I have received many emails asking about the logic behind my investment decisions--both good and bad. So this is the first in a series of posts addressing investment strategy.
Many VC's talk about Road Map investing, but I think that we at Bessemer try to take this more seriously than most. This stems in part from our unique structure--with only one investor, we do not engage in the tri-annual exercise of raising new funds, which may lock firms into a strategy, and burn time that can otherwise be spent on strategic planning.
I think I developed the first formal road map at Bessemer. Back in 1992, fresh out of business school, I joined Bessemer and proceeded to fall in love with every crappy pitch I heard (I recall that one of them manufactured conference expo booths). Fortunately, before I did any damage, my bosses intervened, suggesting that perhaps I should take a few months to Think Before I Fund.
So I developed a comprehensive list of 38 potential investment sectors of high technology, and I spent the next 3 months whittling it down to 5. I crossed off sectors which required deep domain knowledge I don't have (semiconductor capital equipment), sectors that were too early (wireless LANs -- pre-Wifi), sectors that were too crowded (object oriented databases and middleware), and too unproven (multimedia CD-ROM titles). I solicited advice from the smartest experts I could find -- folks like Al Lill at Gartner Group, Rick Sherlund at Goldman Sachs, Brad Feld at Feld Technologies (now at Mobius), Eric Schmidt at Sun (now Google), and Per Suneby who ran Motorola Codex and is now at Flagship Ventures (Per: "Let me give you some advice: never invest in a security company"). I went to conferences, surveying telco equipment buyers and MIS Directors (there weren't too many "CIO's" back then).
The result was a decision to focus on Data Communications (indeed, that was a narrow specialty in 1992), and a Powerpoint presentation targeting five specific "initiatives", each of which led me to proactively find very specific investments that I made in the following 3 years:
1. Network Management (led to NetSys--acq. by Cisco)
2. Network Security (Verisign)
3. Asynchronous Transfer Mode (Cell Access--acq. by Fore, Data Labs--acq. by Yurie)
4. Wavelength Division Multiplexing (Ciena)
5. Enterprise E-mail (Worldtalk, ON Technology, Tumbleweed). Actually, PSI-Net also emerged from the Email initiative, since PSI-Net's commercial IP network was accessed primarily for SMTP (there was no web in 1992).
This road map enabled me to focus my time very specifically on investment opportunities that matched my plan. I think that entrepreneurs outside my road map appreciated the quick No, and entrepreneurs on my road map appreciated the in-depth knowledge I brought to their businesses. (At least no one had to explain to me what IP was--quite a credential for a VC!)
Other Bessemer investors noticed how much more pleasant my life was with a road map, and so they adopted and enhanced the methodology themselves, with great results.
Obviously, my road map in 1992 was way too broad to be effective today, when the volume of startups and investors demand a much higher degree of specialization. Along the way I stopped every few years to refresh the map, redirecting myself toward internet services (Keynote, Flycast, Register), then later to consumer e-commerce (eToys, BabyCenter, Blue Nile, Hotjobs). In 1999 I composed an ill-fated road map around B2B exchanges, and in this decade I have focused primarily on information security.
Each Bessemer investor's road map begins with an analysis of disruptive catalysts that have the potential to cause major displacements in our economy. Those disruptive catalysts might be technical (e.g. network vulnerabilities), demographic (e.g. aging US population), regulatory (e.g. spectrum auctions or SarbOx), psychographic (e.g. consumer concerns about security), or geopolitical (e.g. China's reception to foreign investment). The road map then lays out specific strategies, or "initiatives", to exploit the disruption.
The information security road map included many initiatives over the years. With help from our operating partners--Peter Watkins (ex Pres. McAfee), Chini Krishnan (founder Valicert) and Devesh Garg (GM, Broadcom's security business unit)--my partners (Justin Label, Jeremy Levine, Rob Stavis) and I decided to seek out companies who would address policy compliance, IPS (without false positives), spam, ID theft, spyware. For each of these intiatives we made one investment in the best team we could find attacking the problem--some were follow-on rounds (Postini, Cyota) and some were new teams that we incubated in our offices (Elemental, Determina, Infinitrust).
Perhaps the most difficult step of road map investing is knowing when to burn the map. Some of the best and worst decisions we have made over the years centered around this question of when enough was enough. We successfully exited biotech in 1993, big box retail in 1995, and etailing in 1998 before those sectors busted, but on the flip side we failed to exit telecom in 1999.
Even with a great road map, it's always necessary to maintain an open mind to great opportunistic investments (our anti-portfolio routinely reminds me of this). Nevertheless, I recommend that investors try the road map approach. And if you're an entrepreneur, I recommend that you favor working with investors who have very precisely targeted your space for investment--those investors will take less time to sell, cause less damage (maybe even help), and, if they don't invest in your deal, will likely make a competitive investment.
As for me, I believe that data security remains a robust market, but at this point Bessemer is over-allocated to the sector. So last year I crafted a new road map, but that's another post.