In this video, I look at the valuation of start-ups from two different angles: from the point of view of the founder/owner and from the point of view of the investor/business angel.
The presentation consists of different parts:
- 1. Point of View of the Owner / Founder
- 2. Point of View of the Investor
Welcome to The Dark Magic of Start-Up Valuation:
Who am I?
My name is Bruno Lowagie.
- I’m the original developer of iText, an open source PDF library.
- My wife and I are the original founders of iText Group, a company with subsidiaries in Belgium, the US, and Singapore.
- Together, we grew the business from start-up to exit.
After leaving the company, I wrote a book about the whole experience. This slide is some shameless promotion for my book:
I want to apologize in advance if you are looking for an academic explanation
about how to calculate the value of a company using different methodologies.
I don’t have an MBA. When I want to know how much a company is worth, for instance based on the discounted cash flow method, I involve a specialized accounting firm. In this article, I’ll look at valuations from a practical point of view, first as a technical founder, and eventually as an investor.
Types of Valuation
I make a distinction between these three types of valuations:
The first type is the psychological valuation. Most founders I know have a specific number in mind that reflects how much money they think their startup is worth. That number usually isn’t based on facts, calculations, or reason. It’s an irrational number, not in the mathematical sense, but in the sense that many investors would consider it unrealistic. This valuation is important, because it reflects the minimum price an owner is willing to sell for. If you look at this valuation from the point of view of the acquirer, it’s the maximum price a potential buyer is willing to pay.
The second type, the calculated valuation, covers all the most common methods accountants use to come to a what we call a “fair market value”. There’s the multiples method, the discounted cashflow method, the assessment of the cost to reproduce the assets of the company, and so on. I’ll briefly discuss some of these methods in the different case studies, but I won’t explain each method in detail.
The third type, actual value, is based on the expression “A thing is worth what someone is willing to pay for it.” In some cases, the buyer can get a company at a bargain price; in other cases, the buyer will pay too much, but a sale is always an important snapshot in the valuation history of a company.
First I’m going to look at valuations from the point of view of the founder / owner.
For a technical founder like me, the valuation process of a start-up often resembles dark magic. Valuations are based on science, but they usually require a leap of faith because there are many parameters of which the value is unknown. If I would have to write valuation software, I would introduce a
TooManyAssumptionsException that would be thrown in 90% of the cases.
Nevertheless, I hope that I can shed some light on the matter by presenting a case study I am very familiar with: the valuation of a company I founded and sold, iText Group NV.
Evolution of the Value of iText Group
I’ve compiled an overview of how the value of iText Group evolved over the years.
The blueish-green lines show the minimum and maximum psychological value my wife Ingeborg and I had in mind at different points in time. The red line shows snapshots of what could be perceived as the actual value of the company. It’s important to note that the scale of the Y-axis is logarithmic: it goes from 0 to 1 million euros, from 1 million to 10 million euros, from 10 million to 100 million euros, and from there on to Unicorn status. The growth of iText Group was more impressive than you might assume at first sight when looking at the chart.
The highlighted values reflect the moments we sold packages of our own shares. We sold about 75% in 2015, about 5% in 2018, and about 20% in 2020. Those are not the exact percentages, but they are close enough.
There are some other values on the chart that I will explain later on, but let me start at the beginning.
I started iText as a hobby project. My philosophy was: if I don’t develop a business to generate revenue, I won’t have any obligations toward customers. That was very naïve. My current opinion is that the “No Money, No Worries” philosophy only works if your project isn’t successful.
When a Hobby gets Serious Business
The moment your project gains popularity, thousands of people start asking technical questions.
I tried to fix this problem by writing a book about iText,
so that I wouldn’t have to answer the same question over and over again.
This book made iText even more popular. This resulted in an abundance of legal questions.
I tried to solve this by doing a thorough IP review.
Eventually, the “burden” of maintainig the iText project got so heavy that I had no other choice than to create a company for the technology. I realized (just in time) that I would have to start generating revenue if I wanted the project to survive. See my "Open Source Survival" video if you want to know how I created a business for iText.
The First Time We Heard About M&A
In 2007, the year before my wife and I founded our first company,
I received a call from someone called Hari with the message that the company he worked for,
Hyperion, was interested in acquiring iText. At first, I thought he was making fun of me,
but he assured me that this wasn’t the case. Hyperion had been (or was at the verge of being)
acquired by Oracle, and the company wanted to acquire iText in the context of a buy-and-build strategy.
