We have a few of these folks inside of CSI. Mostly they are young. These growth-oriented employees usually leave us after a few years to pursue their organic growth aspirations elsewhere, or they focus their energies on doing acquisitions. Some entrepreneurial folks come with acquisitions and hang around CSI for a few years, often because of earn-outs. Some of the ex-entrepreneurs in our ranks are older & bored and lust after the energy and sense of purpose that they used to have in a smaller milieu. A few energetic employees find their way into our dozens of tiny to small initiatives, where we treat them poorly. They get virtually no upside, considerable downside, and are limited to missions which, even if wildly successful, are expected to lead to the creation of low single digit million dollar revenue businesses which usually get reabsorbed into the sponsoring BU over a 5 to10 year period. VMS Ventures will finance larger initiatives with standard venture capital equity ownership structures. Initiative employees will own the majority of the business at the seed stage. No BU manager or PM will be peering over their shoulders. Rarely will the initiative employees be diluted below 20% ownership. In successful ventures they will likely own far more equity than that. We expect the initiative entrepreneurs and their teams to work at a frenetic pace that leaves the average CSI employee in the dust. They will have high upside and no safety net. That risk / return formula will only appeal to a small group of entrepreneurial employees. Of those, we only need 20-40 to lead the fund’s investees.
The old argument about cannibalising your own businesses applies… you should do it before someone else does. That theoretical answer aside, CSI’s software has high switching costs resulting in low attrition, so our new startups would be nuts to compete head-to-head with CSI. We’d rather see them competing for underserved segments by providing a differentiated offering or building add-ons that can address an entire vertical (i.e. both CSI’s clients and the clients of CSI’s competitors). A CSI operating group, portfolio, or business unit can always offer to buy one of the fund's investees. If Constellation does not offer to buy it, or is not willing to pay a price that investee management view as fair, then the fund and the initiative managers can either run the business and distribute the profits (these businesses can generate significant dividends), or continue to grow it and perhaps take the business public if it is extremely successful.
If the fund is interested in financing the existing initiative and there is a backable team, then the BU will be able to contribute the initiative's assets to a newly formed entity in return for some initial (to be negotiated) percentage of the equity in the new business. The Fund will only invest if it is in the lead investor role.
Yes, if there's a backable team, and a strategy that is not dependent upon a forced upgrade of a legacy base owned by a CSI business unit. Our preference is to create lower-priced products with less functionality than existing legacy products, and to take advantage of SaaS features to rapidly evolve these products into whatever market segments adopt them quickly. Wherever possible we will seek out unique "price carriers" to differentiate our offering from conventional "per seat" or "per user" perpetual license pricing.
Yes. CSI has no monopoly on good VMS ideas. However, the Fund is more likely to invest in verticals that CSI already knows.
Definitely. I expect most of the fund investees will have some of their capital from a sponsoring business unit. The only reasons that the business unit would not want to own the entire initiative are that 1) equity is needed to compensate the initiative’s employees, 2) most business units cannot afford to have a large initiative fail, or 3) business unit management believes the probability weighted expected returns of the initiative are unattractive. The fund will have 20-40 investees, so it does not have the same concern about any individual investee failing. The fund will only invest if it is in the lead investor role… the Fund does not need to put in the majority of the investment, but it does need to have significant influence upon the board of the investee.
It gets worse. I worked in the venture business for over a decade with a couple of the larger Canadian funds. The results for our investors were not attractive. Then I started CSI… so maybe I learned something along the way. VMS Ventures will have some advantages vs traditional VC firms. In nearly all instances, we will know the employees whom we are backing and the markets that we are pursuing. We have done diligence on literally thousands of successful VMS businesses, and we have previously launched hundreds of initiatives. We have a large network of “opportunity seekers” who will send us potential investees (31,000 employees and our hundreds of thousands of CSI clients). These advantages are somewhat offset by the natural rejection/antipathy/jealousy that corporate venture capital firms experience in dealing with their parent corporation. We have very high-level sponsorship within CSI to help overcome these disadvantages.
