Softletter100 Key Metrics


Review Each Key Metric from the Tabs Below

Cost of Sales

SaaS companies provide is a software enabled service, mainly delivered over the Internet. Therefore, the items that typically comprise COGS include:


  • Application hosting and monitoring costs.
  • Customer support and account management costs.
  • Data communication expenses.
  • OEM license fees for products embedded in the application.
  • Website development and support costs.
  • Professional services and training personnel costs.
  • Costs of subscriptions.


Typically, a well run SaaS company enjoys gross margins of  between 80-90% (95% is not unheard of). In this scenario, Cost of Sales, also often referred to as Cost of Revenue, should range between around 10-20% of the total revenue.


This list of expenses are not considered part of operating expenses.

Days Sales Outstanding (DSO) also known as the ratio of receivables to total sales, used to be a stable measurement in software, but as the numbers show, there's a great deal of volatility to be found in this once sedate metric.

DSO is designed to enable companies to calculate their average collection period (colloquially known as "getting paid") and see how well they're managing receivables. A DSO figure is an index of the relationship between outstanding receivables and credit account sales achieved over a given period. In our reports, we use the following formula:

DSO ratio = (accounts receivable / annual revenue) * 365 days

A low DSO number is usually better than a high one, though an overly low number can indicate Scrooge McDuck has got control of your books) because you're feeding cash to your piggy bank more quickly, making it smile. High DSO numbers can indicate:

  • Your customer base has credit problems.
  • You sales force is buying market share by extending very generous credit and payment terms.
  • You simply aren't tracking this key metric and have no idea of the state of your receivables (or, in some cases, don't want to know).

An increase in DSO can result in cash flow problems, and may require your company increasing its bad debt reserve.

One question that arose in the early days of SaaS was DSO going to continue to remain a relevant benchmark? After all, why should subscription companies have to worry about DSO? Don't you pay in advance for access to an application? And don't you immediately lose access to it if you don't?

The answer is "only sometimes." In large enterprise sales, companies can and do offer deferred payments in return for signatures on bottom lines. Large customers may face financial difficulties and negotiate easier reimbursement. And companies who want to "buy" market share often offer extended and delayed payment.

What's an ideal DSO number? We think if you're averaging 45 days, you're operating at peak performance. As the full numbers demonstrate, it's a hard mark to hit.

We have to be honest; this is not our most exiting benchmark but it does offer insights into an industry sector ’s operational efficiency. G&A expenses normally includes the salaries and staffs of the upper management, rent, connectivity costs, etc. These costs tend to grow in good times and during a company’s startup phase and tend to be stable over time. Rapidly rising or high G&A during economic good times or periods of stability are often a sign of internal turbulence; during a recession, they can reflect internal inefficiency if a company can’t keep expenses under control.


The best way to shrink G&A ratios is to be big. Large companies bring volume purchasing and local political clout to their negotiations, both of which can be important in bring G&A down. Tax abatements, special access to retail properties, waving of onerous zoning restrictions, deals on power purchases and similar perks can all assist in driving down expenditures.


G&A expenditures can be difficult to shrink because the price of power, office space, supplies, and related items tend to invariably rise over time, but during bad times, it’s a poor negotiator who can’t drive some costs, such as rent, down. Some costs may be beyond company control because legal and accounting fees fall into this category; external events (such as changes in regulations) can have a major effect on G&A.


In the past some firms combined G&A with S&M in their financial reports. These figures were then split out, but none of the current Benchmark 53 package up their 10-Ks in this deceptive way.

The Softletter Benchmark 53 are all publicly-held companies and most of our basic business metrics are derived from documents these companies are legally required to provide to investors. In most cases these are 10-Ks and equivalents, supplemented by expert analysis of the numbers.


Operating Income (OI) is a derivative metric designed to measure how efficiently a company runs it operations. It is calculated by subtracting EBIT (Earnings Before Interest and Taxes) from revenue.  The numbers are most  valuable when comparing companies of a similar type or peer group, something the Benchmark 53 is designed to do.


OI is frequently used by investment analysts to evaluate a company’s operating performance without regard to interest expenses or tax rates, two variables that can vary widely from company to company and even within industry sectors.


It is also important to note that some industries have higher labor or materials costs than others.  In the case of software, labor costs are almost without exception the single highest operating cost. This is why comparing operating income or operating margins is generally most meaningful among companies within the same industry, and the definition of a "high" or "low" ratio should be made within this context.


It is quite possible for OI numbers to be negative. The most common reasons are:


  • A company is generating high levels of sales and marketing expenses as it attempts to build market share and dominant position vis a vis its competition.
  • Unexpected expenses, such as a product recall, need to purchase new equipment, or rapidly expand a product development effort, hit your bottom line.
  • Your business model undergoes rapid disruption.
  • You over hire.

