There are few people as qualified to talk about the landscape of software investment as Bill Janeway. On joining Venture Capital firm Warburg Pincus in 1988 he built their IT Investment practice, which focussed heavily on Enterprise Software. It’s fair to say that they were pretty successful, investing in companies like VERITAS and BEA…
In this talk, Bill looks at the state of the software investment market and comments on how the move to SaaS has affected the enterprise market.
Learn how great companies are run
At BoS we run events and publish content that is highly valued by anyone trying to build, run, and scale a great software company.
Sign up for a regular dose of actionable and useful content:
Bill Janeway: First thank you Mark very much for introducing me with a moment of unbridled commercial mechanicalism. “Doing Capitalism In The Innovation Economy” Second Edition, fully revised and extended is now available on an Amazon Web site near you. But I’m going to talk about today is the most relevant extension of this book. We’ve been back and forth over this. “The Rise And Fall Of Enterprise Software”. And as Mark said this is a world I spent a generation in which I was on leave from the academic world practicing at the frontier. Once upon a time enterprise software was the most wonderful place to invest. It was a time when in building the infrastructure software that enabled distributed computing we had the opportunity to participate in the fundamental rearchitecting of the entire computing universe, which had grown up over the previous 30 years based on centralized mainframe, and many computers with dumb terminals, all in a vertical proprietary stack which you could not approach from the outside. And we participated in the transformation into the horizontal layers of technology linked together through open APIs and protocols. And that for that work we were very richly rewarded. Now back when I was an academic the first time round I wrote my thesis on 192-1931, and that meant that in the late 1990s I was privileged, I had the unfair advantage of having seen the movie before. So this is what was the iconic new technology transformational vehicle for speculative excess in the late 1920s – The Radio Corporation of America. And you can see that between 1926 and 1929 and then on to 1932 it had quite a ride. And I’m pleased to say that Warburg Pincus participated in solely funding half of what became Veritas Software which enabled data management across infinitely extensible networks and then BEA systems which became the go to platform for transactional applications, moving from client server to Internet distributed systems. And again between 1996/7 and 1999/2000 we had the opportunity to ride that train once again. So when I speak about the transformation of enterprise software from one it was the most wonderful place in the world to invest I do so with a certain nostalgia, thankfulness, gratitude to all those mindless investors who got the hint somewhere around 1997/98 that what we were doing was actually going to enable a new economy. And it was going to change everything – took a little longer than we might have hoped, it always does. It always does. But I want to spend a moment on what were the key factors that enabled that extraordinary explosion of creativity and transformational economic impact.
The first was the fact that right through this world we had standard platforms. Initially of course the standard platforms were the proprietary platforms established by IBM. IBM had two or three of them, Digital had one data general had another, HP had a third. And then in the in the 90s we began to construct a new standard platform. It was much more open, it was radically open. It began with Unix given to the world under the orders of what was then an active aggressive Antitrust Division of the United States Department of Justice which provided an enormous benefit to the world by requiring that AT&T license on a fair and non-discriminatory basis all the technologies that developed at bell labs that were not exclusively dedicated to communications and it went with a host of other like Ethernet other other protocols all the protocols of the Internet that combined to make for a platform a target platform that companies like Veritas and BEA we could focus. Second crucial factor was of course the radical decline in the cost of hardware. Now, I’m really not making these numbers up. In 1960 the cost of a MIP which was of course a Mainframe MIP with $10,000,000. $10,000,000 per MIP. By 1980 it was down to $400,000. By 1990 it was down to $20,000. By 2000 it was down to $2,000 a MIP as IBM had to respond to the fact that companies like BEA were enabling mainframe enterprise class applications to be run on these flexible distributed systems. And as another example of this – the first single user computer that you would recognize was the Xerox Alto which became the Xerox Star – The Xerox Star was introduced at a price of $16,000 which thirty five years later would equal about $44,000 for a PC. In 1984 the first Mac was priced at $2,500, which shows a radical decline in hardware. In fact what was going on was hardware was in the process of becoming commodified. And value was shifting as Tim O’Reilly my friend and guru of all things digital said hardware value was moving from hardware to software.