It boils down to a huge embedded real option. The fund will be climbing a learning curve, developing and refining processes for financing and growing niche VMS startups. If it is successful and CSI’s organic growth rate increases over time to a sustained high single digit rate, finance theory and common sense tell us that enormous amounts of value will accrue to CSI’s shareholders and employees. We will start off with two very significant process advantages vs the ’04 to ’10 Initiative experiment: We will target much larger Initiatives (minimum $10MM expected revenues), and we will compensate the employees with significant equity in the startups. We believe these changes will attract better employees/founders and better investment opportunities than we pursued in ’04-’10. While we may not initially generate acceptable returns from our investees, we think we will get better over time. This is analogous to the path we followed when we started doing acquisitions. We gradually refined our M&A processes until we felt comfortable that we could invest billions acquiring VMS businesses at attractive rates of return. Ideally, three to five years into the VMS Ventures experiment, we hope to have data that supports the investment of billions of dollars with attractive expected IRR’s. That’s the upside. On the downside, we learn something about organic growth, get poor returns on two months of CSI’s free cash flow, and instead focus solely on investing billions in VMS M&A.
Few companies simultaneously maintain high new product growth and high profitability, so there may be something to the exploration/exploitation trope. CSI’s experience also supports the existence of an exploration / exploitation tradeoff: During the ’04-’10 experiment we found that initiatives with dedicated full-time leaders were far more successful than those with leaders who were also responsible for some portion of their parent business’ activities. A small group of employees from an existing CSI BU who launch a new enterprise can concentrate on exploration, without having to worry about exploitation. They can start with a hypothesis and experiment fast to refine their hypotheses. They won’t have to worry about supporting legacy products, doing performance reviews, participating in diligence for potential acquisitions, and responding to head office surveys. The experience of other corporate venture groups is that internal startups are soon ostracised from their sponsoring business unit, but I’m hoping that the sponsors holding a continued minority equity interest, combined with curiosity will result in VMS Ventures having a better experience than the corporate venture groups that I’ve studied. The employees who “defect” to a VMS Ventures startup will sometimes reduce the capabilities of their sponsoring BU. In many instances, they will already have been the outliers and contrarians who resisted bureaucracy and “efficientisation”, so I think the redeployment of troublesome / high maintenance staff to more appropriate endeavours may be welcomed by some BU managers.
Only a tiny portion of all startups address billion-dollar markets that can support the economics cited above. For example only 0.3% of YCombinator’s >3000 ventures are public companies. About 5% of their portfolio might be considered “successful”. 4/8 VMS Ventures will be financing investment opportunities with tens of millions in total addressable market (“TAM”). Even targeting these smaller TAMs, VMS Ventures expects the founder of a “Winner” investment for VMS Ventures to build a double digit million dollar net worth within a decade. Our hope is that the founders love the businesses that they’ve built, the customers that they serve and the team that they lead and will continue to compound their intrinsic equity value, doubling it every three to five years for many decades thereafter. If, however, after a decade of growth, the founders are satisfied with a stable business that serves a loyal group of clients and provides a great working environment for a skilled team, while paying significant dividends, we would be equally delighted.
CSI is the most likely provider of "exits" for the Fund and the employee shareholders in the Fund’s Investees. Independent fund management is necessary to stop CSI preying upon the investee shareholders. The “loser” investments will be obvious before the winners. The losers will also outnumber the winners. We don’t want the CSI board or the BU GM’s deciding when to call it quits on an investee or on the fund. Their job is to make that call once the fund has made the investments and the experiment has run its course.
For the fund, it means that an experienced and respected group are recommending that we consider the investment. That carries a lot of weight. For the potential investee, BU sponsorship is a necessary rite of passage. If the BU does not think the idea is worthy of the fund’s consideration, and/or that the team is backable, it is extremely unlikely that the fund will consider the proposal. For the BU, sponsorship does not have to be an unqualified endorsement. The fund will work with BU’s and potential founders to get an initiative to a stage where it warrants investment by the fund.
