Economics « Non-Linear VC

November 10, 2009 • 12:48 am 2

The math of SaaS revenue growth

A conversation with the CEO of a SaaS company today reminded me of the importance of the rule of 78s. What is this “rule”, you ask. If you run a recurring revenue business, it is the most important number you have never heard of.

Back to my conversation with the CEO. We were talking about the use of proceeds for the financing she is trying to raise. In her case, the business is break-even, but has the opportunity to grow into some oncoming market demand. So I asked a classic VC question; “Assuming you close the financing in Q4, 2009, what will your 2010 revenues be?” Simple question right?

Lets say that this company ended 2009 with $3 million in run-rate revenue and that the Company has two sales reps. Each sales rep. has a quota of $10,000 of incremental monthly recurring revenue (“MRR”) bookings per month. Sales reps. historically produce at 75% of quota in this company, so the incremental new bookings per month, per rep is $7,500. At its current level of sales and marketing resource and productivity, this business can expect to generate $1.17 million more in revenue in 2010 than it generated in ‘09. To get there, just multiple the $15k per month of incremental revenue the two sales reps. will generate by 78. Why 78? Because 78 is the ”sum of the digits” for revenue producing months in a year. Incremental revenue added in January will produce revenue for 12 months; incremental revenue added in February will produce revenue for 11 months; …; incremental revenue added in December will produce revenue for 1 month. The sum of digits for 12 is: 12 + 11 + 10 + … + 2 + 1 = 78. So the baseline revenue projection for 2010 should be about $4.17 million.

Now lets talk about the enhanced growth opportunity. The Company wants to hire two additional sales reps. who can be expected to produce at the same $7,500 of MRR as the two existing reps. For purposes of the example it will take a quarter to recruit and hire the two new reps and another quarter for them to build a sales pipeline given the sales cycle. Conservatively, these two new reps. won’t begin generating MRR bookings until June. Lets say there is a one-month install cycle so that June bookings convert to revenue in July. So what kind of incremental growth will these reps. generate in 2010? We don’t get to use the rule of 78 this time; no, because these reps. will produce for only six months during 2010, we use the sum of digits of 6, which is 21. That is a lot less than 78; that six month delay between making the decision to hire two new reps and getting them on board and productive really hurts. In 2010, the two new reps. will generate only $315k of incremental revenue. The business can be expected to generate $4.485 million of revenue in 2010; not meaningfully more than the base scenario without additional investment.

While this looks like meager growth, it doesn’t tell the whole story. Remember, this business came into the year with a $3 million run-rate. Without hiring the two new reps, the business should end 2010 with a run-rate of $5.16 million. By hiring the two new reps. that can be increased to $6.24 million, even though the two new reps. are productive for only six months. That incremental run-rate of over $1 million makes a huge difference and sure makes hiring the two new reps. look a whole lot more attractive than the meager incremental revenue they will generate in 2010.

Here are the key takeaways:

1) Get the Base Right: Recurring-revenue businesses are great because they are highly predictable. Applying the rule of 78s and with a little understanding of your sales resources and their productivity, you should be able to estimate your baseline next year revenue with a high degree of confidence.

2) Scaling Takes Time: The most common mistake I see recurring-revenue entrepreneurs make it to underestimate the time it takes for increased sales and marketing resource to impact the top-line. Hiring new sales people today probably won’t move the needle on your next twelve month revenue. More likely, an investment in sales and marketing won’t have meaningful impact until the following fiscal year.

3) Run-Rate Matters: You will see the impact of an investment in sales and marketing in run-rate much faster than in top-line GAAP revenues. If you are asked about what your revenues will be next year, answer the question directly, but also include a description of the difference in the run-rate you expect to end the year with under the no-growth and growth scenarios. The run-rate difference will impress much more than the top-line difference.

By the way, the CEO of the SaaS company I was speaking with got this analysis dead-right. Well done!

Filed under: Economics, Lessons Learned, SaaS, Venture Capital , recurring revenue, rule of 78

August 5, 2009 • 11:24 am 1

Cognitive Dissonance: Are you a technology or a service?