I told Hari that I had never thought of selling iText before and that I was totally inexperienced
in these matters. He advised me to involve an M&A consultant,
so I called a random company specialized in M&A.
Their first question was: “What is the name of the company you want to sell?”
When I explained that I didn't have a company yet, let alone a business, I was met with disbelief.
“How do you expect to sell a company if it doesn’t exist yet? What deal size would we be talking about? This doesn’t make any sense!” said the M&A consultant.
The conversation with Hari lost momentum and the opportunity window closed due to my lack of experience,
but the unexpected invitation to talk about a possible acquisition, made my wife and I think
about the value of iText. After the first couple of calls with Hari, we thought:
if Hari asks us how much money we want for iText, how much would we ask?
We came up with two numbers: let’s ask for a million and a half euros, so that we can sell for one million.
These numbers weren’t based on any metric. It were just two nice, round numbers.
We thought it would be cool to make a million,
but the conversation died before we could share our financial expectations.
In hindsight, looking at how the value of iText evolved, you could say that one million would have been a bargain price. Just look at how the value evolved in the years that follow. However, knowing what I know today, I think that a value of one million for a project that didn’t generate any revenue, wasn’t backed by a company, and therefore didn’t have any employees on payroll, would have been considered extremely unrealistic.
In 2008, we founded the first iText company. You can read the full story in my book Entreprenerd.
First Term Sheet
In 2010, we received a term sheet from a competitor out of the blue.
That competitor was offering half a million dollars for the business
(distributed over four years) and a salary of a quarter million dollars a year (also for four years).
We didn’t even start negotiating.
We had made a revenue of about $300K in 2009.
In the fourth quarter, the revenue dropped to zero, but we made a license change in December,
and sales were picking up. We made about $700K in 2010. It may be hard to believe,
but up until October 2010, I combined the iText business with a full-time day job.
I was confident that sales would boom once I quit my job.
The competitor wasn’t aware that I already had made a forecast of the effect of that decision.
Looking at the term sheet and including the salary,
I would have to work four years to make 1.5M dollars while I was expecting
at least one million in revenue in 2011.
I think half a million was the psychological value the competitor was willing to pay. Given the fact that none of iText’s numbers were public, this value probably wasn’t based on any metrics. My wife and I might have started negotiating if the offer had been closer to the value we had in mind ourselves, but the numbers were too far apart. We decided not to engage in a conversation because we feared it would be a waste of time.
First Valuation Exercise
Surprised by the unexpected term sheet,
Ingeborg and I decided that it might not be a bad idea to involve an accountancy firm
so that we would get an official valuation of our company.
We hired PKF and they calculated the accounting value of the company based
on the most recent financial statements.
They didn’t consider using any other method because there wasn’t sufficient historical data.
PKF concluded that iText was worth one and a half million euro. While we would have been very happy with that result four years earlier, we were very unhappy with the number produced by PKF.
I had hoped PKF would also look at other approaches, but they didn’t, so I did some work myself.
- In those days, CollabNet, a company cofounded by Tim O’Reilly, offered a free service to calculate how many man years were spent on the development of open source projects hosted on SourceForge. This service is no longer available for free, but in 2011, CollabNet estimated that the effort to rewrite iText from scratch was about forty-three man years. Using the constructive cost model (CoCoMo), I valued the iText business at €3 million euros.
- I also looked at recent acquisitions made by Google. Benchmarking the size of those deals, I could justify a value of €3 million to €10 million euros.
I asked PKF to record the results of my own research in the valuation report, which they did, but they commented that these amounts weren’t trustworthy given the volatility of the software business. In the conclusion of the report, PKF maintained the price of €1.5 million.
Personally, we thought our company was at least worth 3 million euro, possibly even 4.5 to 5 million.
This time, we had done an effort to justify this psychological valuation with arguments,
but our arguments weren’t strong enough yet to convince “professionals.”
This first valuation exercise taught us what we needed to do to make our arguments stick. Our subsidiary iText Software Corporation (ISC) didn’t have any employees. We worked with sales people on commission for 100% of our sales, all of which were realized in the US. When we reached a revenue of $1 million dollars in the summer of 2011, we decided to create a second subsidiary in Belgium. iText Software BVBA (ISB).