The investee should have: • The credible expectation of $10MM in annual revenues within 5 to 10 years. • A target vertical market (or close adjacency) where CSI already operates. • A sponsoring CSI business unit. • A backable team, or the key initial members thereof. • An earned secret.
Generally, a backable team will consist of at least one experienced person who knows the vertical, plus a strong technical lead, and a handful of talented developers. Initial team size will rarely exceed 7 unless there are customers enthusiastically using the early products. We may lean on CSI’s development shops in Pakistan and Romania for rapid prototyping, to keep the investee’s burn rate and headcount low until traction is achieved. Beyond these skill sets there are other characteristics that we are looking for in the team. We want a frugal, analytical, urgent and ambitious team. We are trying to run many cheap and fast experiments to find 5/8 product / market fit. People who wish to lead balanced lives, with a multitude of social and family obligations, are unlikely to flourish doing startup work with VMSV. The backable team doesn’t need to be complete from the outset. The fund can help recruit some team members. Keep in mind, however, that if high quality team members aren’t enthusiastic about joining you, either you or your business idea may need more work before you are backable.
Great vertical market software businesses tend to have high switching costs. The switching costs are high because of the time required to implement new systems (i.e. for customization, training, and interfaces), and to weave them into the operating processes and habits of the customers. That makes winning greenfield market share expensive and time-consuming. But it also means that taking market share from incumbents is even more difficult. In markets with these characteristics, your installed base is very profitable, but acquiring customers can be slow and very expensive. Depending upon the competitive environment in our mature markets, we sometimes pay extraordinarily high customer acquisition costs for new name clients. In less mature markets, we can accelerate adoption of our products by subsidising the customization/training/interface costs for our customers. We are happy to do that, so long as we see product utilization is high, have evidence that the systems will be “sticky”, and that they will deliver significant value to our clients. The tactic of subsidising customer adoption costs is not news to the venture community nor to software vendors generally. For big and obvious greenfield markets there are dozens of well-funded startups and corporate contenders. Competing with deep-pocketed contenders in hotly contested markets where positive cash flow is many years down the road is not appealing. This is where the earned secret comes in. If you have a unique insight into a particular customer problem through deep knowledge and experience in a vertical, and this insight is not obvious to agile and deep pocketed competitors, then you have an earned secret. That allows you the opportunity to grab a sticky greenfield market by subsidising customer adoption. We believe that our employees are the fortunate holders of thousands of unexploited earned secrets. So, bring them on.
The simple answer is: all three of them. Scaling up of startups could refer to startups from within the CSI organization or from the outside
Well, CSI often operates businesses in mid-market segments providing software stacked with features developed over a 20 to 30 year period. The fund is interested in several plays. We mention two of them below. Low-end SaaS based VMS products with limited functionality, positioned below our mid-range products are likely to attract a different customer base, but one with which we are somewhat familiar. These products will have less customisation, more remote selling and training, lower price points and more customer churn, often accompanied by a slightly more broadly defined addressable market. SaaS shines in this kind of 6/8 application because it provides the usage data to allow rapid product evolution for the niches that display need and adoption. The experiments are cheaper and faster than with non-SaaS platforms. Another play could be a major add-on (built agnostically) on our and our competitors’ VMS product.
The investor in the fund will be CSI. We are working on setting up the fund so we can flow the early operating losses from the investees through to the sponsoring BUs and their Operating Groups for tax purposes, thus allowing them to save cash taxes. This may not always be possible, but where it is, it lowers our after tax cost for the fund investment and provides another incentive to the BUs to sponsor investees. In each investee, the investors will be the fund, the sponsoring BU and the entrepreneur. The fund will be managed with a lean virtual office. The fund’s job is to deliver data, learnings and returns to CSI and provide equity and coaching to the investees.