The next factor was that we had smart customers. An external and internal ecosystem of customers supporting teams, professional I.T. departments. So the infrastructure software companies like Veritas and B.A. could sell to a central I.T. department which was in turn responsible for adapting installing and maintaining the code for business users. Application & software companies could offload much of the low margin professional services needed to customize and to adapt their code to the environment of the customer and not have to do it themselves. And of course, the most important factor was a revenue model that across you might think the entire history of global commerce was unique. I mean it was so much too good to be true the fact that it lasted for something like 30-40 years is actually amazing. The perpetual, non-exclusive license paid upfront in cash by the customer for code that was inevitably incomplete, a work in progress, which came with 20 percent of that license fee due annually so that the vendor would keep the stuff running as the environment in which the code was running continued to evolve… well nobody knows what this model came from. The best story I ever heard about it was that the German software company which somehow still persists SoftwareAG which had a very early hierarchical database product called AdaBase. The president of SoftwareAG came to the US and he managed to sell a $50,000 license at a base to an unnamed enterprise customer and everybody who was in the game looked around and said “Boy that sounds really cool”. Now we have a reference and we can tell the next customer. The financial consequences of this model were absolutely terrific. It meant that the rich customer was funding the poor startup with no issuance of equity, no dilution of ownership, and of the upside rewards available to the investors and the entrepreneurial founders of the business – the cheapest capital imaginable.
So venture capital was needed indeed upfront to buy development hardware and software to fund operations until a first always incomplete product could be brought to market but then upfront customer license payments funded the build out of the business. The sales force, the marketing, the customer support, the back office, with reasonably good execution. You could reach the promised land of positive cash flow from operations, liberation from the vagaries of the capital markets for only $20/30m – today $20m = $30m. So the first thing that’s changed. Obviously the first critical change is in the revenue model. I remember reading Salesforce’s prospectus and noting that at the time as in the S-1 prospectus at the time of its IPO in 1999 just as the bubble reached its absolute peak. It had taken more than one hundred million dollars not 20 million dollars of risk venture capital to get to the point approaching almost at positive cash flow reported profits which usually leads positive cash flow in those days.
For the enterprise of course, the shift from the sale of perpetual non-exclusive license plus maintenance to software as a service means that the poor startup is financing the rich customer. Now for the rich customer it’s much easier to buy is incurring an operating expense it’s not a capital purchase. It doesn’t have to go to the board it doesn’t have to go very high up in the organization to make the to make the commitment. But the for the poor startup it means a radical increase in the capital required to reach positive cash flow chart. For consumer oriented businesses of course software business, the shift has been from licensing shrink wrapped software to advertising supported which in turn stretches out through IBM. In this case means instead of going to it’s obvious we’re watching it now instead of going from needing $1m/2m to get to positive cash flow you need $1bn/2bn of risk capital to get to positive cash flow now.
Second fundamental factor in the state of the software the enterprise software industry today is of course the rise of open source and cloud. So it’s one thing to go from perpetual license upfront fees which in the case of enterprise software could easily reach a million dollars for a company that was only two or three years old to software as a service payments cumulatively month by month over a period of years to get to the same amount of money. But when you’re competing with free it’s really tough. It’s really tough. I remember at BEA when Jay Boss came on the market and we had to really think about how could we add how could we protect a price point against a free app server offering. But the cloud also cuts two ways. So on the one hand obviously for developers it again radically reduces the cost of launching software. And in fact given the combination of free open source software tools and the cloud, it means that you know as I say to my students you guys don’t need your parent’s credit card you can do the startup on your own. That is, you can get to running code and launch something at a radically low price. But the cloud is also doing something else. It’s pre-empting market space that startup enterprise software companies use to address. That is as customers enterprises even banks move towards cloud resources, they’re not buying the sort of functionality that they used to buy on a case by case basis. So we have GitHub, where freemium offerings are kind of table stakes for being accessible to customers. We’ve got the cloud providers competing to thicken their commodity computing services by adding more and more functionality more and more richness. And of course the most appropriately and relevant new technology is machine learning. And of course machine learning gets you back to something I’d be happy to talk about if anybody is interested. The first hyper wave of AI, the first looping through the world of the promise of AI which I lived through long before BEA was founded. But Microsoft is offering A.I. and machine learning machine learning morphed into A.I. and even preempts it in the case of Google Cloud where it’s become a kind of attempt to establish a radical differentiation.