It's always important to remember that in SaaS and mobile, your available cash reserves may greatly exceed your recognizable revenue. Thus poor OI numbers over a year or even two may not be as significant as they appear. However, if the continue over time, you're burning cash. But even this is not necessarily a harbinger of doom if you can rely on public markets to replenish your business operations (think Amazon). But only a handful of companies are that lucky. Normally, investor impatience lowers the hammer on management and your business well before almost a quarter of a century goes by.


A point to note is that if you love the color red (represented by brackets in this report), OI is your metric! Negative numbers are common in contested segments, particularly in SaaS, where companies turn to public markets to fund war chests for long-term plays to achieve market dominance. There's recently been a fair amount of criticism aimed at this strategy (see Softletter editor Rick Chapman's review of Disrupted by Dan Lyons), but shareholders in firms such as, which has never been profitable since it went public, have made out handsomely (if you bought and held onto your original shares).


Softletter100 subscribers will note the strong correlation between OI and OIE. If you're not profitable on an overall basis, the efficiency of your workforce cannot make up for this

Operating income per employee (OIPE), often also called “revenue per employee,” and “net revenue per employee,” is a percentage ratio that is calculated by dividing a firm’s revenues by its current employee headcount. A high OIPE indicates high productivity and maximum use of a firm’s resources. In many ways, it’s a far more effective measure of economic health than revenue per employee, and both are benchmarks upper management needs to monitor closely.


As with operating income, the number is best used when measuring firms operating in similar industries and markets.

Factors that can impact OIPE include:


  • High employee turnover. It takes time to qualify, interview, train, and deploy new employees. While this process goes forward, your personnel will be less productive and your per employee numbers will suffer.
  • High employee hiring. As with firings, it takes time to bring new employees up to speed and productivity. But a company that’s growing rapidly is often enjoying strong economic growth so care should be taken to analyze the actual situation before passing judgement. However, history shows many, many examples of firm’s the hired well ahead of projected growth that never materialized.
  • Successfully (or unsuccessfully) leveraging your firm’s fixed assets. SaaS companies should be masters of this. Once an operating infrastructure for a system is in place, additional subscriptions should be added to the company’s revenues at minimal cost. That’s the theory, at least. In 2008, Twitter demonstrated that didn’t have to be the case as the online “microphone” suffered numerous outages and server crashes because of poor policies by upper management (which in led to the ouster of CEO and co-founder, Jack Dorsey).
  • CFOs love the impact a layoff has on OIPE because their spreadsheets show an immediate juicing of this metric. But, as we all know, layoffs are also indicative of over expansion and failed company initiatives. Fortunately for the CFOs, they’re usually not included in the typical company downsizing.

As with operating income, it’s quite possible to have negative numbers. In fact, as the Benchmark 53 demonstrate, it can be quite common.

R&D tends to be a stable metric, except when the industry is living through interesting times, as it is now. In the space of nine years, beginning in 2007,  SaaS and mobile apps completely disrupted a development, infrastructure, and distribution system that had been in place since the late 70s. In a placid environment, R&D as a percentage of revenue generally hovers between 12% to 18% of total revenue. (But don't be fooled by the executive summary. Within the categories there are some startling differences.)


But in an industry undergoing disruption, the numbers can gyrate wildly, as a look at the complete Softletter 100 Core Metrics  makes clear. Numbers in Social Networking and Mobile B2B and B2C are all over the place. For a sense of the good old days, snuggle on up to On Premise with a plushie, some nice herbal tea, and perhaps watch an episode of Family Ties on Nickelodeon.


Why do R&D numbers vary so widely in these segments? For several reasons. These include:


  • A company's core product or service is under attack and it's trying to code its way out of the mess.
  • A market segment is opening up and the firm's geeks are attempting to out-innovate the competition and build overwhelming market share by filling the feature tick lists ahead of everyone else.
  • The original product the company launched into the market has been found out to be a grievous heap of sh...errr...dung and must be fixed immediately.
  • Someone in upper management is bored, has a brilliant idea, and the R&D staff has been taxed with creating the next entrepreneurial miracle.

There are other reasons, most of which are usually related to the above.

Return on equity (ROE) is simply a company's net income revenues after expenses are  deducted) divided by shareholders' equity in the company. The equity is calculated by adding up all the assets and deducting all the liabilities, including debt. “Good” ROEs typically range between of 15% to 20%.


ROE informs you how much profit a company earned in comparison to the total amount of shareholder equity found on firm’s balance sheet. Shareholder equity is equal to total company assets minus total liabilities (interest, tax, operational costs). It's the shareholder’s “piece of the company." Shareholder equity is an accounting creation that is supposed to represent the assets created by the retained earnings of the business and the paid-in capital of the owners.


