Perspectives of a VC on Key Software Operating Metrics in 2011
General Catalyst’s tagline is “entrepreneurs investing in entrepreneurs,” and partners of the active venture firm are well known for lending deep operating experience to their portfolio executives’ growing businesses. I’ve known Larry for years since we were both operating executives at various software companies. Here’s his take on the key operating metrics that executives need to be on top of to stimulate growth and profitability — as well as the metrics that drive valuation in today’s market:
Lauren: What key operating metrics do you track as an investor and board member at your portfolio companies?
Larry: The metrics we track depend a great deal on the stage of the company. We invest in many very early stage companies which are still building products, so the key metrics for those companies are around the overall cash burn,cost per employee, and relative costs by function. In a pre-revenue company we watch these metrics to ensure the company has the runway to get to its initial goals that are set for further funding.
In an early stage company with early customers, the key metrics to watch are around the cost of customer acquisition (coca), cost of sale, average deal size, implementation costs, and projected support costs. The key focus for this phase of the business is to develop a solid and repeatable sales model that can bring the company to profitability and growth. This phase in company development is often fraught with lots of change as the management team learns the intricacies of the fit/finish of the product and the true underlying economics of selling. A major mistake of this period is to anticipate too early what is repeatable only to pour cash into an unproven sales model.
In a company that has proven a sales model and market fit, the key measures are around the ratio of cost of customer acquisition to revenue. In an enterprise software license environment this is most accurately measured by looking at sales/marketing expense in relation to growth in revenue and bookings. The belief here is that since revenues are often lumpy and sales often direct, investors are willing to invest in sales and marketing early because it generally brings cash and predictable revenues (near term) to the business.
In a SaaS company, the key measure to track is that of MRR (monthly recurring revenue) and churn. MRR is most significant because it is the best predictor of overall success and long term growth and value. Churn is key because while many SaaS companies are good at on boarding many early customers, churn in the customer base basically erodes the value of the business as it is growing. It’s like building a sandcastle and realizing that the base keeps eroding the higher the sculpture.
Another important SaaS metric is to track the underlying cost of supporting an individual customer. In lightweight SaaS applications this may be trivial as much of the cost appears in cogs as expenses to Amazon or Rackspace. But increasingly SAAS is moving to more demanding applications and the cost of the underlying infrastructure (whether cloud, or self-managed) along with the costs of supporting customers becomes a significant factor in profitability and growth. The worst case is where there are major step functions in supporting customers due to poor product design or complexity.
Lauren: Yes, we’ve been tracking these SaaS metrics, such as customer acquisition cost and average cost to maintain a customer, along with about a dozen others, for 4 years now in our annual software benchmarking. By comparing your own numbers to close peers with a similar sales model, you can test whether you are as efficient as you might be or identify specific areas for improvement. How have you seen these metrics change or evolve as software business models have evolved?
Larry: The major change in the software business model is around the movement to SaaS for all types of software products, from collaboration to SFA to ERP. The SaaS business model is terrific in that it aligns customer interests with vendors (if you don’t serve me, I’ll cancel my subscription) but it is one that is complex and often costly for the company and its investors. In the licensed model, companies become profitable sooner as they are able to sell expensive perpetual licenses (sometimes that don’t get deployed); get cash up front, and generally don’t worry much about the customers’ deployment except when it affects maintenance fees. This model tends to favor and support an expensive direct sales force that can convince customers of the utility of the software and work toward a major transaction that makes the customer profitable instantly, often for the long term.
In SaaS, the vendor takes on much of the risk of not only acquiring customers but keeping them satisfied in an ongoing way. Also, SaaS companies provide more than software—the whole stack of infrastructure software, hardware, SLA, etc. So the vendor has a requirement to ensure that customers that come on the platform stay because they generally do not get large upfront commitments, but their fixed costs to build the product and sell it are high. So SaaS companies often require significantly more capital to build before they become profitable (often 2x previously) and this leads to a focus on managing sales/marketing costs with a more granular focus.
For example, it is now the basic trend of SaaS companies to focus on an inbound selling model that uses low cost internet techniques (SEO, blogging, SEM) to build a customer pipeline and to then use an inside sales model to nurture and close customers. SaaS makes this possible because the vendor takes on almost all deployment responsibility remotely so they can “own” the deployment in a way licensed vendors often can’t.
Management teams and board members therefore track the customer funnel through stages of qualification, conversion, acquisition, and deployment with an eagle eye to get the funnel into the proper “shape” where the economics and predictability are good. The best companies, at maturity, have an internal language around these funnel metrics and it provides a good way for the company to manage functions. The core metrics are around COCA (cost of customer acquisition) and LTV (life time value) but there are others around time to profitability, growth within customers, etc.