We hired our first employees on payroll and we had two consecutive quarters with a revenue of almost $900K.
First M&A project
In 2013, we started an M&A project, not as much with the intention to sell the company,
but to discover what was needed to make our company “investor-ready.”
We told the M&A consultant that we wanted to ask potential acquirers 15 million euros,
but that we would settle for 10 million.
We didn’t succeed in selling the business, but we considered the project a success because it was an interesting learning experience.
The M&A project resulted in a long list of action points.
One of those action points involved the reorganization of our companies.
Our companies had grown organically.
This resulted in a company structure that was difficult to understand for potential buyers.
Ingeborg and I owned two companies directly: 1T3XT BVBA (aka iText Group) and Wil-Low.
ISB and ISC were owned by Wil-Low, not by iText Group.
There was no direct link between the company that owned all the assets and the companies doing the sales.
No one understood what Wil-Low was about; that company was considered dead weight.
We were advised to sell ISB and ISC to iText Group,
so that a potential acquirer could buy a single entity to acquire the complete group.
We hired BDO to prepare a valuation report for ISB and ISC. The accounting value of ISB was approximately €600,000; the value of ISC was €2 million euros. Wil-Low sold both companies to iText Group, and we liquidated Wil-Low. You can see the resulting structure in the chart on the right. iText Group was the company owning the assets, and the parent company of the sales organizations. As you can see, we created a third subsidiary in Singapore in 2015.
Thanks to this operation, Ingeborg and I cashed about 1 million euros each after taxes. In 2013, this money gave us the peace of mind to take the risks that were needed to further grow the business. I notice that investors often underestimate the importance of this peace of mind for technical founders. They want to acquire shares in a company through a capital increase and usually aren’t open to purchasing shares directly from the owners. Maybe they fear that the founders won’t be motivated to further grow the company once they’ve cashed. In my case, having financial security gave me a boost to go full force with iText. The money didn’t make me “lazy”, on the contrary.
In 2013, our company was worth at least two and a half million:
that was the valuation of the subsidiaries ISB and ISC without taking the mother company iText Group into account.
The psychological value was 10 to 15 million euros.
Apart from reorganizing the companies, there were many other action points, such as opening an additional office in Boston, hiring a dozen more employees, installing a board of directors, and so on.
Executing these action points involved significant investments, which had an impact on the EBITDA.
The EBITDA dropped from 2.6 million euros to 2.3 million euros,
and from 58% of the revenue to 47% of the revenue. Those are still good numbers, though.
The energy we spent also had an impact on our revenue. You can see that we no longer grew with double digits. There were two other reasons why the growth of the revenue dropped temporarily.
- We switched from paying our external salespeople a commission to giving them a discount.
- We implemented some new rules to calculate our recognized revenue based on feedback from the auditor we hired. Because of this, part of the cash received in 2014 was deferred to revenue for 2015.
We hoped that this drop would be a temporary effect—as you’ll see in a couple of slides, we were right to think so.
Preparing for an Exit
On the first full board meeting in January 2014,
I told the board members that I would consider the board a failure
if we didn’t succeed in selling the business before December 2016.
In the same year, we won BelCham’s Most Promising Company of the Year Award and the Belgian edition of Deloitte’s Technology Fast 50. Those would be the first of a long series of awards. At the end of 2014, we started contacting consultants for a new M&A project.
After a long conversation with one of these consultants,
sharing our financial results and forecasts,
we received a spreadsheet with valuations that were calculated using different methods.
Second Valuation of iText Group
- The consultant looked at the deal size of recent acquisitions of companies similar to iText Group. He found deals of as low as €4 million euros and as high as €50 million euros.
- The discounted cash flow (DCF) method is used to calculate the value of a company based on the concept of the time value of money. All future cash flows are estimated based on an expected growth percentage and discounted using the weighted average cost of capital (WACC). Doing the math and playing with these two parameters, the M&A consultant estimated that iText Group was worth €13 million in the worst case and €26 million in the best case.
