SandHill.com

The Bottom Line on SaaS

As vendors scramble to provide on-demand offerings, an in-depth analysis of the numbers behind one SaaS business show a superior economic model.

By Scott Bolick, George Gilbert and Rahul Sood, Tech Strategy Partners

Jul. 06, 2006

Software-as-a-Service (SaaS) is a top-of-mind consideration for most software company leaders today. The idea is hot but the reality is challenging (as Ray Lane wrote earlier this year ) leaving many vendors on the sidelines. Although there are several intuitive arguments in favor of the model, there is still disagreement on how and why the SaaS model is superior to the traditional perpetual licensing model.

Few thorough analyses are available on the long-term impact of the SaaS business model. After a careful examination, it was determined that the best way to delve into the strength of the SaaS economic model would be to use SalesForce.com's (SFDC) quarterly figures and analyze them to uncover the company's true profitability.

The numbers told a compelling story: SaaS is not just a superior economic model, but it also has several strategic advantages over the traditional licensing model. Normalizing SFDC's income statement expands operating margins to an implied 33 to 39 percent, up from a currently reported 6 percent. SaaS is economically more efficient for the customers too, as it minimizes their spending on IT infrastructure and services. SaaS also allows vendors to shrink their product innovation cycles and drive innovation across a wider cross section of their customers than the traditional model.

A Closer Look: Normalizing the SaaS Model
Most vendors agree: the biggest advantage of the SaaS model is its insulation from large, lumpy sales deals and the better predictability of the annuity revenue stream. Another big advantage is the resilience of the model vis-à-vis changes in the capital spending of customers, since customers can account for their spending as operational expense.

On the negative side, there are concerns over several aspects of the business model, especially since the leading SaaS vendors have profitability of 3 to 6 percent as opposed to traditional vendors who earn 25 to 35 percent of their revenues as operating profits.

But there are a few reasons for this discrepancy. First, revenue growth is slower at SaaS companies because it is no longer front-loaded by perpetual licenses. Second, gross margins are perceived to be low because SaaS is compared to legacy ASP models which have to factor in the costs associated with the customer- "silo"-ed data center required for application hosting and service delivery. Finally, sales and marketing costs remain persistently high as generally acceptable accounting principles (GAAP) force recognition of expenses in advance of subscription revenue, as well as the continual requirement to replace churning customers.



Taking publicly reported financials of SaaS vendors at face value masks a much more intriguing story and makes it impossible to compare their financial health against traditional, perpetual license-focused independent software vendors. If the role of financial analysis is to compare like with like, the SaaS economic model needs two adjustments.

First, it needs to be normalized for the differences in the revenue model vis-à-vis traditional software companies. There are two fundamental differences in the revenue charged by the two models. The first difference, which is commonly known, is subscription-based pricing as opposed to a perpetual license. The second difference is not as well understood. While the traditional software vendors charge only for the license, the SaaS vendors provide (and charge for) the services required for managing the application - in addition to the application itself.

To normalize the revenues of SaaS, one needs to consolidate the upfront subscription payments related to the application which would be otherwise spread over multiple years. Note that this involves the payments related to the application only, and not the subscription payments related to the managed services (which account for the lion share of the customer payments). As a result, the normalized revenues are 20 to 30 oercent higher than the traditionally recognized revenues.

Second, given the relatively recent emergence of SaaS vendors, their economic model needs to be normalized for the difference in maturity vis-à-vis the incumbent traditional software vendors. The traditional vendors have a scale ten-times larger and growth rates in the 10 to 20 percent range. SaaS vendors have to make higher investments in sales and marketing to drive their higher growth - in fact, SaaS vendors lack economies of scale in all SG&A (selling, general and administrative) areas compared to established vendors.

When SFDC's economics are adjusted for this difference, its SG&A fell from 64 percent to 45 to 48 percent. The $50 million investment in the new data center temporarily raised overall cost of goods sold as reported in the cost of subscriber and support. The table below outlines the impact of these two adjustments.








The SaaS Advantage
Aside from the higher predictability associated with annuity-based revenue stream, the SaaS model has three significant advantages over the traditional software model.

1. A Higher Share of Customer Spend. The SaaS model allows software vendors to capture almost double the total spend on the application. For every $1 spent on the SaaS application, the vendor is able to collect $0.70 on services, while in the traditional model the vendor typically captured only $0.15-$0.20 per dollar spent. The remainder is typically spread across hardware, external professional services and internal IT staff.

The reason the SaaS numbers only translate to double the total software spend is that SaaS vendors squeeze the total spend down materially to begin with. This is a big advantage for the SaaS model because it reduces the "overhead" expenditures that crowd out spending on new application functionality.

2. Better Profitability. The SaaS's model's profitability becomes far more compelling once it is corrected for the differences in revenue model and company maturity relative to traditional application vendors. The analysis suggests that SaaS vendors have the potential to generate 33 to 39 percent in operating margins.

At the heart of SaaS' superior profitability is a dramatically simpler product model: SaaS vendors support and manage only a single version of the software. This means significantly lower R&D costs because resources are not fragmented by supporting multiple versions or releases. It also means vendors have a significantly lower cost of managed services.

In fact, SaaS vendors are able to run their "integrated" support and managed services facilities at a cost structure that is lower than the cost of running just the support centers of the traditional vendors. The COGS for these services runs at 21 percent when revenues are normalized. Because SaaS vendors "own" the production system, they can provide proactive support. They also don't have to manage the complex software distribution system for applying and managing patches and upgrades. The cost advantages from the simplified product clearly outweigh the higher marketing costs required to manage customer churn.

3. Continuous Innovation. With the entire installed base of SaaS customers using the most current version of the software, SaaS allows vendors to introduce innovations to their entire installed base simultaneously. This reduces the cost of introducing these innovations because developers don't have to worry about backward compatibility or migration paths. It is likely that this flexibility also increases the vendor's yield from new innovations because the entire installed base serves as the target market for these innovations - not just the small portion of customers that are using the latest release. SaaS also minimizes the risk of replacement that normally arises when customers are left behind running old versions of the software.

A Look Ahead
The numbers show that SaaS is a far more attractive economic model than the perpetual license model - even more attractive than SFDC's financials to-date have demonstrated. Over the next 3 to 5 years, the sharp distinction between SaaS and traditional software models will blur. Traditional vendors will introduce and expand their SaaS offerings. SAP recently announced their On-demand CRM and Marketing solutions. Oracle inherited the on-demand business of Siebel.

These offerings are only the tip of the iceberg. The enterprise software market will see more offerings from these and other vendors based on SaaS. What will distinguish the winners from the losers will not be the model itself but how the model is executed.

George Gilbert is a consultant with Tech Strategy Partners, a San Francisco-based consulting firm. Scott Bolick and Rahul Sood are its alums. Email George at george.gilbert@techstrategypartners.com.