Editor Profile - Denny K Miu was the founder and former CEO of two startups, Gigamon Systems and Integrated Micromachines (now Touchdown Technologies). Denny has extensive experience in developing technology, products and business relationships. He has been a Professor, an engineer, an entrepreneur, a team leader as well as an individual contributor.
Denny has recently published his second book entitled "Survival Guide for 'Slow Start' Entrepreneurs". Please visit here for more details.
--
Two years after we launched our product in May 2005, my previous startup was able to acquire our 150th customer. Our sales continued to grow organically, from new customers as well as from existing ones who kept coming back for more. In addition, we announced that we had achieved profitability for six consecutive quarters.
By all measures, Gigamon was a spectacular success which was all that more remarkable consider that we had taken no VC money. Like so many entrepreneurs before us, the six co-Founders of Gigamon built their company the old fashion way. We bootstrapped the product development effort by dipping into our own pockets (or more accurately, our wives’ pockets) and by not taking salaries for two years; and after the product was ready, we secured minimal but adequate amount of working capital from an angel investor and not take salaries for another year.
Having spent the last decade and more as a struggling entrepreneur, I understand all too well the twin irony of accidental genius and reluctant warriors; nothing out-performs good execution like good fortune. In all honesty, the fact that Gigamon had taken no VC money was not part of my original financing plan. Between the summer of 2003 and the fall of 2004, I must have visited nearly 50 VC’s, all of whom had turned me down. So my company succeeded in spite of my initial failing.
Keep in mind that I was not unfamiliar with raising VC money (I raised $65 million for my previous startup) and I was mentally prepared that 2003 was probably the worst time to raise money (having just endured the rolling catastrophes of Asian financial crisis, Y2K, dotcom bomb, telecom burst, stock market crash, 9/11 attack, war in Afghanistan, SARS, launch of the Iraq invasion, outbreak of avian flu, etc.). But the uniformity and the abruptness of how I was turned down by every single potential investor I spoke with still startled me.
Obviously I respect the VC’s decision and I believe they have made the right one. I was not spiteful nor do I have any need for vindication. I believe VCs have an important place in our financial eco-system (as a financial instrument, their successes ultimately contribute to our 401K’s) and I respect their profession. In fact, if deem necessary and appropriate, I would have no problem taking money from VC for my future startups.
It is just that in general, I find it much more educational and rewarding if I refrain myself from demonizing others and instead focus my energy on learning from my own failures as well as from my successes (accidental or otherwise).
So in that spirit, over the last four years, I had spend much awaken hours soul searching and I have summarized what I learned in the following which were the top three reasons why we were turned down by the VC’s and why we should have failed as predicted had it not been for the good timing of the ever changing market and the enduring quality of its team (intelligence, integrity and inventiveness).
Top three reasons why VCs were convinced that we would have failed:
1) We did not have a rock star CEO
2) We were all engineers
3) We contradicted Gartner
1) We did not have a rock star CEO
Experience really matters in a startup.
This explains why VC’s prefer to bank on an experienced CEO in a startup, particularly if the Founder/CEO has a successful track record and is touted a “Serial Entrepreneur”. The reasoning is simple. Although past success does not guarantee future performance, it does make it easier for the VC’s to quantify the probability of success. Since the VC’s business is to reduce unknowns and the unknowables, avoiding inexperienced and ineffective CEO’s is a no brainer.
However, as it turns out, experience is important but judgment is too, and everyone can be a victim of his/her experience (which ironically could often cloud their judgment). Much of a startup’s success and much of the difference that a CEO can bring to a startup have to do with his/her ability to make the right decision (at the right time). But there are two kinds of decisions, one that is life-and-death and one that is not.
Life-and-death decision is very easy to make. And this is where experience comes in. It is really just real-time signal processing and pattern recognition. Anyone who is capable of trusting his or her own instinct tends to make the right life-and-death decision. But making the right life-and-death decision is like avoiding a car accident that never happens. Your only material reward is that you get to live.
Decisions that are not life-and-death are much more difficult to make in a startup. They are actually much more important to the success of a startup. So the question is how does a startup CEO go about making the right non-life-and-death decision. It turns out that there is no such thing as the “right” decision. In order to make the “perfect” decision, one has to have “perfect” data. In a startup, having perfect data (or even the right amount of data) is a luxury that we could not afford.
Typically, in a startup, there are two ways to compensate for the lack of perfect data. One is to wait for more data and one is to seek consensus. In my opinion, both kill startups. CEO who procrastinates hoping to get incremental data is as deadly as a CEO who jumps to the wrong conclusion before enough data is in. Similarly, a CEO who is afraid to make tough and unpopular decision and hides behind the shield of consensus will almost always lead a startup to the proverbial cliff.
It turns out that in a startup, it is not about making the right decision but about making the decision right. In other words, it is about the CEO truly accepting the awesome responsibility of being the ultimate “Decider” while refusing the temptation of being a dictator (avoiding consensus is not the same as avoiding consultation).