So these changes – that change in the revenue model, the change in the market space – these create enormous challenges for entrepreneurs in this startup enterprise software industry. One of the most significant is that as I say the cost of building a business hasn’t changed. The cost of writing code the cost of applying Agile technologies and rapid iteration using free tools and running and rent by the function point, rent by the cycle, rent by the megabyte of storage resources from Amazon or Microsoft or Google, that cost has declined hugely but the cost of building and maintaining an enterprise class salesforce has risen enormously. Plus there’s another major, major change – the customer has changed. IT today in major enterprises is principally concerned with two things – keep the lights out and keep the bad guys out. Since they can’t succeed in keeping the bad guys out, it’s then perhaps morphed into a third which is remediate after the bad guys have gotten in and meanwhile keep the goddamn lights on. The decision maker, the budget, has moved out to the business user. Moved out to the business user in a context where the actual individual users who are not currently looking at their devices will be looking at their devices and will expect the same degree of user friendliness, user experience. But if the business user is the customer and the customers’ people have expectations of having the same kind of experience that they have with Facebook, Google, Apple etc, for the vendor of the software it becomes a really challenging problem. You cannot sell technology to business users. And if the business users are not intermediated by an I.T. department whose responsibility is to transform technology into working solutions for mere mortal civilians in the organization, then the vendor has to do that herself. The vendor has to provide the customizing professional services, the downstream translation of general purpose technology into specific focussed business application software, software solution. Think about the enormous growth in the capabilities of predictive analytics as companies learn how not only that every interaction with their customers, with their vendors is generating data, but they’re capturing the data and now they have the opportunity to mine that data and turn it into business advantage. Increased profitability, reduced cost, reduced working capital. Doing that requires a degree of customization to the specific environment – a supermarket is not like a bank, even though the functionality may be exactly the same when you’re talking about managing customers, input into what I’m offering today or this minute or tomorrow. It has to be translated, it changes the business model for the vendor to try to become an integrated solutions provider versus a purveyor of general purpose horizontal technology.
At the same time as I say the cloud becomes the platform and preempts market space, and the technology coming off the cloud that is putting your technology on the cloud is not going to solve the problem for the customer that the customer needs to have solved. And the consumerization of IT as I said means that the business buyer is in effect the inverse of the IT department as a customer. More of the low margin professional services that previously could be farmed out to the customer’s IT department or to third party systems integrators now has to be performed by the vendor. When you think about it from a selling perspective, part of the problem that the vendor has is that whereas once upon a time there was one door to knock on, now there are far too many doors to knock on. And the economies of scale that came from having a general purpose customizable solution when third parties did the customization has been translated into radically reduced accessible market, the total addressable market for each customized solution from the vendor is obviously much smaller than the general purpose market for data management software ala Veritas or app server transactional application platform such as BEA. There’s another generalist factor here which distinctively has an impact on the enterprise software industry. The U.S. IPO market has never come back from the post 2000 collapse. Back in the day back from 1980/81 through 2000, during those 20 years we typically had an average of 30 venture backed IPOs per quarter. Per quarter. And the bulk of those were software companies, a few biotechs periodically. Since then the average has been on the order of 10, sometimes in back in and around just before the global financial crisis it got inflated by a wave of ‘Born in China’ mostly-scams that went public in the US and then collapsed, so the numbers were artificially inflated. Since the financial crisis we’re back to around 10 to 15 a quarter. More than half of venture backed IPOs in the US have been for biotech and life sciences companies. Very very few for enterprise software companies. At the very moment when of course the need is for more and more cheaper capital to be able to climb that mountain to sustainable positive cash flow. But you will say what about the unicorn bubble. Well the unicorn bubble is a real phenomenon. It does involve, against all of financial history and the common sense that will sooner or later reassert itself, it involves professional investors who are accustomed to buying liquid tradable stocks in the public market paying premium valuations for private companies whose shares they cannot sell in order to deal with their fear of missing out on the next Google, Facebook, Amazon, what have you. But as we cut down from it, of the 237 unicorns counted by CBA as of about