- Finally, the consultant used the multiples approach, which consists of multiplying a key figure with a specific factor, assuming that the ratio between that figure and the actual worth of the company is a constant value. The factor can for instance be based on the ratio between the TTM (Trailing Twelve Months) revenue and the current market value of publicly traded companies. When using multiples that were standard for software companies on the EBITDA, the result varied from €4million to a whopping €115 million (that value is off the chart). Applying an average factor on our revenue resulted in a value of €20 million.
When we looked at all those values, we were confronted with a fork of more than €100 million. We reduced that fork to €10 million. We assumed that our company was worth €25 million euros with an error margin of €5 million.
This was also the psychological value once we started the M&A project for real:
We’d ask for €30 million but accept a deal of €20.
In December 2015, we succeeded in selling 75% of our shares at a total valuation of 27 million euros.
I signed a commitment to stay on board of the company for three years.
After the dip in 2014, the growth percentage stabilized.
We had an almost constant revenue growth of about 25% every year,
and we were able to keep the EBITDA percentage above 40%.
As CTO, I was working on initiatives that could result in an even bigger growth. Unfortunately, I had to leave the company before the end of those three years because of a dispute with a majority shareholder.
Aftermath of the Exit
In 2018, that shareholder executed a call option and bought a fifth of my remaining shares
for a price we had agreed upon in December 2015.
That explains why I sold that part of my shares at a value lower than the actual market value.
That’s perfectly normal. That's how a call option works.
But the majority shareholder also tried to make me sell the rest of my shares at a valuation of 51 million euro, which was below the psychological value the shares had for me.
I went to court, and both parties involved an accountancy firm.
I hired KPMG and based on their assessment, the company was worth 157 million euros. The accountancy firm hired by the other party wrote a report in which they tried to justify a price of 51 million euros. I can’t disclose the details, but if you look at the discounted cashflow method, you can get such discrepancies by making different assumptions of the expected growth. KPMG assumed that iText Group would keep on growing. The defendant expected that the growth would slow down.
I hope you now have a better understanding why I consider the valuation process a form of dark magic. When we calculated the value of our company in 2014, there was a fork of more than 100 million euros between the lowest value and the highest value. When we went to court, two different accountancy firms came up with a number that was also 100 million apart.
The court appointed an expert who vetted the valuation reports of both parties, and redid the math.
This resulted in a value of 95 million euros. Eventually, we settled for what the expert considered being the fair market price for the shares.
As I left the company in 2018, I don’t get the consolidated results of the group of
the most recent years anymore, but when I look at the publicly available financial statements
of the mother company (source: openthebox.be),
iText Group, I see that the business is going better than anticipated
by the accounting firm that valuated the company for the defendant.
Of course, one can always argue that no one could predict the growth realized in 2018 and 2019. What is true in general, is certainly also true in the context of valuations: It’s hard to make predictions, especially about the future.
As you’ve noticed, I don’t know much about the theory behind specific valuation methods,
but I did gain some experience in the matter of valuations in general.
You’ll find the full history behind the first part of my talk in my book Entreprenerd.
In the second part, I’ll discuss valuations from my personal point of view as investor.
What do you do when you are a multi-millionaire and you don’t have any company to work for anymore?
I could have decided to buy a house on the beach and never work another day in my life,
but that’s not in my nature. I get bored easily, and I don’t think I would survive doing nothing for more than a day.
My wife left iText Group after the first exit in December 2015. She started several real-estate projects. She bought a handful of houses and renovated them. When our son proposed to start a table-tennis club, we created QLT, a company dedicated entirely to table-tennis.
This was a much bigger project than my wife’s real-estate projects. We invested about 4 million euros (no subsidies involved!) in building a high-quality venue. It was the first project we started together as a family in the fall of 2018. QLT was supposed to open in April 2020, but Covid decided otherwise. Fortunately, we saw this as a long-term investment from the start. Even once QLT can operate at full force, we know we’re not going to get rich promoting table-tennis, but we assume that the value of the bricks and mortar will remain stable. In Europe, most of the money goes to soccer; we wanted to invest in a sport that didn’t get the TLC it deserved. We’re now looking for a similar project with a similar budget in the cultural sector. We thought we had found one, but we had to abandon the idea, among others due to Corona circumstances.
These are “family projects”; we work with partners, but we remain in charge. Apart from these family projects, we also wanted to invest in other companies, but we wanted to gain more experience before doing so.
Venture Capital Firms
We thought it would be a good idea to become limited partners in Smartfin Capital I,
a 75-million fund that invested in companies such as MariaDB and NGData.
For Ingeborg and me, it was interesting to be at the other side of the table for once.
In the past, we had been presenting our company to VCs;
now CEOs presented their company to us on a yearly basis.
We realized that this was an investment with a much higher risk than our family projects. We were aware that we could lose all the money we put into this fund. We’re rather conservative in the sense that we would never make such an investment with money we can’t miss. We also know that the Return-on-Investment could be much bigger when successful. As a matter of fact, we never had to cough up the full amount we committed. One portfolio company, Newtec, had a spectacular exit; another company in the portfolio, UnifiedPost, had a successful IPO. We are now investing the money we make in Capital I in a second fund, Smartfin Capital II.
I had another learning experience with Volta Ventures. Once in a while, Frank Maene, General Partner at Volta Ventures, asks me for my opinion about a startup. In the case of ChiliPublish for instance, Frank also asked me to perform a technical due diligence on the company. When I gave a positive advice, Frank told me: “Put your money where your mouth is. I’ll follow your advice on condition that you co-invest.” This was a unique opportunity for me to acquire shares that weren’t publicly traded. I am a minority shareholder and my shares diluted over the years because I chose not to participate in follow-up investments, but the valuation of the company is growing year over year. Maybe one day, I’ll get a nice return on my rather small, early stage investment.
In both cases, Smartfin and Volta, financial return on investment wasn’t my main goal. I mainly want to get regular updates about what is going on in the technology sector now that I am no longer the owner of a technology company myself.
I learned from the VC firms that a large part of their success is due to the clear focus
they have and I defined a scope for my own investments.
Above all, I want to learn from the companies I invest in. This implies that I want to be involved in the business to some extent. I don’t want to be courted for the money I can bring in; I want to actively contribute to the success of the company. That’s why I limited my scope to early stage companies with some revenue, but not sufficient revenue to convince a VC. I have a preference for companies with technical founders who are building a deep tech solution, preferably an open source solution (although that’s not mandatory). For instance, I’m not interested in apps or games, but In companies that create the technology to produce apps, or the engines that are used to create games.
I don’t do angel investments in companies that I don’t know. I offer two ways for companies to introduce their business. Either they pass through an incubator or accelerator first, or they need to hire me as a consultant first.
- To be admitted to an incubator or an accelerator, founders usually have to go through a selection process. Having been part of a panel of judges at a couple of accelerators, I know that startups need to meet some minimum criteria to make the selection. This saves me time: I don’t have to do that minimum due diligence on my own. Ideally, the incubator/accelerator asks me to be an adviser of those companies that are within my scope of interest. (So far, I've never been paid for this advisory role, and I don't deem that necessary either.)
- I don’t invest in companies that don’t want to join an incubator or accelerator unless they hire me first. I am very picky when startups ask me if I want to work for them. It’s important to understand that I didn’t put the “hire me first” rule in place to make money at the expense of startups. I use this rule mainly to test two things: the level of commitment and the coachability of the founders. If founders don’t want to pay me to help them, I see this as a lack of commitment. By working with founders, I can discover if I can coach them before I invest.
I typically work with tickets of $50K (to start with) and a maximum of $200K (in total). If a company needs more money, they shouldn’t ask me as a business angel, but go to a VC.
I don’t like to negotiate. I am very binary when I make a decision.
- First I ask the founders for the psychological value of their company. If that price is much higher than what I am willing to pay and much higher than what they can justify using objective arguments, I immediately pass and I don’t give founders a second chance to adjust their expectations.
- If the price seems right, I involve an accountancy firm, usually BDO, to calculate a valuation. I do this for different reasons. For fiscal reasons so that tax or other authorities can’t complain that I purchased shares below their market value. I also see it as a good practice to check if the start-up did its homework well. Often BDO finds what I call “honest mistakes” made by the founders. Small errors in the annual accounts, wrong assumptions here or there, certain risks that weren’t accounted for,… I discuss these with the founders. If they see this effort as a learning experience, we commit that we’ll fix these issues going forward. If the founders get defensive, I pass on the opportunity and I don’t give them a second chance.
Since I mostly invest in early stage companies,
BDO gives them the treatment I got from PKF: they mostly work with the accounting value.