In other words, a CEO must have conviction or there is no way to lead a company. But a CEO must know the difference between “temporary” conviction and “permanent” conviction. Lack of temporary conviction will kill a startup but so would permanent conviction (as Stephen Colbert once said, “Staying-the-course means doing the same thing on Wednesday as you did on Monday no matter what happens on Tuesday.”)
So being a successful CEO in a startup requires experience but it also requires good judgment.
And good judgment has to do with when and how to build up creditability with the team and the shareholders, and when and how to cash in your political “earned” capital to mobilize the company behind an unpopular decision that you have made based on imperfect data. More importantly, good judgment has to do with maintaining a positive feedback loop to constantly “re-up” your credibility with your constituents so that you do not inadvertently over-extend your “reserve” because you can be sure that you will need them again (and again).
Unfortunately for Gigamon, I was not a “Serial Entrepreneur”, but fortunately for them, I was a “Serial First-Timer”.
While I did not have a successful track record since I had not made any money for my former VCs (in my last startup), I was stubborn, I was hungry and I was experienced.
In summary, experience is necessary for a startup but not sufficient. On the other hand, paradoxically, as I have learned the hard way, as a CEO for a startup, having a strong will to succeed and the tenacity to follow through on difficult decision is sufficient but not enough. You need a little of both (and a whole lot of luck).
2) We were all engineers
A startup has the same set of constituents (employees, customers, vendors, etc.) as any other companies and what ties these disparate pieces into a puzzle is a marketable and revenue-generating product (or in case of startups, the pursue of a marketable and revenue-generating product).
But interestingly from the VC’s perspective, the startup itself is their product. They are interested in the company’s product (as a means to an end) but ultimately they are only interested in the company **as** a marketable product. But unlike our customers who consume our product, VC is like a channel partner since they do not consume but rather they resell.
VC purchased a sizable piece of a startup at a very low price with the pure and simple intent of reselling it at a very high price (sooner rather than later), hence the importance of an “exit strategy”.
So it is reasonable that when a VC makes an investment, they look for an A-Team consists of Founders who can help build and sell a product but also those who can help build and sell a company. So it bothered many of them that the six co-Founders of Gigamon were just a bunch of engineers (or former engineers who knew only to sell products but not companies).
Fortunately for us, there has been very little market demand for the VC’s product (i.e., no liquidity and no exit for startups). Therefore, had we built our company based on the assumption that we would have a quick flip, we would have failed miserably because we would have either run out of fuel or run out of runways.
So we did the right thing (and the only thing that we knew how to do) which was to act as our own “surrogate” customer and built a “Swiss Army Knife” data access switch for “out-of-band” network monitoring that we wish we had in our previous career. It turns out that being our own “surrogate” customer was the most important ingredient for our success.
My fellow co-founders of Gigamon had worked in the network monitoring side of the networking business for many years. They knew about their customers' pain first hand because as a monitoring tool vendor, they knew how hard it was to deploy their own products. So in a sense, they were frustrated that their ecosystem was broken and the Gigamon switch was their dream tool.
Had they not been long-time veterans of their own industry (which they help built in the past 20 years), we couldn’t have gotten that important insight by just talking to our customers. Customers know what to object and what to improve when you present them with an imperfect product but not when you present them with a perfect PowerPoint. Henry Ford had said that if he had listened to his customers, he would have built a faster horse because that was what his customers wanted.
It is sort of like Steve Jobs and his iPod (and iPhone). It is hard to imagine how things could have become what they are had Steve not been a music fanatic (not just a fanatic). Steve didn't rely solely on his customers neither to define his dream toy. Instead he listened to his heart because he believes he was the perfect surrogate customer.
And that was what we did. We listened to our hearts (even though we were just a bunch of engineers).
3) We contradicted Gartner
Fund-raising for a startup is an art form and a process. From the VC’s standpoint, the process is about establishment of credibility and mitigation of risk.
An experienced CEO would know how to “do the dance” and would engage potential VC’s in a proper spatial and temporal sequence (i.e., first meeting to get to know each other and to do a soft sell, second or third meeting to meet the team and lay down some measurable markers, a few more meetings to update on progress punctuated with right amount of dissonance to be sure that everyone knows that you are shopping but still keep everyone within arms length so that no one would accuse you of being promiscuous, etc.).
The last thing a CEO needs is an unknown introduced by a third party. In 2003, we had such a bombshell.
Gartner released a report stating that IDS (intrusion detection systems) was an obsolete technology and would soon be replaced by IPS (intrusion prevention systems).
An IDS is an “out-of-band” passive appliance that sits on a SPAN port or a TAP and “listens” to activities in a network (it is equivalent to a surveillance camera).
An IPS is an “in-line” active network element that can shut down illegal activities much like a firewall (equivalent to a security guard with gun and real bullets).