two weeks ago, only 37 of those 237 could be possibly included as enterprise software slash service vendors. Only 5 of those were valued above $2bn versus some 90 consumer oriented ones and I think the reason is very simple the consumer oriented ones can pitch that they’re addressing markets that are numbered in the hundreds of millions if not billions of potential users and no enterprise software company can. Slack is not going to have a billion users, no matter what. So the unicorn bubble has been substantially, not entirely, but substantially irrelevant for the enterprise software entrepreneur and venture backer. So, I’m going to close with two slides – one from the perspective of the entrepreneur, and the other from the perspective, somewhat more natural to me, of the venture capitalist. From the entrepreneur’s point of view – OK the technology risk i.e. when I plug it in will it light up, when I run the code will it freeze the machine or actually execute the way it’s supposed to – that risk is much less, much less in terms of money and time than it used to be. But the market risk given the changes within the enterprise customer the shift from IT as the buyer to a plethora of distinct business users, uninformed and ignorant of technology as buyers and with that market space being progressively taken over more and more by the cloud providers the market risk is radically greater. Who will pay to buy it if it does work? The financing risk, well the financing risk, will the capital be there to fund the venture to positive cash flow… Financial risk has grown with the lack of customer financing even while some spillover from the unicorn bubble may transiently seem to reduce it. And the business building risk hasn’t changed at all. So from the point of view of the venture investor and perhaps let’s say the seasoned entrepreneur who’s been through the mill at least once before, the question becomes at each of these stages at each of these decision points is “Do we sell now?”. It does light up, the technology works. Do we hold the auction now? Can we possibly add our functionality get to of the cloud companies to compete to incorporate our functionality in their cloud platform? We have three credible customers and they’ll testify, they will go on record as saying that they paid for it, it works they’re going to buy more. Do we sell now? We have access to another round of capital. It’s going to be dilutive. Because we’re not a unicorn. We’re just a humble working company adding critical functionality to the underlying infrastructure that enables the digital world to work. We can get some more capital. But we’ve then got the business risk of trying to go all the way and build out the business and build a sales force and build customer support. And yes build a professional services organization that can customize our technology into relevant solutions. Do we sell now? And that’s why as you look across the board, success for venture investors in this world is now defined not by a successful IPO and the gradual liquidation by distribution to limited partners, or sale in the in an active trading market. Success is now a timely trade sale. And it’s not just British companies being bought up by American companies, it’s American companies being bought up by whoever will write a cheque. Now the Chinese are writing a lot of cheques and the US government is getting a little nervous about that. But this is the world that has changed. It’s not that creativity has disappeared it’s not that the needs for incremental – or even radical – improvement in the infrastructure and the applications that exploit that infrastructure has changed. It’s that the financing model, the revenue model, the market model, the market environment, these have changed radically. Enterprise software, as a couple of really really smart young academics have written, in this environment from the old venture model – and I don’t know if there’s anybody else here in this room who’s a functioning working venture capitalist – but the old model one reason one reason we limited in each portfolio the number of investments we made was because every investment was a call option on an infinite amount of time and energy and sweat and pain as my very distinguished partner Henry Purcell who in his days as a working physicist at the RCA Sarnoff labs had the patents which made the semiconductor lasers that were reliable enough to hook up to the fiber coming out of Corning and enable fiber optic communications. Henry used to say as a physicist there’s only one reliable law, statistical law of venture capital and that is grief is a constant. So we limited the quantum of grief for any moment in time by limiting the number of investments. In this world today you can’t play that way. Because the only way to get access in this world where there are so many wannabe startups at such low cost is to shift towards the pray and spray model originally adapted for consumer oriented new software startups but now relevant to enterprise software startups, where little bit of money in far more companies than you can pay attention to – you give them a cheque, you buy access to updated reports on progress – and if and when some subset, usually pretty small subset of those startups generate traction as they say then having negotiated pre-emptive rights on further financing, you have the chance to decide yes we’re gonna make a real commitment and we’re going to give them a call on our time and energy as well as a gift of some cash. So with that, that is my message. I’m the last word between you and lunch. I’m happy to take any questions that anyone may have.