One of the trends I’ve observed over the past several years is that more and more technology entrepreneurs are starting service-delivery business. By services businesses, I’m referring to the category of businesses that some venture investors refer to as technology-enabled services (“TES”). We at Meritage prefer the term network-enabled services (“NES”), which we think more accurately demonstrates the fundamental innovation in the business model, which is that there is a high-level of connectivity between the service delivery platform and the customer. Services is a big tent, so to ground it, put your mind on business models like SaaS, cloud computing, and even search.

Many casually refer to these businesses as “technology companies”, by which most intend Wikipedia’s definition of “information technology”, not the generic definition of “technology“. After all, these businesses create technology to build their service, right? Salesforce.com is a technology company; Yahoo! is a technology company; Akamai is a technology company …; AT&T is a technology company. Whoa! Hold it right there; rightous indignation time; I can hear it already.

AT&T is not a technology company! They don’t know the first thing about technology! AT&T is a communications services business.

So what is the difference? Why do we think of Salesforce.com as a technology company and AT&T as a service provider.

Cognitive Dissonance

Objectively, either all the companies in the list above are technology companies or none of them are. AT&T uses every bit as much technology to run its business as does Salesforce.com. And while some of that technology is different, much more of it is fundamentally the same. More importantly, the technology serves the same singular purpose; it powers the service.

So which is it? Are they all technology companies or are none of them technology companies?

Brace yourself; here comes the controversial statement. If you are a services business, you cannot be technology company. Salesforce.com is not a technology company; Akamai is not a technology company. They are service companies.

To Create of Consume

I’ll return to where I started this post; with the observation that I see many technology entrepreneurs moving into services businesses. Often times, such entrepreneurs bring with them a technology creator mindset, not a technology consumer mindset. And as a result, they are susceptible to a number of mistakes. The most common of these is when the entrepreneur convinces themselves that their proprietary technology is going to make there service provider successful. Here are three examples:

Wireless service provider that has invented a new technology for mesh wireless networking, enabling quality of service management for voice over wireless applications.
Tech-enabled service provider for eBay resellers that has invented a new web server technology that will make the technology infrastructure highly scalable.
SaaS developer who built its own platform for managing its application in the cloud.

What is the problem with these businesses? They are acting like creators of technology, not consumers of technology.

The rationale is that by creating a proprietary technology, the service provider can gain a competitive advantage over its service provider competitors. It is an alluring and dangerous trap. But fundamentally, there is no way the wireless service provider above can keep up with the R&D budgets of Motorolla, Qualcomm, Cisco and the like and create a best of breed mesh QoS technology in the confines of a single wireless service provider. The tech-enabled service provider can’t possibly out-compete the web server technologies on the market in the long-term.

Embedding a proprietary technology business inside a service provider decreases execution focus and greatly increases execution risk. Creating this kind of technology requires a different skill-set, different processes, and a broad market in which to distribute the product. The economics of spreading the costs of creating and maintaining a technology company like this over one, captive service provider don’t work. From an investors perspective, the Company will consume more capital than is required, trying to capture/maintain a lead in a core technology that creates only the illusion of competitive advantage for the service provider. It is hard enough to build a successful service provider. Why would you want to complicate it by having to simultaneously execute the creation and maintenance of a core technology you could purchase from a third-party vendor. And what if the science upon which the new technology is predicated upon doesn’t pan out. Then what?

Rather than invent and maintain truly proprietary core technologies, I encourage all of my service provider investments to be rational and ruthless consumers of technology, not creators of technology. Find best of breed wherever you can, configure it in a way that suits the needs of your particular service and drive your vendors, who have the R&D budgets and skills, to innovate on your behalf. It is much more capital efficient and the execution risk is significantly lower.

What I am not saying

Does this mean services businesses should not be inventive? I’m not saying that at all. In fact, the successful services businesses are highly inventive, but not at a core technology level; rather at a business process level. One of my portfolio companies, Pipeline Trading has developed some amazing algorithms around large block and algorithmic trading of equity securities. This invention is unique to Pipeline and highly proprietary. But it is not a “technology” in the information technology sense of the term, it is a business process. Sure these innovations are expressed through technology; lots of software to be precise. But that doesn’t make the innovation a technology innovation; quite the contrary. Any quality engineer who is given the advantage of understanding the business process expressed by the algorithms could easily write the software. The innovation is not the software itself, it is the algorithm; and that algorithm has fundamentally nothing to do with information technology. It turns out that the competitive advantage is also in the algorithm, not in the software that expresses it.