CSI has delivered a proven track record of acquisitive growth and operational excellence. In addition, CSI’s business units have started and launched hundreds of organic growth initiatives. As a result of the focus on acquisitions and operational excellence, those initiatives have been relatively low risk and have invested only a modest amount of capital. We would like to invest an order of magnitude (or two) more capital in organic growth initiatives at attractive returns. We need to build up our organic growth best practices and track record with larger initiatives that deploy larger amounts of capital. Conventional venture capital jargon such as daily active users, customer acquisition cost and unit economics will be more important to the fund than CSI’s more frequently tracked ROIC, EBITA margin, and Sales and PS ratios.
The main VMSV activities will be sourcing, investing, and supporting/coaching the investees. The fund will not need to worry about funding and exits for many years. Sourcing will be done outbound through the press release, this Q&A, and our web site (pending), and inbound largely from the CSI global ecosystem. Investing will be a matter of screening the investment theses, performing due diligence (light), generating the term sheet, and executing the required legals (ideally templated). Supporting/Coaching will be achieved via a mix of partners from the inside and outside of CSI.
That will depend upon the stage at which the investment, idea, or initiative is at when the fund invests. If we are going to build something from the ground up, we could imagine a seed round ($300K) followed by series A ($1M) a series B ($2-3MM), and a series C round ($5MM+). In a winner scenario, an investee may consume $5-10MM of capital over multiple rounds.
The regular CSI ratios are hard to apply to early-stage initiatives. The culture within CSI is strongly datadriven: numbers prove that things are ok. Data-driven hypotheses and qualitative and quantitative evidence are still going to be an important part of the VMSV investees.
We’ll use modern SaaS frameworks, but they will be “middle of the road” and will try to avoid cloud provider lock-in
We will be organizing some events, like (online) brown bag lunches, to discuss the fund and the opportunities it seeks. We would appreciate your feedback through these lunches or via any other channel. Any references to successful internal venture capital funds would be much appreciated.
Firstly, it is awesome you’ve got an initiative of which you expect to grow into a $10MM annual revenues business. (keep in mind that ML would never say “awesome”… so this clearly came from one of his younger partners). Please send us a note (Karl, Mark or Daan). We’ll read your investment thesis and reach out to you. Please make sure you have a sponsoring business unit within the CSI ecosystem for your proposal.
The fund will take minority as well as majority interests in the investees to be determined case by case and depending on the other shareholders’ wishes. The fund will only invest if it is in the lead investor role (i.e. is setting the terms of the investment and governance). The Fund does not need to put in the majority of the investment, but it does need to have significant influence upon the board of the investee. In most seed round situations, the fund will own less than 20% of the underlying economics, while the sponsoring BU and the employees of the investee will own the remainder.
ROIC will be irrelevant to the fund managers (max 5 people). The net revenue growth achieved by their investees will be the primary driver of the fund managers’ incentive compensation.
While we may not initially generate acceptable returns from our investees, we think we will get better over time. This is analogous to the path we followed when we started doing acquisitions. We gradually refined 8/8 our M&A processes until we felt comfortable that we could invest billions acquiring VMS businesses at attractive rates of return. Ideally, three to five years into the VMS Ventures experiment, we hope to have data that supports the investment of billions in organic growth initiatives with attractive expected IRR’s.
We know that within the TDC Operating Group TSS has started to use some of TOG’s practices. The recent sharing of the micro-investments framework by TOG at the 2021 CSI M&A Summit has also led to further study, analysis and application at other groups within CSI.
The fund would be happy to hear from M&A teams about such businesses. We can do a joint call with the prospect and the M&A team member. We think these leads can be valuable for the M&A team as it provides the chance to be on the call with the founder and the fund. The most frequent stumbling blocks will be the loss of founder’s control (to VMS Ventures), and the fund’s inability to bring operating synergies to the investee. If synergies are a big part of the thesis, then a “micro-investments” (see #31 above) approach may be better than a direct fund investment.