The point was not which was better, or whether or not Gartner was right (it turns out that IDS/IPS are apples-and-oranges and Gartner was wrong, or at least misguided). The point was that we had a reputable analyst firm making a bold statement that “out-of-band” solutions would be replaced by “in-line” solutions.
In other words, in the eyes of the VC’s, our proposed product which was a data access switch to facilitate “out-of-band” monitoring was at best a temporary bandaid since they rightfully inferred from Gartner’s assertion that there would be no long-term demand for “non-intrusive” network monitoring and therefore no possibility for the Gigamon product to gain a meaningful and permanent footprint in the emerging enterprise networks.
Now the problem was not just that the CEO was non-performing or the team had limited skills, the problem was that our vision was wrong.
This posed a dilemma for us.
On one hand, we knew in addition to other industry experts, our customers were contradicting Gartner as well. As battle-scarred veterans, we understood that not only was it important to sell solution and not technology, it was important to sell solution to solve customer problems that fit within the constraint of their “current” job description (in other words, the lowest hanging fruits are the ones where your customers can buy your products without fundamentally changing their career).
In summary, IDS and IPS were two different product categories targeting two different sets of customers. The customers that we were talking to were monitoring people who would only buy monitoring products. When an IDS becomes an IPS, it is no longer monitoring and would be purchased by a different team with a different budget (in fact, one could argue that IPS would never be a permanent footprint since its functionality would ultimately be absorbed by the firewalls).
On the other hand, we had to wonder because Gartner is smart people and how could they be wrong.
So we spend the next six months architecting a product that would allow IDS to become IPS by letting an attached passive appliance send back “kill” packets into the network. We conjured up a beautiful concept that we called “asymmetric” switching. When properly implemented, our switch would act as a cross-connect in the forward path to deliver replica traffic to the monitoring tools and when needed, would behalf as a packet switch in the reverse path so that we could deliver the payload back into the originating network element in order to shut down an illegal conversation.
However, the problem we had was that we didn’t have any money. So to survive (and to avoid going back to our wives for more money), we decided that we had no choice but to go to the market with only “half” a product and to introduce the “backflow” feature only after we have made some sales. Had we had money, we would have preferred to stay stealthy and take another eighteen months to fully develop the ultimate killer product.
At the beginning of 2005, as soon as the beta unit of the “crippled” product was ready, we took it to our first customer and it was a gigantic hit. The customer loved it and wanted to buy one right away even though it was preproduction and it was the wrong color (black, not orange). We were excited and since it was such an easy sell, we needed something else to talk about so we started to talk about the next feature release which would turn our product into a “bidirectional” switch.
Then the most unthinkable thing happened. The customer stopped us mid-sentence and told us that he would only buy our product if we could guarantee that we would never implement the “backflow” feature. So we asked him why. And his answer was astonishing. He told us that Sarbanes Oxley has changed everything.
With SOX, network monitoring has become tantamount to financial auditing and an important tool of his corporate arsenal to ensure adequate “Internal Control” as required by Section 404 of newly enacted regulation. Just like an auditor who had to do their job off of a “copy” of the book, not the book itself, our customer had to do their job on a “copy” of the packet and not the packet itself.
So the customers actually “needed” a one-way valve and could not have purchased our product if there was any possibility that we would disturb the production network. In other words, for the same reason that if an auditor were to have their finger prints on the actual book, he would be disqualified as an auditor, our customer was told that by his bosses that if he could touch the network, he was no longer monitoring.
How ironic could it be that not having VC money to build the perfect product as championed by Gartner was the primary reason for our success.
What we lack in vision we made up in peripheral vision.
--Denny--
Click here to read other chapters of Denny's 2nd Book »
Check below for some very insightful comments from readers across the World.
^ 
Also, some interesting blogs by VC's sharing their thoughts on why startups fail ...
Entrepreneur Issue! Why do many startups fail? - by Sid Mohasseb
“Too much Brilliance: a lot of companies fail because the market is not ready for the innovation yet and the entrepreneurs focus on their vision as opposed to what the customer wants or needs – they continue to miss the mark by listening to themselves more than listening to those who pay.”
Why Startups Fail - by Theodore F. di Stefano
“Why do startups fail? The answer is amazingly simple and can be reduced to three main causes: lack of working capital, a poorly conceived business model and a poorly constructed business plan.”
Why Startups Fail - by Steve Tobak
“Most startups fail ... the most common cause is that they develop technology and not products. Lots of people confuse the two terms, but the distinction is critical in startups.”
The 18 Mistakes that Kill Startups - by Paul Graham
“If you look at the origins of successful startups, few were started in imitation of some other startup ... It seems like the best problems to solve are ones that affect you personally. Apple happened because Steve Wozniak wanted a computer, Google because Larry and Sergey couldn't find stuff online, Hotmail because Sabeer Bhatia and Jack Smith couldn't exchange email at work.”