But, but but…

But what if we really have created an incredible technology (in the information technology sense of the term) in the process of building our service provider? What if I really do have the best cloud-enablement platform in the world. Great, you just hit the jackpot, because you’ve discovered and solved a problem that every other service provider like you will also experience. But don’t keep that innovation captive to your own little service provider; free it. Spin the business off; create a hardware, software or new service provider around the technology; open source it. In fact, this is exactly what the SaaS operator with the cloud-enablement platform did. Kudos; right call!

The highest and best use of a new technology innovation is to sell/license it to anyone and everyone who could benefit from accessing it. That includes the service providers with which you compete. If it is amazing and valuable new technology, you should be able to sell the technology to your competitors and a lot more customers. If you can’t monetize it directly, ask yourself again why you believe the technology innovation will be a source of advantage for your captive service provider?

Final thought

Where does this leave us? Well, if you are a technology entrepreneur coming into a services business, I’d suggest you first change your mindset from one of being a technology creator to a technology consumer. Whenever possible, focus on innovating at the business process level, not the core technology level. Avoid technology “research” projects and focus resources on expressing your innovative business process with software. And finally, if you are absolutely forced to solve a tough core technology problem and create a truly proprietary technology in the process, look for ways to free it from the shackles of your captive service provider. In the end, the technology will have a better chance to flourish and your service provider will be unburdened from the costs, and challenges of having to execute on two fronts simultaneously.

Don’t confuse the services business with a technology business. You can’t; and shouldn’t try to be both.

Filed under: Economics, Lessons Learned, Risk, Venture Capital , capital efficiency, cloud computing, compeititve advantage, execution, SaaS, technology

July 30, 2009 • 10:37 am 1

Financing a Services Business: The Valley of Death

It is a late afternoon ritual for me to read the Meritage Minute, a daily briefing on key news events published by my colleague Heidi Longaberger. If you would like to receive The Minute, email Heidi at: hlongaberger@meritagefunds.com. Yesterday’s briefing included an AlwaysOn piece titled “In Ten Years Will All Apps Be in the Cloud?” I’m not here to debate the merit of the piece; ”all” is a strong word despite the fact that directionally, movement toward the cloud is inevitable. The article lists a series of challenges of moving into the cloud, most of which are the same FUD often touted by those that don’t quite understand. And of course there is the compulsory Twitter data breach mention. Despite that, the last paragraph contains an important nugget of wisdom; a challenge all services businesses (including those in the cloud) face which has nothing to do with technology, scalability, security or compliance. Here it is:

Return on Investment. From a VC standpoint, it takes a long time to run a subscription business past the Valley of Death. Time to revenue for a cloud business is very long, because revenue comes in a stream, not a lump. Driving sales has to be very swift and very focused.

This is essentially a financing challenge. It is all too often overlooked by entrepreneurs. Services businesses can be capital efficient, but are not necessarily so. To understand this fully, we have to look at both the cost and revenue-sides of the business.

Build it Before They Come

In the services business, you have to build the service before you serve the first customer; not just the product, the entire service. The product analogy is the R&D that goes into hardware/equipment design or into software development in the software business. But in the services business, the stakes are higher, you also have to create a customer service infrastructure, billing and operating support systems, and service delivery infrastructure to wrap around your offering. These costs are ongoing, they are not one-time and sunk. Worse yet, all that infrastructure has to be supported by staff. Early in the businesses’ development, the services operator is by definition operating at well less than minimum efficient scale, so these up-front investments can consume significant amounts of capital and look highly inefficient on paper. Simply “outsourcing” these functions is not the solution. You are still running at sub-scale so whoever you outsource to is going to charge you rates that reflect your scale.