One of my favorite metrics, developed by my friend Rory O’Driscoll at Scale Ventures, is called the “Magic Number”. Rory was on my board at NetGenesis and on the board of Omniture where the Magic Number was championed and is an expert at looking at sales expense and revenue growth. He developed the Magic Number as a concept to capture how a monthly increase in marketing and sales expense will predictably affect increase in MRR. It’s a way to determine if a company really has a predictable model in place and helps an investor decide when it’s the right time to “turn on the gas” for company growth…..
I would provide some caution that not all SaaS applications fit the same low cost sales model. Sometimes there is a sort of mindlessness today that investors want to see companies look like robots that have a product, and an internet sales/marketing engine that scales without human involvement. This sometimes works for lightweight or lookalike apps but tends to make all aspects of the business simple, even where it warrants sophistication or complexity. Some of the best SaaS companies now have solid and growing businesses that sell strategic software, have direct sales organizations that are required to partner with customers, and are able to get long term commits and upfront cash from customers to balance the customer/vendor partnership. The SaaS subscription and delivery model does not dictate the type of value or customer approach that works. I see too many companies trying to shoehorn their business into an inside, low cost sales model that may never work for the value they are providing.
Lauren: I agree and I think that the type of sales model that is right for your company will drive a number of the other financial and operating ratios as well so if you benchmark, you want to compare your numbers against other companies with similar sales structures. We separate our benchmarks for software companies with low average deal sizes/low cost sales models versus companies with high average deal sizes/more expensive sales models so that if I’m selling a high value/high sales expense product, I can compare my operating ratios to companies with a similar model rather than to companies with a lower sales expense model.
Which numbers do you track as key leverage points to ratchet up growth? How do you measure whether sales and marketing investments are getting efficient “bang for the buck” in customer and revenue growth that you and company executives want to see with venture money?
Larry: The numbers that we like to track before ratcheting up growth are COCA and churn which gives the best perspective on overall LTV. In some businesses, though, MRR growth within customer is really significant as it shows how customers can often be a “channel” for a vendor’s products (either internally or though their own channels). Also some of the newer SaaS companies sell indirectly through partners, so tracking the metrics around number of partners, time to get them productive, etc. gives us insight into further investment.
The best way to ensure the efficiency of sales and marketing spend is to ruthlessly track the pipeline metrics that I discussed earlier. With inbound marketing techniques now it’s possible to quantify the cost of leads, track their conversion as sales expense, and determine how that investment translates into customer revenue and bookings. The internet has taken a lot of the hand waving out of sales and marketing and the truly great SaaS companies have marketers who act like engineers as they adjust and refine their funnel techniques.
One of the great things about what’s happening today is that the cloud is now proving out the “pay by the drink” model of purchasing IT infrastructure. The result is that it’s less expensive and problematic to build a scalable SaaS company. Five years ago, the SaaS companies we backed almost all were building out their own infrastructure at a managed service or hosting facility. This was capital intensive and difficult. With the advent of Amazon and Rackspace these same companies now are buying this infrastructure in both a secure and scalable way. Now this does not universally apply, and some high end SaaS companies need to own their own infrastructure, but for the bulk of applications the cloud works and it means a direct increase in innovation, economy, and benefit for customers. Whereas the earlier SaaS companies often took $50m plus to get to profitability, the newer SaaS companies can do this for half that amount. There is still lots of complexity and subtlety around cloud pricing, lock-in concerns, security, etc, but the overall trend is inexorable. Also, with the growing number of companies and applications moving to SaaS there is now a growing ecosystem of component, integration, and service providers that round out the cloud infrastructure itself to make building robust products easier and more reliable. And some of these companies look to be great investments…..
Lauren: We definitely have been seeing through the benchmarking that venture-backed software companies are on average getting to profitability earlier than they were five years ago, I’m sure partially due to the lower cost of building out infrastructure and using other cloud services. What’s your outlook for metrics in 2011?
Larry: I think 2011 will be a great year for SaaS as there are a number of very prominent and varied SaaS companies that are likely to IPO. This will show both the variety of companies but also reveal some of the complexity for analysts following companies. For example, only recently have there been FASB rule changes that make it easier to match deployment service revenues directly with SaaS revenues so investors can really see how expenses and revenues match in a defined period. SaaS companies are great for public investors as good SaaS companies that have strong, growingMRR often have the next 4-5 quarters of revenue totally baked so their chance of blowing a near term quarter is minimized. The SaaS business model, at growth and maturity, is a very profitable model, and analysts like to look at core free cash flow from these companies as the ultimate measure of their value. While this certainly is true, we must not forget that SaaS companies still go through periodic investment cycles around new products, international expansion, etc. so a rigid cash flow model can be undermined by the realities of technology growth requirements. This brings us back to the core metrics for SaaS: COCA, MRR, LTV, and COGS associated with the service.
Lauren: Larry, I absolutely agree with you; the relatively shallow data that you get from pubic company filings really isn’t sufficient to assess the true value of a SaaS company, which is why we are focusing our 2011 benchmarking (launching early March) on exactly the metrics that you mentioned here. Thank you.