Audience Member: Thank you very much Bill. I guess a couple questions from observing the market at the moment where we’re seeing an increased expectation and as a result valuations in the market. So I guess question number one is do you think there will be a correction anytime soon. Question number two and I think this is something I’ve been reflecting on a very keen to hear your thoughts on – as a company that’s going through the different stages of growth and at the moment as vote can be very tempted by very very high and lofty valuations. That does set a certain expectation in the market. So what advice do you take something at a lower valuation but with a more realistic expectation will continue growing that hype cycle.
Bill Janeway: Well writing that speaking as a as a company that’s riding a hype you have a company and investors who are riding the hype cycle. But I think the principle issue about whether or not whether when there will be a correction relates to two different factors. The first to remember since 2008 we, all of us, and particularly institutional investors, have been living in a financial environment that actually has never before been experienced. In the US it’s ending it’s pretty much ended. But around the world still for the first time ever we’ve had a decade when the risk free real rate of interest has been zero to negative. Now if you want a force for pushing investors into higher risk in pursuit of higher return investment that’s it. Now what it means is – as a generic that is going to touch everyone not just the specific, the first the most exposed companies – there’s a double exposure here. When it begun in the U.S. it almost certainly is going to spread around the world. There is a return towards more normal you know 2/3% real risk free rates of interest with credit spreads on top of that taking available returns for investment grade debt back up to the 6/7% level and jumped back up into double digits. You’ve got to expect an adjustment in the valuations paid for the possibility of hyper growth/super growth/super return by institutional investors but they’re also along the way. There’s the micro exposure the short that’s the macro exposure the micro exposure represented today by Tesla. Tesla isn’t software. Well yes it is a software company. Tesla is. Oh yes it is a hyper growth company. Just looking at the at the remembers the New York Times or The Financial Times this morning we have heard a lot about Porche and BMW. I think it was BMW not poor. BMW gets something like in revenue terms twenty thousand dollars in revenue per car. I mean net revenue. Tesla gets four hundred. I mean the valuation of a BMW is twenty thousand dollars per car it sells per year. For Tesla that’s four hundred and fifty thousand. The notion that well you know that M3 they the model 3 that we told you about that we were going to sell for thirty five thousand bucks. Well you know it’s actually going to be seventy five thousand. It’ll be a little faster that’s still good for you. The guy you guys who gave me the hundred thousand dollar deposit refundable deposit?… The bloom coming off the Tesla Rose has implications for all entrepreneurs out there who have been promising hyper growth fantastic returns, just stay with me for another year or two or three or five or seven. So discount rates on the future go up present value of that future go down when it happens. I mean compared with 2008 this won’t even be a blip in the data. There’s no credit really involved. I mean the credit is those deposits at Tesla. But this is not like the leveraging of the financial system to the ludicrous levels of 2007. It’s not even at the scale. I mean the dot com Internet bubble absorbed six trillion dollars of capital. But all of that capital was in the equity market and was in the junk bond market where leverage was maybe 1:1 max 2:1. In the credit bubble it was 35 to 50 to 100 to 1. So when the dot com Internet bubble burst we had a recession but it was self well within normal dimensions. This won’t even be a recession. The only thing that will happen is some people have to go back to real work for real companies. Some fund managers who got carried away will lose their jobs. But it’s not going to disrupt the continuity of economic life. We can leave it to our political masters to pursue the disruption of our economic life on both sides of the Atlantic.
Audience Member: So you talked about the cost of building a enterprise software company has gone up and opened up. But there’s a glaring counter example which which is VIVA systems. VIVA raised $6m and 20 million free went public. Seven years later six billion dollars and now is six hundred and forty five million of recurring revenue with a billion dollar market cap.