The point is that the services business has a significant minimum cost structure that is required to deliver on a minimum required level of customer support, quality of service, uptime, sla management, etc. Relative to other businesses, the cost structure is high fixed operating cost, low marginal cost per incremental customer. Now lets look at the revenue side of the equation.

Delayed Revenue

One of the appealing elements of the service delivery model for customers is that they get to swap up-front CapEx (think hardware and software purchases) for a delayed, consumption-based payment stream. This is well and good for the customer, but has some significant implications for the services operator. The first and most obvious issue is that the receipt of cash is delayed; that is challenge enough.

But what happens when you miss your bookings of incremental monthly recurring revenue (MRR)?

The Rule of 78s

Those of us who have been around services business have internalized a concept called the rule of 78s. Lets say you project your company will book $50k of new MRR during your first twelve months of revenue generation. The rule of 78s say that the company will generate a total of $3.9 million ($50k * 78) of revenue over that year. [Note: 78 is simply the total of the revenue months during the year; 12+11+10+...+1] But what happens if the company does only $25k of MRR bookings. Now the first year revenue is only $1.95 million. The company just missed its revenue number by$1 million and that miss is probably all incremental burn because of the high fixed cost, low marginal cost of operating the business. I’ve taken a bit of a liberty (albeit a small one) on the cost structure assumptions here, but you get the point.

Run-Rate Matters

Now lets look at where the company stands at the end of that first year of revenue generation. If the business hits the $50k MRR bookings number, the end of year run-rate revenue is $600k. But if the Company does only half of the projection, the end-of-year run-rate revenue also drops by half; to $300k. Again, the difference, in this case $300k per month, is probably all incremental burn, because of the cost structure of the business. So not only has the company burned $1 million more than anticipated in the first year, but the business is also burning $300k per month more than projected at the end of year. If your fixed operating costs are $600k per month, the business is 2 months away from covering fixed operating expenses under the $50k MRR bookings scenario, but 12 months away under the $25k MRR bookings scenario; assuming no churn. This gap between fixed costs and recurring revenues is the Valley of Death referenced by the AlwaysOn author. Further delays in revenue acquisition are going to exacerbate this problem, particularly if your fixed operating costs are higher than $600k. MRR bookings misses don’t just impact the month of the miss, but reverberate though every subsequent month.

Skeptical not Cynical

You can understand why investors are weary of investing in services-delivery model businesses early in their life-cycle. And so far, we’ve only addressed the costs and revenue economics of the business at an income statement level. We haven’t even begun to address the uncertainties around customer profitability, including the lifetime value of a customer, which makes investors really queasy. Profitless prosperity (think Vonage), is a scary concept to investors and without operating history to prove out metrics around customer acquisition costs, ARPU, churn, renewal rates, costs to serve, there is a bunch of guesswork to be done, educated as it may be. I’ll address some of the issues around customer lifetime value in future posts.

Don’t fret, all is not lost. Some of this, particularly on the cost side can be managed. Better yet, there are actually investors who prefer the services model (I happen to be one of them) and understand how to build services businesses. Having uncovered some of the services businesses flaws, I’ll talk about why I prefer it in future posts.

Filed under: Cloud, Economics, SaaS, Venture Capital , capital efficiency, cost structure, financing strategy, lifetime value of a customer, recurring revenue, services

June 2, 2009 • 11:48 am 1

The economics of on-demand services

A post by Ben Kepes over at CloudAve got me thinking about the economics of on-demand services, so I thought I’d do a quick blog. On-demand services businesses come in all shapes and sizes. This is particularly true today with the emergence of SaaS, where the vendor diversity is staggering. Despite the diversity of services – ranging from traditional communications services (think voice, video and data) to highly verticalized SaaS applications (think point of sale applications for yoga studios) – the fundamental economic building blocks of these businesses are, for the most part, the same.

You wouldn’t know it though based on the business plans I review and the discussions I have with entrepreneurs. Generally speaking, the CEOs of traditional communications services businesses that I speak with have a much more mature understanding of the core economics of services businesses than most SaaS executives. I attribute this to the fact that most SaaS executives I see come out of the software space, not the “services” space. And in the services business, there are fundamental economic concepts that simply put, don’t exist in the software business. So if you are a SaaS, PaaS or any on-demand executive whose business plan gets far enough down the Meritage Funds pipeline to have a hard-core conversation about the business’ economic characteristics, here are the things I will want to talk about.