I then ask BDO to add a certain amount of “goodwill”
to get to a price that corresponds with the psychological value both the founders and myself are OK with.
In some cases, I purchase shares, like I did with ChiliPublish. In other cases, I work with a convertible loan. Sometimes, it’s a combination of both. Let’s take a look at three specific cases.
Three Case Studies
I have known Johan Vos, one of the five founders of Gluon Software for a long time.
When I left iText, Johan asked me if I was interested to become their CTO.
I refused the offer, but the technology of the company interested me. I made a counter-proposal.
I wanted Gluon to hire me to help them implement some changes on the business front.
Gluon consisted of two companies that were closely entangled. One company had a clear ownership; the other one had several sleeping and even one deceased shareholder. I helped Gluon clean up the ownership. Different people owned IP that hadn’t been transferred to the company. Once this problem was fixed, I looked at the business. The bulk of the revenue was generated by professional services. I tested the willingness of the founders to move from services to product.
I invested in the company to buffer the loss of revenue due to phasing out professional services so that there was more time to work on the product. We agreed on a valuation that was higher than the accounting value, but much lower than a later valuation by KPMG that was based on the production cost and comparable transactions. The founders agreed to this lower valuation because of the expertise I brought to the company. We agreed to split the investment in two parts: a share purchase and a convertible loan. The share purchase resulted in me getting 5.66% of the company, which is less than any of the original founders (and that’s OK for me). The convertible loan meant that I could still get a nice premium in case the company realized an exit.
My personal goal with the company is to help the company develop a business with the product, and keep a handful of VCs up-to-date on the progress we’re making so that they can invest in the company as soon as the time is right for all parties.
The founders of Stormbee tried talking to different investors,
but without any success. One of the people they talked to, referred them to me.
I invited them for a free intake session, after which I decided to work for them for an hourly fee.
We worked together for about a year, and when it was time to go to the next level, we started talking about a possible investment. Putting a value on the company was part of the project, so we already were in agreement on the market value. We also had an oral agreement on how much money I would invest, but then the company received a much better offer from another investor.
Although I missed the opportunity to invest, I consider the extra money the company was able to secure to grow the business as a good result for the company. And that’s what matters most, isn’t it?
I met the founders of Tengu after
I was selected as their mentor at an accelerator, Watt Factory.
I introduced them to a CMO from my network, and that CMO joined in the capital (she became a co-founder).
After Tengu graduated from Watt Factory, it was accepted by Birdhouse, another accelerator where
I’m a “scale-up angel.” I was able to monitor the company based on their activity at Birdhouse
and thanks to conversations with the CMO.
When the CMO told me the company was running out of money, we agreed that I would loan them money using the concept of a “win-win-lening.” This is a specific loan that only exists in Belgium. You can’t convert the loan to shares; you get an interest if the company survives, but you risk losing the money if the company goes bankrupt. To mitigate that risk, the Belgian state gives you a tax exemption on your personal income tax. This money allowed Tengu to survive a couple of extra months that were used to go for a bigger investment from PMV, an investment firm owned by the Flemish Government.
PMV only wanted to invest if Tengu succeeded in finding a certain amount of money from private investors. I decided not to invest the full amount that was required, but Tengu found two additional business angels with skills that were complementary to mine. Together, we granted a convertible loan so that Tengu qualified for a much larger investment by PMV. In this case, I didn’t ask BDO to calculate a valuation, nor did I ask my lawyer to draw up the terms. All business angels agreed to the terms proposed by PMV. One of the conditions was also that Tengu would actively search for VC money on the short term.
I hope you’re not disappointed when I conclude that there’s no general rule for start-up valuations.
Nevertheless, I hope this isn’t the only takeaway from my talk.
- As owner/founder of iText Group, I noticed that the valuation process was different in different stages of the life cycle of the company. The more mature the company, the less dark the magic got.
- As an investor in early stage startups, I see that the valuation consists mainly of two parts:
- The accounting value of the company, which is usually very low,
- The goodwill that is added to meet the psychological value of the seller.
When am I interested in the company’s technology and the business? This is a very personal criterium, and I notice that some founders find it difficult to understand that I’m not interested in what they do. For instance: I’m not a gamer, so why would I invest in a company that creates games? When I invest, I need to have the feeling: even if I lose all the money I invested, I still gained plenty of experience while actively helping the founders.