Bill Janeway: Fabulous achievement, fabulous achievement. And every generation there are companies and even without the enormous benefit of that old perpetual license revenue model. There are examples I know two or three of companies that have bootstrapped their way with extraordinary financial discipline to grow through a manner that.. when I say positive cash flow is the promised land. I learned this from my psychotic mentor 30 40 years ago. Fred Adler a remarkable venture investor the point about positive cash flow is it means as with your example that your customers are giving you more money more cash on a continuing basis than it costs to deliver value to them. How cool is that. Well there are companies and there will be. The question I’m posing here is how generalizable is that in 30 years? 25 years ago it was the model you never thought that you would get public without already generating reportable profits and paying your bills because your customers valued what you were giving them more than it cost to deliver to them. So you’re right there are examples and there will be examples out of the unicorn financing. There are gonna be some number of sustainable. Players that that win and generate great value won’t be two hundred and thirty seven of the thirty seven enterprise software companies won’t be thirty seven might be three four five seven. But when you see a company that has the productivity in the use of capital that generates that kind of value and growth. I mean those are those are those are marvelous. The flip side is I always like to talk about I give a whole separate lecture on how to think about bubbles and why how bubbles financial bubbles not only can piss away an ungodly amount of capital investing in stuff with no underlying fundamental value you know beach houses in the Nevada desert, Bitcoins. The point about bubbles is that they for this kind of venture they solve a coordination failure in time when you’re starting a company that financing risk question is there going to be enough money behind this first round to get me all the way? With a bubble you know to worry about that bubbles can be productive. But we don’t have that many history examples of that. We have railways, electrification, the Internet, right now we will see from the unicorn bubble much smaller in scale much more narrow and focus than the Internet bubble. Some survivors which will play a role in continuing the build out of the digital economy which I think is only half done.
Audience Member: Hi. I wondered if you could be a bit more explicit about how margin is declined with the enterprise software system has changed and in particular what the benchmark for achievable EBITDA is today for a modern enterprise company.
Bill Janeway: Well I think there’s the question is when does the tipping point come. That is when you’re building up a revenue stream on a SaaS revenue model, there comes a point in scale where the cumulative generation of revenue in cash becomes enormously profitable. Obviously Salesforce reached that some years ago. It’s the build up to that point. I mean I can think of several companies, successful businesses that Warburg Pincus invested in where it was. No order. It’s not so much that as I say it’s not the amount of the margin it’s the time and the integral of the capital required to get to when you cross over into positive cash flow. I think it’s pretty generally the case now when we’re talking about Enterprise SaaS revenue model businesses that it’s more likely that it will take a hundred and fifty two hundred million of cash over four or five six years with good execution to take advantage, to build out the salesforce and customer support that enterprise customers will require as well as to customize to the needs of those enterprise customers. The technology the general purpose technology that you develop so that it becomes a useful application.
Audience Member: In your final paragraph you’re saying that VCs are shifting to perhaps the time to sell as opposed to an IPO. Does this affect the kind of perceived wisdom of VC expectations of returns? So this is 10X investment and also the time scale. So you know in five to seven year period where you’d be looking at you know seven to 10 times return.
Bill Janeway: There’s data on this. Now the time to IPO such a IPO is as there are has risen very substantially so that the time to IPO is now up around 8 years which represents the time it takes to generate that revenue base that can have some promise of a cash flow positive cash flow. Back in the back in the Internet bubble it was it was under three years normal such as it was before the bubble was around four to five years so eight years a long time and it does it really kills the IRR. Whereas the time to trade sale according to the NBC data in the US has stayed rock solid and about four years. When you’ve got some proof points such as I’ve indicated that you can go to the market with. Now the other factor and this is a problem you know if you like not some not for the entrepreneur for the ALP for the venture capitalists there really is a big problem and that is that limited partners have gotten completely overwhelmingly focused on IRR not on cash on cash return. We used to say Warburg Pincus was one of the last holdouts we used to say we used to tell our limited partners you can’t eat IRR. If you’re a pension fund and you’ve got obligations, telling your clients that you had a great IRR but you know unfortunately it was made in 11 months and it was one point one times the capital you invested isn’t going to meet the obligations when the costs of the pension comes to bear. Whereas cash on cash is what pays the bills. But the the fact that whether it’s an event in venture growth equity or let alone the buyout business the LPs are so exclusively focused on IRR pushes everyone in the business to have a much shorter term time horizon because the one thing that will kill IRR without fail is length of time when you’re compounding at what’s supposed to be the venture IRR of 20 percent plus net of fees and expenses.
