Lewis and Clark It: The Importance of Portfolio Construction and How to Think About It

This is actually harder than it sounds. Managing a fixed pool of capital with all the ambiguity of a start -up is hard enough, but add in the fact that you have arbitrary capital requirements and follow on funding needs — and it becomes doubly hard. Then add in macroeconomic uncertainty and the fact that capital is scarce in the middle of a fund and often much more available later on due to recycling, and it is darn near impossible to get right. 

First, let’s define terms. Portfolio construction is making a determination of how many companies you can include in your portfolio, with what ownership, how much you should hold in reserve, and where to deploy those reserves.

Avoid These Pitfalls

First time managers often make a handful of common mistakes.

• Not taking enough “high quality” shots on goal. Many people talk about shots on goal, but I think that is too simplistic a view. A high quality shot has a great team and an asymmetric upside to drive big returns, ideally in a sector where you have disproportionate information to make good risk/reward decisions. 
• Overfilling your bucket. As my dad used to say, “You can’t put ten pounds of $#*t  in a five pound bag.” If you allocate all of the fund to initial investments, you won’t have any reserves when you need them to protect the fund from having reserves to enable tough financings in good companies or to double down on winners.
• Spreading peanut butter. Investing small amounts in too many companies can leave a fund with no one position large enough to drive returns. It may also spread the GPs time too thin so that she isn’t close enough to the portfolio to know when to double down. Some people have made this work, but it is awfully hard and needs to be very systematized.
• Need more cowbell. And in this case, cowbell is ownership. I’m showing my bias here but at that moment when a GP has done her work and has conviction, not writing a big enough check can haunt a fund.
• Feeding the strugglers. It is tempting to give struggling companies enough time and capital to turn the corner. It is a trait founders love — and I have been guilty of it myself. But you really don’t want to get to the end of a fund and see the biggest investments in middling companies. Instead, feed the winners. 

Embrace These Best Practices

Here are a few that can help a first time fund manager (even if not a first time investor) get off to a good start:

• Model the portfolio up front (and/or make excel your friend). Pick an appropriate number of companies, target ownership at initial investment and your theory on reserves. Research from Constellation Ventures suggests that diversification helps up to about 20 investments, and its benefits taper off after that.
• Lewis and Clark that baby. Plot your course as you're making investments so you know where you are in the fund at all times. Investments tend to get more scrutiny as capital gets scarcer. Use that to your advantage. 
• Play Warren Buffet for a day. I think of my job as a partner to founders in building companies of consequence. But it is helpful in managing reserves to think of the job of allocating capital to the places where it can generate the most value. If you’ve modeled reserves upfront, you have a sense of what that pool is before your final investment has been made. 
• Lather, rinse, repeat. Review your portfolio model at least once every six months so you know when you need to raise capital again. LPs don't like surprises and this is a place where portfolio construction can help you avoid them. In doing the review, I find it helpful both to allocate reserves to specific portfolio companies to recognize constraints and to treat the reserves as a single pool to avoid following the path of least resistance. It also helps to see if the planned portfolio construction is working and make adaptations if the firm or the opportunity set has evolved.
• Navigate the narrows. This is the pro tip that I wish someone had told me when I started. In the initial model you might allocate 85-90% of the capital to account for fees. Even that assumes a little bit of luck to generate returns along the way to pay fees in the later years. Reserves can get very tight around year four especially if there hasn’t been any liquidity. Treat reserves as really precious during that period. But if you make it through the channel and start generating liquidity, most Limited Partner Agreements (LPAs) allow a fund to invest at least 105% of committed capital through recycling early returns, and some go as high as 120%. When the fund starts to generate liquidity, reserves can be really plentiful - and it can be very helpful to quickly deploy some capital in the fund’s best companies to make sure it can get up to the 100% deployed which GPs and LPs both want.

Funds (and firms) are not made by good portfolio construction — but they can be lost through bad construction. Being sloppy or letting portfolio construction happen to you instead of planning it out yourself can lead to having no capital when a critical investment opportunity strikes or the fund needs to defend itself from a tough financing in a good company, It can lead to the need to raise a new fund sooner than expected, which can be before the first one is mature or before LPs are expecting Fund II to be in market. Thinking strategically about portfolio construction and explicitly measuring and adapting is a pro technique that can help grow ownership in your best companies, protect the firm, and prepare you for raising the next fund.  

Greg Sands, Founder and Managing Partner at Costanoa Ventures

An early-stage investor for over 25 years, Greg prides himself on being the first call for founders whether they have good news or bad news to share. Because great product is the foundation of great venture-backed startups, he also focuses on helping them develop product strategy, including focusing on the initial wedge and the Ideal Customer Profile. Some of his notable investments include Alation, Auterion, Cape, Demandbase, Highnote (co-led), PayNearMe and StackHawk, with successful exits most recently in SGNL (sold to CrowdStrike for $740m in 2026) and Aquabyte, as well as Datalogix, Inflection, Intacct, Roadster and Kenna. 

Before founding Costanoa, Greg was a Managing Director at Sutter Hill Ventures, focused on early stage, enterprise technology investments, including Merced Systems, Quinstreet and Youku. Before Sutter Hill, Greg was the first hire at Netscape Communications after its founding engineering team. As its first Product Manager, he wrote the initial business plan, coined the name “Netscape” and created the SuiteSpot Business Unit, which he grew from zero to $140M in revenue. Greg also served as Manager of Business Development at Cisco, where he architected a global channel management plan. He holds an MBA from Stanford and a BA from Harvard.