Building a Board: The Impact of Company Stage

In response to the original post about the “Building a Board” results (see below), Tim Connors of USVP inquired about the role played by the company’s stage of development. Even though our analyses (and the original blog posting) focus on the linkage between Board characteristics and CEO characteristics, we also controlled for the impact of the company’s stage of development. (Other controls were the company’s industry segment and the time period in which the data was collected.) We used the following Company Stage controls:

  • # rounds of financing
  • Dollars raised in last round
  • Revenues in current year
  • Company age (months)
  • Number of employees

Here is the impact of each variable on our 3 Board dimensions. (Empty cells indicate relationships that were not statistically significant.)

So the results support Tim’s contention that Company Stage has a key impact on Board characteristics, but hopefully add more details to that, highlighting which Company Stage variables are the strongest drivers of each Board characteristic.

Once again, feel free to post any thoughts on individual cells in the table (whether they make sense to you, whether you’re surprised that they’re empty, etc.) or on other overall patterns we are missing!

(One last note on the dataset: The dataset we are using for this paper is comprised of 444 private IT companies, collected in 2000, 2001, and 2002. We use the three years to help eliminate any artifacts introduced by market fluctuations, using year dummies to control for differences across the years.)

  1. Noam,I am curious about two factors related to boards and am wondering if your research has explore these areas. First, has to do with outside (non-VC/non-mgt) board memebers. The second has do with female (outside) board members.Have you looked at these issuses and if so, how do they correlate with the findign in this post?

  2. Paul:We’ve examined the “outside” side of boards pretty extensively, in two ways. In the first, we used a metric for the % of <>all outsiders<> on the board (instead of the % of VCs on the board), and found similar — though a little weaker (from a statistical significance perspective) — results as we did with VC-%, which wasn’t surprising given that there’s a high correlation (significant at the extremely stringent p< .001 level) between VC-% and outsider-%. In the second, we took just the % of <>non-VC outsiders<> (i.e., consistent with your definition of “outsiders” above). There, very little showed up as statistically significant, both because the numbers of non-VC outsiders are too small and because non-VC outsiders are a relatively heterogeneous bunch. (This is actually one area where I think my Life Sciences data might yield better insights.)In case you’re interested, here are some high-level board-composition numbers for our boards:% insiders: 30%% angels: 8%% VCs: 35%% industry execs/reps: 18%Regarding your outside-director gender question, we unfortunately do not have the genders of our directors, so we don’t have anything systematic to report there. Any thoughts on how things would differ by gender?

  3. Noam: Interesting series of posts. I had a quick thought on the relationship between the # of rounds a company has raised and the size of its board. While I am not surprised that the relationship is a positive one (i.e. more rounds equates to larger boards), I wonder whether there might be a subset where, in fact, a negative relationship exists. Specifically, it has been my experience that when a company has raised capital through multiple rounds, and then undergoes a recapitalization or washout round, that the board size tends to shrink. Investors often use the recap as a cleansing event and included in this cleansing is board composition. Non-participating or weak investor groups may be taken off the board and not replaced. The larger investors view a smaller board as a way to more efficiently control a bad situation that they are trying to improve.While my observation is anecdotal, it would be interesting if there was a way to test this in your dataset. I’m not sure what, if any significance the results might show but figured it was worth mentioning.

  4. Good point, Scott — that’s not only an interesting subset, but a potentially important one. A big part of what you’re pointing to is a Board-level impact of a major down-round, in addition to the executive-level (e.g., higher rate of CEO turnover?) and equity-level (major dilution for past equity holders) changes that occur. Because I have the full financing histories of my companies, I’ve been able to compute down-round indicators for another study. When we revisit our Board-size analyses, I can add those indicators (for whether the company has ever had a down-round, had a down-round in the last round, etc.) and see if they have a significant (presumably negative) impact on Board size.

  5. would be interesting if someone did a study on which ceo attributes had a statistically significant effect on outcomes. we all have our preferred set of criteria for a ceo. would be interesting to see if there is a set of attributes that have proven to return more often.another question is founder succession. Some vc’s swap the technical founder for an experienced ceo quickly after investment. Some think swapping the founder too soon is death to a company. While every situation is unique, wonder if there are some insights from the data.

  6. Tim:Regarding your “CEO attributes” issue, there would seem to be 2 sides to it: 1. <>When the CEO is a founder<> (i.e., what founder-CEO characteristics yield better returns). In this case, it’s where VCs are choosing to invest in an existing (founder-)CEO, and the question would be which founder-CEOs you should <>invest in<>.2. <>When the CEO is a non-founder/successor<>. In this case, it’s where VCs often are helping choose the CEO, and the question would be which successor you should <>hire<>.Both of these are related to the <>”After the Firing”<> work I mentioned in the “coming attractions” posting. In addition, your second question (about when to replace a founder-CEO with a nonfounder-CEO) is squarely at the heart of that paper. My co-author and I happen to be heading into our next big phase of that project (our initial ideas of what questions to ask drove our data collection; now that we’re ready to analyze our data, we’re going to revisit the research questions) and I’d love to get other input on what we’re planning. To do so, I’ll devote my next posting to it; let’s use the blog to hash out further issues we should examine.

  7. I’m a bit surprised by the 18% non-mgt, non-vc board member statistic. It seems high to me for tech deals, but could be biased by the life sciences data. Does you data allow you to identify when outsiders are brought on-board, i.e. by series, stage or size?WRT female board memebers, my experience is that the numbers are low overall, but higher on the life sciences side of the portfolio, and higher still on SAB’s. That said, my personal observation is women add a distinctive and important perspective beyond their individual experience or functional expertise. This perspective influences board dynamics in a very postiive way.

  8. As we migrate into the Founder-CEO topic. I’m imagining that your “after the firing study” will address the “first time” vs “serial founder” issues. (Assuming that your data supports the analysis.) The skills, experience, expectations and issues of each are quite different and have an direct influence on who initiates the transition, as well as, the “when, why and how” of the transition.On your point of “hiring the right CEO”, there are two factors that that we pay particular attention to if the founder is staying on board: 1) The get along factor, as in will the founder “get along well with the new CEO”, and 2) the mentor factor, as in “Can the CEO be a coach/mentor to the founder, and is he/she willing to play that role?”

Leave a Reply