In the first part of the series on corporate venture capital I explored how the disruption of institutional VCs (IVCs) and the imperative for corporations to innovate provide an opportunity to corporate VCs (CVCs) to make their mark in the startup ecosystem and be viewed as viable and valuable financing sources to private companies. In the second part I provided more context on CVCs by presenting a brief history of corporate venture capital, and detailing the characteristics of CVCs during the dot-com period and today. In this blog I discuss when corporations should be establishing venture funds, I introduce a framework for creating venture funds and discuss two of the dimensions in this framework.

Three questions to ask before starting a corporate venture group

Over the last year several corporations have asked me for advice on the venture groups they plan to form. These days corporate venture funds are announced on a weekly basis (and here). Before providing my advice I ask three questions.

First, whether the CEO is ready to lead the corporation’s innovation goals, and see the venture group as one of the means for achieving these objectives rather than the panacea. I have already discussed the CEO’s role in establishing the company’s innovation culture. The CEO’s tenure, contract, and staying power with the corporation are two additional important factors for the success of these initiatives. Based on my experience, a new CEO who has signed a multi-year employment contract with the company, or an existing CEO in the beginning of renewed contract, typically approaches the creation of a venture group very differently from an existing CEO at the end of her employment contract. Several partners of newly established CVCs have also admitted that CEO tenure is one of their major concerns regarding the long term prospects of their venture groups, their investment efforts and the commitments they make to their portfolio companies.

The CEO’s staying power with the corporation is also very important.  Innovations, particularly disruptive innovations, may take several years to achieve.  Think of how long it has taken Amazon to develop AWS, or Apple to develop the iPhone and how these innovations benefited from the staying power of Bezos and Jobs.  HP, on the other hand, had six CEOs over the past 15 years; something that hasn’t helped it recapturing its innovation roots.  The CEO must be able to continue supporting the innovation initiatives regardless of the pressure she may be receiving from the markets and shareholders for short term financial performance.  This is also why in the second post of this series I spoke about timelines and how institutional VCs investing in early stage companies think in terms of 7-10 year horizons for a startup to reach maturity.  Innovative companies are led by CEOs with long tenure and staying power.

Second, I ask why the corporation wants to start a venture group and what will be the mission of the group. Typically corporations create venture groups because they believe they provide the best way to invest in the startup ecosystem instead of providing them with financial returns. I have found that corporations invest in the startup ecosystem for the following reasons:

  1. Create an acquisition pipeline including identifying opportunities for the acquisition of talent.
  2. Support existing or prospective partners, including accelerating the creation of a partner ecosystem around a particular product, e.g., IBM’s Watson Fund that supports the Watson cognitive computing platform, SAP’s HANA Fund that supports the HANA big data platform, and Microsoft’s Azure Fund that supports the Azure cloud computing platform.
  3. Understand a sector and the associated market with its dynamics.
  4. Provide over the horizon view of new technologies and business models.

It is important to understand which of these reasons are important to the corporation and how they relate to its innovation goals. Depending on how a corporation prioritizes among the four investment reasons provided above, conflicting behaviors that diffuse the effectiveness of the corporate venture group could arise. For example, business units within the same corporation may not share the same over the horizon perspective about a particular technology. Moreover, central R&D organizations always believe they have the best and most updated perspective on a technology. Therefore, before creating a corporate venture group the list of reasons provided above not only has to be created but it has to be prioritized.

Third, I ask whether the corporation has considered other options instead of forming a venture group. Corporations have four investment options. They can invest:

  1. Off the corporate balance sheet. For example, Intel’s $740M investment in Cloudera.
  2. Directly from a business unit. For example, Cisco’s investment in Mavenir Systems, and Cisco’s and VMWare’s investment in Hytrust.
  3. In an institutional venture firm. For example, 10 corporations had invested in the Java fund that was managed by Kleiner Perkins and many other corporations have invested in the institutional venture funds.
  4. Through their corporate venture arm. For example, Intel Capital’s big data investments, e.g., Guavus, MongoDB, etc.

I have the following recommendations for when to create a corporate venture group:

  1. If creating an acquisition pipeline, or supporting partners are of high priority, then the corporation needs to establish a venture group. Moreover, if the creation of acquisition pipeline is of particularly high priority, then the corporation should consider establishing a venture fund and investing in IVCs whose focus is consistent with the corporation’s acquisition interests.
  2. If understanding sectors in breadth is very important, then the corporation needs to establish a venture group and invest as an LP in IVCs whose investment theses are consistent with those of the corporation.
  3. If understanding markets and geographies where the corporation doesn’t have direct access, and providing over the horizon views are of high priority, then instead of creating its own fund the corporation should consider investing in several top tier IVCs as a regular LP, again paying attention to the compatibility of each IVC’s investment theses and portolio. This may be a more economical and effective way to proceed. To understand why consider two notable examples: Kodak Ventures and Intel Capital. Kodak pioneered digital photography and photo sharing, and even though it established a corporate ventures group in 2000, squandered its lead in these sector and we know the rest. Intel made big bets on Wi-Fi first with the Centrino chip, later with Intel Capital’s investments in WiMax. Yet, despite its lead in the Wi-Fi sector, the corporation has missed the mobility revolution and to date remains a small player.
  4. Finally, I recommend that the CVCs of corporations in industries with long innovation cycles, e.g., agriculture, oil and gas, etc. CVCs should always be actively collaborating with IVCs.