Customer Acquisition Costs: It all starts with acquiring a customer. What are the marketing channels you will use, online, direct, traditional, sales channels, etc.? In each of these channels, what is the fully-loaded cost for you to close and implement a new customer? Do you have hard implementation costs; equipment, installation, setup? If the answer is that you expect customers to find you, go back to the drawing board. You will spend money to acquire customers, particularly if you hope to scale in any meaningful way.

ARPU: ARPU (average revenue per unit) tells me a ton about your business; the first thing being what you think the appropriate unit of measure is for your go to market strategy. So what is a unit? In a communications model, it is typically a “subscriber”. In a SaaS model it might be a seat or in an enterprise or smb model it might be the enterprise as a whole or the smb. Point being, it really depends on the level at which you are selling and where the purchasing decision is made at your customer. ARPU tells me how much revenue you can generate per month out of a customer as you define it.

Price Degradation: With ARPU in mind, I also want to know about how you expect the price of your service to degrade over time. My experience is that prices for services tend to go down, not up, because as a service moves into a mass market phase from an early adopter phase, the value proposition is naturally diluted. The benefits that accrue to the mass market for adopting your service are less than those that accrue to the early adopters. If this were not the case, the mass market would be early adopters. If you believe that the value and therefore the price of your service goes up over time, be prepared to justify it. Some opportunities have natural monopoly characteristics where the addition of each customer adds value for all the prior customers. If this is the case with your business articulate clearly why this is the case.

Up-sell: Offsetting the natural price degradation in services, tell me about other services that you intend to offer in the future that will provide you with some ARPU lift. Service segmentation is critical, because this enables you to optimize the revenue you generate from your customer base. Be sure to not give functionality away in your base service that you may be able to charge a large percentage of your customer base for later.

Variable Costs: I want to know about all of the variable costs of your business on a monthly basis. These typically include commissions, infrastructure and other service delivery costs, customer support costs, billing expenses. The common thread being that each of these scales up with the number of customers you serve.

Churn: When you sign a customer, how long to you expect to keep them? What does the churn pattern look like; does most of your churn take place in months 1-3 or before the customer ever gets fully implemented? Once the customer is past the initial adoption phase, what percentage of your customers do you lose every month?

With these building blocks, I can make some key observations about your business and its potential to generate big profits. First, with ARPU, price degradation, and up-sell on the revenue side and your variable costs, I can determine your customer level contribution margin, the amount of monthly net cash you generate from a customer unit. If your customer level contribution margin is negative, no level of customer volume can save you. Some refer to this as profitless prosperity; you can sign customers, but never make a profit.

The second observation I can make is related, but less obvious; I can determine the lifetime value of your customer. By running the contribution margin through your churn calculations, I can estimate the lifetime contribution margin of a customer. This number must more than offset your customer acquisition costs if your business is to be profitable in the long-term.

From here, the questions all become about scale; how many customers can you sign, over what period of time and how many are required to cover the fixed costs of running your business.

Don’t worry, I won’t push you on these economics in a first meeting, unless your business plan is so clear that we can jump right into the economics. But rest assured, we’ll get into them eventually. If your business can’t prove out on these metrics, I probably won’t invest and you probably won’t want to spend the next five to seven years of your life trying to build a business that can never get to the finish line. Some may argue that these metrics apply only to enterprise, smb and related segments and that they don’t apply to consumer web 2.0. Unfortunately in the end, I think eBay’s experience with Skype and the poor track record of monetizing web 2.0 and social media are on my side of the argument in the long-term.

Filed under: Economics, Lessons Learned , ARPU, churn, customer acquisition cost, on-demand, PaaS, SaaS, variable costs

Non-Linear VC

The math of SaaS revenue growth

Financing a Services Business: The Valley of Death

Build it Before They Come

Delayed Revenue

The Rule of 78s

Run-Rate Matters

Skeptical not Cynical

The economics of on-demand services

General Info

Categories

Recent Posts

Archives

My Twitter Feed

Blogroll