Softletter's Benchmark 53 reports enable readers to compare their own company's performance with that of peers and competitors. But note that a company heavily in debt or with a large write-down could easily have a high ROE because its equity has shrunk. Also,  companies in financial difficulty will generate some anomalous figures: a large net loss can generate a large negative ROE, while net losses divided by negative equity will generate positive values. Because the Benchmark 53 arranges sector results by medians, these outliers have a minimal effect on the larger picture.


It is very possible, as the numbers show, for shareholder equity to be negative. This has become increasingly common in high tech, where companies often burn through equity at high speed, often via expensive sales and marketing operations, to build market share, then return to the markets for more capital to burn. Amazon and Salesforce are two classic examples of companies executing on this model (and note Salesforce’s negative ROE of -7%). Many observers have commented at great length on Amazon’s inability to turn a profit, but the company has been able to generate cash and rely on new investment to build an online ecommerce engine that’s currently driving WalMart executives to drink and draining America’s mall of dollars and shoppers. If Amazon is riding a bubble, it’s been able to do it for close to a quarter of a century.

The Softletter Benchmark 53 are all publicly-held companies and most of our basic business metrics are derived from documents these companies are legally required to provide to investors. In most cases these are 10-Ks and equivalents, supplemented by expert analysis of the numbers.


Revenue per Employee is one of the most popular financial metrics by which to measure a software company’s performance. RPE is indeed a useful productivity benchmark, but several caveats apply. One is that a healthy RPE does not necessarily translate to a dynamic, growing company. That’s why despite its relatively handsome RPE, Microsoft stock has not excited anyone in the last several years. Larger companies do tend to generate large RPE numbers; business does indeed benefit from scale, as the Big Four’s substantial numbers demonstrate. But when companies are growing rapidly (something that makes IPOs exciting and the hearts of investors beat faster), RPE can suffer greatly as revenue is devoted to hiring, R&D, and sales and marketing programs that drive growth.


Another point to remember is that a major layoff can temporarily boost RPE; however, this is usually only a temporary fix. But CFO’s love to point to the sugar of enhanced RPE numbers as proof of increased efficiency and health at a struggling company. Sometimes there’s a big crash at the end of the rush.


When analyzing RPE, the sweet spot you should be looking for is:


  • Is the company in an active, growing market segment?
  • Is the company outperforming its contemporaries in terms of RPE?
  • Has the company avoided laying off significant numbers of employees in the previous fiscal year?


If the answer is to the above is yes, RPE can be used as a significant indicator of growth and good management.

Sales and marketing expenditures are always an interesting benchmark to analyzed and the numbers can reveal a great deal about both an industry category and the challenges facing a particular company. Before we dive in, it's important to make a couple of key points. Publicly held companies, or companies planning on an IPO, typically spend more money on sales and marketing in an effort to satisfy the demands of their investors for the type of growth that drives up the firm's stock price. For privately held firms, expenditures between 15% to 30% are far more the norm.

This ties to another point. Many companies as they grow trade profitability for profit. The goal of this strategy is to seize a dominant position in an industry and then cash in. A recent famous example of this is Facebook. In the Social Network, the movie about the company's founding and rise to dominance, Mark Zuckerberg is seen deriding one of his co-founders for his efforts to sell web advertising on Facebook. But once Facebook had driven Friendster and MySpace into minor niches, the company immediately began to explore ways to monetize its massive user base, which it has done by converting your network into its network (remember, if you're not paying for a product you are the product) and charging you to access it (and serving up ads). Saleforce, NetSuite, and other firms have followed similar strategies, though the path to monetization often differs.

On the other hand, a high S&M ratio for a company in an established market or niche is frequently a sign of stress. Competitive pressure, business model disruption, and bad management are often the reasons for S&M numbers to start to rise in excess of historic levels.

Another observation. Many Softletter readers often ask us how software companies allocate sales vs. marketing expenditures in this report. Public companies are not required to break out these classes of expenditures in their 10-Ks, but in on-premise markets, traditionally for every dollar spent on S&M, $.80 was spent on sales (personnel, management, and operations) for $.20 spent on marketing (advertising, email, PR, SEO, etc.). However, this ratio is starting to change. For example, in the case of Angie's List, sale and marketing expenditure are approximately 55% sales, 45% marketing. We're seeing this reapportionment grow in the SaaS and mobile markets as more money is spent on market education and demand generation as opposed to traditional feet on the street sales programs.

Some companies also play games in an attempt to disguise their sales and marketing spend. For example, Groupon has broken out marketing separately on their 10-K and lumped sales in with G&A! We're not falling for that and have made an intelligent "adjustment" to their figures which we believe better reflects their true S&M ratio.