On historical returns and venture flavor.

By Fred Destin, contributor

FORTUNE -- The big story of the past few years in Venture Fund Land has been that large institutional investors concentrate money with fewer managers and flagship brands and/or find emerging managers.

Since VC offerings have generally been undifferentiated, LP's had little choice but to overweight historical returns in their decision making algorithms.

Now the market is in radical reduction mode (speaking of the number of active venture firms out there) and every venture firm has to answer a simple existential question : Why do you even exist? Is there a valid reason to continue beyond the funneling of money from A to B and the siphoning of management fees? What do you stand for? Why does the market need you?

So the LP's have done two things:

1. Concentrate more money with fewer managers
2. Radically concentrate their bets into differentiated offerings

You can group these offerings into four broad categories:

• Flagship brands with both exceptional entrepreneur mindshare able to compete at whatever dollar amount required for halo deals (Accel, Sequoia, A16Z, KPCB, Greylock, Spark Capital, etc.)
• Highly focused funds with dominant presence in their chosen field or area of focus but (broadly) choosing to remain small (USV, Foundry, Benchmark, First Round)
• Turnaround stories / successful generational transition stories of strong (or formerly strong) brands being reinvigorated by hungry / rejuvenated GP teams and showing signs of out-performance (CRV, Mayfield, Atlas Venture) -- usually espousing the small / focused mantra described above.
• Emerging managers and super angel funds whose size now allows institutional investments, hyper-focused versions of the above (IA Ventures, True Ventures, Felicis, Softtech etc.)

Again, no matter which strategy you choose, your venture firm has to be able to compellingly answer that simple existential question: Why do you even exist?

When any VC firm pitches their story, they're always presenting their strategy as the one that works. I am slightly amused by the constant back and forth between small and large funds, believers in seed and non-believers in seed, folks who pretend you should only play seed/acceleration rounds and not in the middle, or whatever flavor you think is going to work best in this cycle of fundraising.

Take NEA's "Big is Beautiful" argument that bigger funds are bigger simply because they are, well, better. Compare and contrast with Bill Gurley (Benchmark Capital) and his comments to LP's ("You are Are Blowing It"): "You're in the field. And when you allocate [your capital] obsessively to firms in the top quartile, it will have an impact on how things play out. [...] If you ask if people will take an excessive amount of money, the answer is yes."

The reality is that no amount of historical market data analysis is going to tell you which fund to invest in next, especially given the speed of market disruption. When the next guy comes in telling you he's found a bulletproof may to make money on his fund, send him this:

Source: Levi Shapiro interview with Felda Hardymon

See, for the LP, none of these odds are good enough to derive a systematic strategy. And for the GP, many strategies will likely work. I'd happily put money in, say, 500Startups (an experiment to be sure, but an exciting one), First Round Capital, Atlas Venture (wink), Spark Capital and A16Z, thereby covering the entire curve style (admittedly with a bias to new managers).

I am not a believer in big funds because I would not know how to run one effectively. We've found capital constraints to be a great friend. So I am perfectly happy with a small-ish early fund where all partners are aligned through equal economics and a desire to make it as a team. I'm in the Bill Gurley camp, and Benchmark has always been an inspiration. But I won't be the one telling you the NEA strategy won't work. It might.

It's just not my strategy.

Fred Destin (@fdestin) is a Boston-based partner with Atlas Venture. This post originally appeared on his blog.