to bring over €100 million euros in cash - a considerable amount for a German startup.

We kicked off the process and outlined everyone’s expectations for the transaction. The roles were clear: Matthias was the seller, and we were the future operators of the company.

Matthias asked me how much I intended to sell in the transaction. I told him I would only sell a tiny amount, if any. 

“Gero, are you sure you don’t want to take more money off the table?” He was surprised. 

I was sure. We were determined to take Signavio to a new level. I could sense the opportunity and was eager to pursue it. My co-founders trusted me and followed my lead.

Apax tried hard to preempt the fundraising process. While I liked them a lot, we still wanted to test the waters with other funds. Many were interested, but the €400 million euro valuation threshold deterred most of them. Apax was the only one serious enough to pursue the transaction.

We negotiated hard to reach an agreement. In the end, we settled on a transaction where €157 million euros would change hands: €40 million euros to be injected into the company as fresh money and €117 million euros for secondaries, most of which went into Summit’s pockets, some to us founders.

As part of the deal, we also offered our employees the opportunity to sell up to 25% of their vested stock options. Many employees were happy to see some liquidity. For the first time, the stock options had tangible value.

The last mile turned out to be exhausting, as Apax needed co-investors to bring the €157 million euros to the table. They couldn’t do the deal alone, as this would have been too large of a single investment for their fund.

They suggested multiple potential co-investors who would then passively sit in the background for the coming years. One of their first suggestions was CIC, the Chinese sovereign wealth fund, but we didn’t want such a massive Chinese presence in our shareholder base, so we declined. In the end, GIC, the Singapore sovereign wealth fund, agreed to co-invest. Singaporean money was acceptable to us; we even had a small team there, and I liked the country a lot.

Unfortunately, Apax and GIC still couldn’t fill the entire €157 million euros, so we needed a third fund to participate. Luckily, DTCP (originally Deutsche Telekom Capital Partners) was eager to join and wrote the check for the remaining amount. We already knew DTCP well, as they were also a significant investor in LeanIX.

When we announced the fundraise to our team, I created a slide that showed the €157,000,000 euros as a figure across the entire screen. It was an unimaginably large amount of money, and people were amazed that such a sum would be spent on Signavio, our “nice little process modeling company.”

To me, it felt like a game. Investment activity in software was heating up, with higher and higher sums flowing into our space. Now, we were among the beneficiaries, having just raised an incredible amount of money.

Daniel remarked that we now had a big responsibility. 

“What responsibility?” I asked him. 

“The investors gave us a serious amount of money.” He replied. “These are people’s pensions and university endowments. It makes a big difference whether we return their money or not.”

I thought about it for a moment but didn’t fully grasp it. They made a bet, and bets can go wrong. Our job was to focus on building a great company. If we succeeded, they would benefit, too. If we failed, we would suffer more than they would. So far, we had never wasted any money.

I didn’t feel any burden. I felt unbeatable, ready to take on any challenge that would come our way.

Working with Mark and Dan from Apax was a fresh breeze of air. For the first time in a while, we were encouraged to dream big and freely explore all of the “why don’t we…?” questions that had been simmering in our minds. Their support and expansive thinking helped to widen our perspective on what Signavio could become.

As part of his research for the transaction, Mark had conducted a comprehensive analysis of all of the software companies operating in the process space. We began discussing which of these companies might make sense for us to acquire and reach out to. Together, we approached a few of them to gauge their interest in being acquired and to understand what synergies might exist.

ProcessGold, for example, was already well into talks about getting acquired. They had received a strong offer from UiPath and were seriously considering it. We saw potential in their product, especially when combined with our own Process Mining solution. Intrigued, we jumped into discussions and put significant effort into courting them. The fit seemed good, and at one point, we even became their preferred acquirer. However, during the due diligence process, some doubts arose. As we took a closer look, the integration and value proposition seemed less compelling, so we ultimately decided to step back. We left it to UiPath to finalize the acquisition.

I reached out to my old colleagues at HPI who were working at Lana Labs, as well as the founders of PAFnow and Apromore. While each of these companies had interesting elements, none of them aligned well enough with our vision to justify an acquisition. In the end, we decided that building our product internally was the best course of action. Developing it ourselves would allow us to retain complete control and flexibility, essential for navigating the fast-moving Process Mining space.

With our pockets full, we chose to act boldly. We gave the teams a green light to accelerate hiring and scale our operations, confident that massive growth awaited us in 2020. It felt like the right moment to go all-in and capitalize on our momentum.

For the Christmas parties, I toured through our various locations. Our teams from France and the UK gathered in London, all dressed in their ugly Christmas jumpers. Martin Adams played the piano under the Christmas tree at Marylebone station, and the entire train station joined in when we sang "Angels" by Robbie Williams. It was a wonderful evening.

For the Christmas party in Berlin, we rented the SoHo House to cap off another phenomenal year. We arranged good music and several show acts to make it a memorable night.

Unfortunately, this party ended in disaster. What was meant to be a classy burlesque show  turned out to be a cheap strip performance, much to our surprise. While a few people cheered, most of our colleagues were shocked. We quickly canceled the show, and I went on stage to apologize to the team.

I hadn’t been involved in organizing the Christmas party and never imagined we’d run into such a problem. We had engaged a burlesque group in previous years who were entertaining and tasteful. This artist, however, was much different and a total surprise.  

In the days following the event, we had serious conversations about what was acceptable at Signavio and what would cross the line. Our team was committed to creating an inclusive workplace, and in this instance, we had failed badly.

I was reminded once again of how perceptions of fun and what is acceptable can vary greatly. It takes sensitivity to strike the right balance. We didn’t want to stop celebrating, but we agreed that more care and thought were needed going forward. We decided to actively involve more people in planning future events to avoid similar issues. The team appreciated this renewed focus.

At least we had the Christmas break ahead to let things cool down and prepare ourselves for another great year.

Business performance at Signavio remained strong. We had solidified our position in the market and could take serious pride in what we’d achieved together. We closed the year with €40 million euros in ARR and a team that had grown to nearly 400 employees.