The five dimensions to employ while setting up a corporate venture group

When considering the creation of a venture group I use five dimensions of a framework I have developed (Figure 1):

  1. Strategy.
  2. People.
  3. Incentives.
  4. Deal flow.
  5. Governance.

Figure 1

Figure 1: The dimensions corporations must consider for their venture group

Strategy

The corporation must establish a long-term strategy for the venture group. This means that the corporation should be thinking of multiple funds over which to execute this strategy, rather than a single fund, as well as the optimal size of each fund. In this way while starting to invest fund 1, the corporation and its venture group are already thinking about funds 2, 3, and 4. The components of the strategy include:

  • A set of objectives, consistent with this strategy, for the venture group to execute. Primary among these objectives is whether the venture group will be financial only investor, strategic investor, or both. The amount of the fund that will be allocated for investments in business models already being used by the corporation, new but market-tested models, and completely disruptive models.
  • An investment thesis that is typically refreshed every 18-24 months. Along with the theses, the CVC group must also specify the sectors and industries where to invest, e.g., data security, big data, agriculture, etc.  The theses may change 20-30% from period to period.
  • The stage of the target investments. The corporation must determine if the venture group will focus on investments of the same stage (early, growth or late stage), or it will have a multi-stage focus, taking into account the corporation’s risk tolerance level and the amount allocated in each fund. For example, until they raise their first institutional round, IT startups use their seed funding to test hypotheses: technology hypotheses, business model hypotheses, market hypotheses, etc. The startup’s risk level during this phase of its development is extremely high.
  • The life of each fund.  Most funds have a five-year investing period and a similar-length harvesting period. Corporate funds may need a shorter harvesting period depending on their overall goal (strategic vs financial).
  • The amount per investment, both initially and over the life of the investment. The investors in the CVC group need to develop a line of sight for each investment, i.e., need to have an understanding of what can happen to each portfolio company based on market conditions before actually investing in it. Developing such understanding is particularly necessary when participating in more than one financing round. This is how the good institutional VCs think about their investment candidates.
  • The governance of each investment, in other words whether the CVC group members will be taking board seats in the companies they invest, or just observer seats.  Implicit in this decision is also the decision on whether the CVC will be leading rounds, and, if so, under what conditions. For example, if investing in early stage companies, then even if the corporate venture group can lead rounds, it should always syndicate with IVCs with whom it has strong relations. This is something we will explore further when we discuss the deal flow dimension.

Many corporations these days are committing $100M for their inaugural venture fund. The corporation must think why $100M is the right amount for accomplishing the venture group’s objectives. For this reason it is instructive to look at institutional venture firms and their funds.

Institutional venture funds have a 10-year life. New investments are made during the first five years of the fund, and most typically during the first three. If the fund invests in early stage companies then an amount equal to 100% of each initial investment is allocated for follow on investments to these companies, implying that approximately 50-60% of the fund will need to be allocated for follow on investments to the most promising of the portfolio companies. At least 30% of the allocated follow on amount is invested during the first five years of the fund’s life to these companies. This is because early stage companies typically raise money more frequently in the beginning of their lifecycle as they build their solution, recruit their initial customers and try to establish themselves in their target market. If the IVC decides not to continue investing in a particular portfolio company then the amount reserved becomes available for new investments.

If we assume that the CVC follows a similar pattern to invest a $100M fund, i.e., 3-4 years, then we can expect an investment pace of $25-35M/year, since there is no management fee. Most CVC groups, particularly the newly formed ones, typically have 2-3 investing partners. This means that each investing partner is expected to invest on the average $8-15M/year between new and follow on investments. Assuming the CVC invests $1-3M in each early stage company, regardless of whether the CVC leads the round or is just part of an investment syndicate, and expects to make follow on investments to 40% of this portfolio, then creating a portfolio of 25-30 investments over the life of the fund is possible, implying 8 or so investments per year. Finally, assuming that the group invests in 1-2% of the business plans it considers, and with the above analysis in mind, it is likely that a $100M fund will only enable the corporation to use the CVC group only along a specific investment thesis, such as Nokia’s Connected Car FundAccel’s Big Data Fund, and DCM’s Android Fund, rather than along multiple markets.

If the corporation wants to focus its venture fund on late stage investments, then $100M allocation implies a small number of meaningful investments (8-12) with $8-10M per investment.   This is the main reason the amount has to be large enough as a percentage of the total amount invested in the company so that the corporation can have information rights as a result of the investment which will enable it to learn through its association with the private company.