Audience Member: So what you describe is basically saying you know the days of unicorn general purpose enterprise software business is gone. So basically for which is replaced by smaller niche-y use case driven enterprise software plays whose valuation will be smaller. So you’re describing how you’re going from a V.C. attitude which is trying to spot the unicorn and go all in, basically a binary bet, to what is now what you describe as “Spray and pray” approach which is basically you know portfolio manage and spread across a whole range of things and hope for the best that something comes out. But is there not another way that says there should be a return of almost a craftsman investor which is basically to say you find yourself somebody who understands the use case in a given subsegment of an industry very very very well, has a vague notion of what technology is possible and can articulate and then story-fy out and then you just slot a team under that they’ll say well I’ll take a bit of Amazon, Google just assemble it and here you go. It’s basically almost an operational engine under a product manager use case understand or would a bit of capital but actually the focus becomes much more pinning that team under it actually.
Bill Janeway: You know this I’m going to take that and meld it a little more. We had as I indicated a really ridiculous amount of success in participating in delivering that general purpose infrastructure technology to the marketplace in the 1990s. And you know some I don’t know 250 million of invested capital came back is 10 billion dollars of return to our limited partners. It was really worthwhile. But the whole strategy the whole investment strategy in tech and Warburg Pincus has changed radically. What we’re now investing in our businesses which have much smaller much smaller addressable market that in fact represent just what you’re talking about. Many of these are companies that have over a period of five 10 years a dozen years reach funded themselves by revenues from customers sometimes beginning on a proprietary license basis. But as private able with some financial support to go through the transition to SaaS. But what they’re doing are things like logistics Management for less than truckload trucks in regions in the United States. Facilities management for schools hey that can branch out and become post offices as well facilities management for public sector physical assets. Oh here’s one which was really successful. Turned out it was sort of consumer oriented but the focus was on how do you find a care home that you can trust. I’m not kidding you. The company was called A Place For Mom. Hugely successful but took years to build up bringing in the kind of not inventing the world. These companies are sitting two generations generations behind the frontier. Of course they’re using open source where they can you know incorporating functionality from Google or Microsoft probably more Microsoft for these sort of these SMB oriented markets. They’re not trying to sell G.E. or JP Morgan they typically are going to not become billion dollar businesses. But at 100 million in revenue U.S. they can make real money and they can dominate a space have pricing power be below the threshold if we did have an antitrust division in the US which essentially has been put on hold by it for the last generation but they’d be even below the threshold for the European Commission if they were operating in Europe. And so we’ve really shifted towards exactly what you’re talking about people backing people who really understand a well-defined but limited because it’s well-defined market can bring good enough technology to solve customers problems and to let enable customers to run their business better. I will tell you I have many here people know what a sharp ratio is. Rate of Return divided by variance. Now in this portfolio of the Sharpe ratio in this portfolio is practically infinite. We’ve had no losses. They never generate you know the kind of I don’t know BEA was one hundred and fifty percent IRR over six years seven years. They’re never going to generate anything that they generate 25 30 percent IRR over four or five years that we’re invested. And the variance of that is almost zero there. It’s a fabulous portfolio rock solid base exactly what you’re talking about.
Audience Member: So just one last question – not a question but an observation. I guess then your entrepreneurial environments will shift very much from the gung ho all or nothing, guns blazing American model to something much more akin to the German. Well they don’t have the German entrepreneurial model where you find these five generation companies that made radio antennas.
Bill Janeway: You know I’d love to see the ??? of software companies in Germany. I still have this problem. I cannot count the number of companies started by refugee slash escapee slash alumni of Oracle. It’s countless. The problem is I can count the number of software companies started by alumni of SAP. You know and I’m not going to use two hands. So we need that kind of entrepreneurial culture. But all those companies I’m talking about are American companies and no they’re not fifth generation. And usually there is a real problem because there’s a founder who really understood a set of customers who’s built a business up on a very conservative basis. That’s why they’ve been able to get to 25 30 40 million may have taken them 10 years to get there. When you think about what’s involved in accelerating growth professionalizing the sales force looking at adjacent markets possibly even doing acquisitions the usually the biggest challenge is to helping the entrepreneur founder understand that she in some cases one critical case at was a very reluctant she kind of need a lot of help like maybe even replacing themselves because you come in as a partner not as Carl Icahn and you want to maintain that relationship while you do shift the value proposition as an investor from what was very slow but very steady into somewhat faster and therefore more valuable growth.
Learn how great companies are run
At BoS we run events and publish content that is highly valued by anyone trying to build, run, and scale a great software company.
Sign up for a regular dose of actionable and useful content: