I haven been working with a number of companies lately who are trying to evaluate the relative merits of staying Independent versus raising Institutional venture funding or selling themselves. This involves making some long-term models and thinking way past the horizon, and always touches on the subject of what an eventual IPO might look like. When I started my consulting practice, D/D Advisory, I envisioned this being the bulk of my work. I call this Long-Term Strategic Finance, the non-accounting side of a CFO’s role.
The IPO window is still open, albeit at discounted valuations. And companies are still raising money. Nonetheless, I have found companies have gotten much more interested in this Long-Term Finance topic because of a growing sense that the days of easy money may be ending soon (or at least taking a lengthy pause). Capital raising moves in cycles, and I am not a believer in doom-and-gloom scenarios. Nonetheless, I think executive teams need to have a clear understanding of their options. As many of these strategies take years to play out, it pays to be prepared and think a few steps ahead.
To make this more tangible, I built a financial model for a fictional company. Let’s call the company Tyrell Systems, they are a SaaS (software as a service) company. Since they sell software subscriptions, they have relatively low upfront capital requirements, and relatively predictable revenue streams. Below are the company’s (made-up) financials:
I deliberately chose a SaaS company for a few reasons:
- Many of the companies I have been speaking with on this subject lately are SaaS providers.
- SaaS results are relatively predictable, lacking the volatility that public market investors hate.
- This category has many clear valuation precedents which we can use
The trick to valuing a SaaS company in today’s market is that investors focus on revenue growth. Normally, I am not a big fan of using revenue multiples to value a company, but the average SaaS company has a pretty straightforward correlation between revenue and the cash flow metrics that I would rather use. The good news is that investors, both public and private, are willing to assign premium valuations to SaaS companies. In particular, there is a big, non-linear premium paid for high-growth companies. I did a quick web search to estimate valuation multiples. These figures are rough, but probably pretty close to current market multiples. (If anyone wants to share their comps model, please send it over in full confidentiality.) These are the multiples I used:
I should note that these number come as a surprise to many CEOs. For seasoned execs, the high end of this range are some pretty high numbers. But for early- and mid-stage venture rounds, the multiples can go even higher.
Picking the Right Time to Sell
Using these multiples, I came up with the following analysis of the optimum time to sell the company or take it public. The bottom row of this analysis is PV or Present Value. So the chart should be read as: if Tyrell sells the company in Year X, it is worth the PV of that year. For example, selling the company in Year 3, would generate $605 million at that time, which is the equivalent of $378 million today. As a reference, the full discounted cash flow valuation of the model comes in at $177 million, which is a very loose proxy for an IPO valuation. (In practical terms, public market investors are going to use some form of revenue metric as well.)
The biggest driver of valuation is revenue growth. If your company is growing very rapidly, you can expect a very healthy multiple. I suspect that this figure will compress when the downturn comes, but it need not collapse.
The other thing to note is the valuation goes up and down depending on the year of sale. Waiting longer generally means a lower value because it entails greater risks (i.e. the time value of money). On the other hand, the revenue number is growing so companies can expect higher basis for valuation as their revenue grows. But then we come back to the alignment of multiples with revenue growth. In this example, Year 3 is worth more than a sale in Year 2, but then valuation declines again in Year 4 with a decline in the valuation multiple. Of course, this is a very simplified model, but it illustrates the point.
These kinds of valuation fluctuations are the basis of my consulting practice. Management teams have a limited number of financial levers they can pull when plotting the financial future of their company. And many of those require years to play out. It helps to work through each scenario and each set of contingencies.
The key to all of this is to understand what metrics really matter in valuation. In the case of SaaS companies, it is revenue growth. For other companies, other metrics matter more.
Valuing New Customers
Let me take this a step further.
Let’s say our example company wants to expand its sales force. This is actually a very common, real-world scenario. As SaaS companies grow they start to face a real bottleneck around their sales channel. When a company is starting out, the CEO or some other founder can handle a lot of the sales work. Word of mouth and personal networks can get a company started, but eventually the company will need a sales force to expand beyond that. And not just a sales force, but channel partners are also very important for a company selling to enterprises. Someday, I will write a separate note exploring the various partners a company can work with, but for this exercise bear in mind that all these parties expect to be paid either commission or revenue share.
Companies then need to understand the Lifetime Value (LTV) of their customer. This is a common topic in SaaS and Internet companies. The basis of which is that it costs money to acquire new customers, and companies need to balance those costs against the gains they eventually bring.
Below, I have drawn up a model for how this works. A new customer would bring in no revenue in the first year, but requires $1 million in upfront acquisitions costs (e.g. commission to the salesperson), and some upfront investment (e.g. greater AWS usage). Revenue slowly builds in following years. In this example, the company is sharing 50% of revenue with its channel partners (that is at the high end of reasonable rates, but not unheard of for a young company). If the company signs 10 new customers, this works out to $40 million in cash the company needs in its first two years with these new customers.
Then I took these revenue figures and added them to the valuation model in the first section above. I will spare you the full model, below you can find just the valuation output. The table reads the same way, the metric we care about is PV in the year of sale. Here, I am valuing the company using revenue net of partner revenue share. The big difference in outcomes is that revenue growth rates are much higher, and push the company into a higher valuation bucket. As this demonstrates, the company ends up being worth considerably more.
We are reaching the limits of a simplified model. In reality, this new model has a lot more sensitivities to it, that are going to vary case by case. For instance, I am using a 9X revenue multiple to value the company, when in reality its growth rate is now high enough to qualify for a higher multiple. The salesman in me also would note that in many cases, companies can find a lot of ways to use a sales force to more fully penetrate existing customers. I know a world-class telco software salesman who can look at a customer list and identify what should be the multi-million dollar accounts. If this company is only selling $100k to those customers, he has a proven track record of ramping up sales, but his services do not come cheap.
On the other hand, this model excludes all kinds of things that go wrong. For one, I have not explicitly modeled churn. I am just assuming that customers keep growing, when in reality all accounts need to be maintained or they eventually decay. Another big issue is that young companies do not always know how long it takes to sign up a new account, making such models little more than a guess. Actual customer sales cycles can be much longer, upfront costs can be much higher. There is a lot of devil in these details.
Most importantly, the stress of radically ramping up growth can tax even the best management teams. As new customers growth becomes the focus other things get overlooked. Systems that worked for a few dozen small customers may start to fail under the strain of a few large accounts. The new customers will likely require new features, and those tend to get prioritized at the expense of existing customers.
This is a problem not just for enterprise SaaS companies, but for all companies that take venture funds. Growth becomes a priority, and sometimes it displaces other things that are vital to the company’s long-term health.
I ran a few scenarios on this model, many of which also showed high variations in valuation year by year. The outcomes are all over the map. So take this model as an illustration of the point. There is no hard and fast rule about how fast to build up a sales force, and it is possible for a company to kill itself by raising too much money and then deploying those funds inefficiently. Homejoy is one recent example of what can go wrong. Nonetheless, I think this is a valid framework for making these sorts of decisions. My guess is that many other Unicorns are under similar strains.
Going back to the original premise, a company looking to raise money today from venture investors needs to also factor in the impact of adding new capital. If management thinks they can meaningfully increase the growth rate of the company by spending capital today (e.g. hiring more sales people), then it often makes sense to take the dilution hit. The next step I would take in a real engagement would be to factor in the management team’s stake. By taking on venture investors, they are reducing their own stake in the company. Would you rather own 50% of a $100 million company or 30% of a $300 million company? What if the valuations are closer, and you then have to factor in the risk of being fired by your own board? These are advanced permutations, and I am happy to discuss them privately off-line.
As a sample, I compared the valuation of the two models above, one with venture money, the other without. The table below compares the two outcomes.
The focus of this table is the bottom two rows. Management has sold 18% of the company to outside investors. In reality, they will probably have to sell a more sizeable stake to raise this amount of money. As the model shows, the valuation of the company without venture funding is actually higher for the first two years, then the benefits of that new capital start to become more apparent. Not coincidentally, this is the amount of time it takes for the new sales force to start contributing. If the management team’s time horizon is five years, then taking venture money is an obvious choice. But those first two years are going to be tough. Every delay will be scrutinized and management held accountable for any missteps. There is a lot that can go wrong in that time frame, and a venture-backed board will be a real pressure on management teams.
If I had to pull one lesson out of this it would be think very hard about what to do with all that new capital. In this model, the capital is used to bring on new customers. So management teams will really want to know how long it takes before those new customers start to contribute. A common theme in downturns is to look back at all the silly things a company spent money on when times were good.
Dream Big? Start planning from the end and work backwards
Recently, I have been speaking to two companies that are facing a very similar dilemma. Neither has raised much venture money. Both have offers on the table to be acquired by large companies. In both cases, it would be a large payout for the executive teams. But neither CEO is in it for the money. Instead, they are dreaming big dreams. They are both in different parts of a fast-growing industry. And they have a reasonable chance of building major success stories, but that requires raising money now and incurring major risks in a nascent market.
In both cases, I advised them to start from the end and work backwards. How big can the company be in five years? To a company with $20 million in revenue, the idea of $100 million in sales five years from now is a major achievement. But from an outside viewpoint, that is still a small company. And no one knows what options will be available to a small company five years from now. On the other hand, maybe that is enough scale to do even bigger things. This is what makes technology investing and management so interesting.
Roll the dice on some future IPO or sell the company today?