Adopting a particular strategy has implications on investment timelines. For example, for investments that support existing business lines and models, it would make sense to expect returns within 3-4 years after the initial investment is made. For investments targeting new but market-validated models, expect returns within 4-6 years after the initial investment is made. Finally, for investments targeting disruptive technologies and business models, expect returns within 7-10 years after the initial investment is made.

My recommendations regarding strategy include:

  1. Focus in the same areas where the parent is working. Successful CVCs, e.g., Johnson and Johnson Development Corporation, Comcast Ventures, tend to do exactly that.
  2. Funds of $100M should be pursuing a single stage investment strategy. Larger funds, e.g., $300-500M, may consider pursuing a multi-stage strategy.
  3. For funds pursuing a single stage strategy around early stage investments, a potential fund allocation could be 30% of the investments to support existing business models, 50% towards new but market-tested models, and 20% towards disruptive models, thinking that 20% disruption may lead to 80% value for the corporation.
People

Venture capital is a peoples business. Therefore, very much like is the case with startups, the quality of the people associated with the corporate venture capital group are a good indication of its success prospects. In my framework I call for six teams that need to be associated with the corporate venture group.

  • Leadership team. Contrary to IVCs that have partnerships and an operating committee that is staffed by a subset of the firm’s senior partners, corporate venture groups usually have a CEO. The CEO tends to be a senior corporate executive. Depending on the size of the venture group, the leadership team also includes a CFO. I prefer flatter organizations, where the partnership is also the leadership team, because such organizations create a better culture of ownership.
  • Investment team. This group consists of both partners who can lead investments and other investment professionals, e.g., vice presidents, associates, who support them. Partners are expected to identify potential investment opportunities, lead the pre-investment due diligence effort, invest in companies, and manage each investment.

The size and composition of the investment team are particularly important. The team needs to be large enough for deploying each fund and to consist of more than investing partners. The partners themselves should be experienced enough investors a) for the entrepreneur to want to collaborate with them, b) to quickly filter out inappropriate investment opportunities, but pick out the right opportunities. As a VC I have always learned that it is easy to reject an investment opportunity. Picking the right opportunity to invest and generate financial returns is much harder as the fund performance shown in Figure 2 demonstrates.

Figure 2: US Venture Fund Returns 2004-2013

Figure 2: US Venture Fund Returns 2004-2013

As Figure 2 shows few institutional venture funds achieve even a 1.5-2x ROI, i.e., at that ROI a corporation may double its investment on the venture fund. But some rejections prove to be very expensive, as do many selections (see the list of funds with low ROI).

  • Investment committee. This committee is responsible for ultimately deciding whether a particular investment will be made under the terms recommended by the investing partner and based on the results of the investment team’s due diligence effort. This committee should include the CVC roup’s investing partners, the group’s CEO and the corporate, or business unit, CFO, particularly if the corporation is investing off its balance sheet. The investment committee could also include representatives of the company’s strategy, corporate and business development organizations, particularly if the CVC is part of the corporate structure than a standalone entity.
  • Advisory board(s). Every good institutional venture firm uses advisory boards and some even have more than one. The members of these boards come from the sectors the firm is interested in investing and provide a valuable outside perspective that is used in order to formulate new investment theses, and identify particular sectors and companies to consider investing. For example, they inform the investment professionals of important problems they are facing, significant trends in their industry, solutions being considered, particularly by incumbent vendors. Even though corporations have first-hand understanding on high priority problems to address, it is highly recommended for every CVC to use an externally staffed advisory board to receive additional feedback that can lead to a new investment idea, or lead to abandoning a particular investment thesis.  The corporation may also consider forming an internally staffed advisory board that includes representatives from the R&D, strategy, corporate development and business development organizations.
  • Knowledge transfer team. Many CVC senior investment professionals told me that much of the knowledge that is gained by the corporate venture group through interactions with the startup ecosystem as well as with particular portfolio companies is lost because typically there is no dedicated team whose mission to transfer this knowledge back to the corporation. The knowledge is lost either because it has a short shelf life or because the investment team frequently changes (sometimes with CVC investment professionals joining IVC teams, but mostly either to join another CVC or take another position in the corporation). For these reasons, it is important for every CVC to have a team whose mission is to transfer the knowledge gained from investments, prospective investments and market analyses back into the corporation. IBM’s Venture Group and InQTel have been pioneers in establishing such teams.
  • Portfolio services team. As I have mentioned in the past, entrepreneurs expect more than money and periodic advice from their venture investors. Once the corporate venture group determines the type of services it wants to offer to its portfolio companies, it should establish a team that can provide these services. Example services can include:
  1. Access to business and process knowledge, as well knowledge of specific geographies that can be invaluable as the startup considers international expansion.
  2. Sales, marketing and recruiting mentorship and execution from corporate executives.

I recommend that the members of this team can come from the corporation’s business units.

In the next post we will continue presenting this framework by discussing incentives for the people in the corporate venture group, deal flow generation and governance issues.