Pure on-demand software-as-a-service businesses are difficult to build.  It is the rare B2B SaaS startup that masters all the Dos and Don’ts of SaaS Success from the beginning.  When the going gets tough, many find themselves falling back on traditional enterprise software approaches to product delivery and business operations.  However, there is a big difference between making the strategic decision to deliver your product in a hybrid model and stumbling into a hybrid model through tactical mistakes. In the first instance the market requirements demand a hybrid approach, in the second executive management is simply not disciplined and creative enough to avoid it.

In an earlier post, I presented the relationship between pure on-demand SaaS and it’s closest hybrid cousins: managed services, packaged software, and enterprise software. The four models differ along the dimensions of pure commodity product (SaaS and packaged software) and pure online delivery (SaaS and managed services).  When a business travels down the hybrid path, it is making a choice to deviate from pure on-demand along one or both of these dimensions.  The farthest deviation being enterprise software. Seems straightforward enough.  So, why avoid a hybrid approach?  Let me state up-front that there are markets that absolutely demand a hybrid model—-a great example is anti-virus software, which requires a fat client for rapid response (packaged software), Internet delivery of virus definitions, software updates and threat alerts (SaaS) and a semi-automated, labor-intensive process for collecting, analyzing and categorizing threats (managed service). However, it is much more often the case that complex market requirements provide the rationalization for poor management rather than the rationale for good business strategy.

The Perils of Hybrid Models
Competitive advantage in SaaS is built by leveraging the Internet to lower costs through economies of scale or to develop differentiation through network-native capabilities. Moving away from a pure commodity product (toward managed services) breaks economies-of-scale, while moving away from pure online delivery (toward packaged software) breaks network-native capabilities.  Deviation due to a lack of discipline without compelling market requirements results in the textbook failure of competitive strategy:  getting stuck in the middle.

“The firm stuck in the middle…is almost guaranteed low profitability
…probably suffers from a blurred corporate culture.”

Sound familiar?  This is the primary reason that some of today’s most well known SaaS companies have yet to attain profitability.  It has nothing to do with the so-called up front investment costs of building a SaaS business (seriously…do NetSuite, Omniture and Success Factors ever plan to stop posting losses?).  They have given up too much competitive advantage in pursuit of revenue from markets where they are not competitive or from customers that they cannot serve profitably (as evidenced by this nice little financial analysis by Bob Warfield).

The Right Way to Go Hybrid
Many markets present challenges along both the dimensions of product uniformity and online delivery.  Unique customer processes,  complex technologies and heavy integration requirements for example can all create serious strategic hurdles.  To overcome them successfully, it is essential to be clear about exactly the business you want to be in and exactly where your competitive advantage truly lies.  Read more »

Since it’s inception, software-as-a-service has labored under an identity crisis. What truly distinguishes SaaS from software? How should it be priced, sold and serviced? Is it possible to succeed with a hybrid approach where a vendor offers both SaaS and software versions of a product? Recently, Jeff Kaplan reported that Google will be offering Gmail as an installed service–flying in the face of the current conventional wisdom that SaaS players should stay true to their model. Alternatively, Oracle is finally making a serious challenge to salesforce.com with its Oracle CRM OnDemand (I know because their sales reps keep calling me, but I’m still not buying it).
What does it all mean?

First and foremost I would like to say that IF YOU ARE A STARTUP, it doesn’t matter if you are offering enterprise software, B2B software-as-a-service, online games, or even hardware—STICK TO ONE BUSINESS MODEL!!! Handling multiple business models almost always entails increased organizational complexity and heightened internal politics. It is often essential to separate organizational functions (e.g., sales, marketing, engineering, etc.) or even entire P&Ls (e.g., divisions, spin-offs, etc.) in order to achieve the right cost structure and culture required to be successful in each line of business. In short, it is death to a severely resource constrained company.

The most common software technology distribution models arise naturally from the economics of information goods. Computer software does three basic things: copies data, transforms data, and moves data. But, the unique thing about software (unlike hardware) is that it just happens to be made up of data itself. Why is this important? Because the costs of copying, transforming and moving data around is decreasing everyday, and in many circumstances it is economically equivalent to zero. This is why so many Internet applications are free! The more a software application lends itself to this sort of self-referential automation, the lower it’s cost. Pretty theoretical, so here it is in plain English. Custom applications are hard to deploy (copy and transform) and fat or data-intensive applications are hard to deliver over the Internet (move). And, here is the picture.

How costs give rise to different software distribution models.

The more you can align your business model with the underlying economics of the technology, the better off your will be. Perfect alignment is rarely achievable, because your customers will pull your business in one direction, while your technology will pull it in another. So, choose your customers and your technologies wisely! However, here is how the most common software business models line up with the technology choices above. Read more »

At the end of 2008, I promised a special new year post on SaaS Model Economics 101, so here it is.  Introducing the SaaS Scorecard, an online test to estimate the competitive advantage of your software-as-a-service business.

Click on the image above to go to the SaaS Scorecard and take the test.

The SaaS Model Scorecard is designed to help software-as-a-service entrepreneurs and investors evaluate the competitiveness of their businesses relative to licensed software and other SaaS competitors using the principles of SaaS Model Economics 101 and the Top Ten Dos and Don’ts of SaaS Success. Every attempt has been made to create a test that accurately reflects these economic principles, however, the goal is simply to provide feedback as opposed to analysis. That is, it’s really just a game. Have fun!

Scores are calculated across 20 key business dimensions that impact low cost advantage, differentiation, adoption costs, switching costs and network effects. For each dimension, a score is calculated for the potential competitive advantage that can be achieved, the current performance of the SaaS vendor in achieving it, and the combination of potential and performance resulting in the actual competititive advantage that is realized.

I’ve recently been asked a lot of questions like the following:

• Are there some applications that don’t fit the SaaS model ?
• When is the SaaS model appropriate ?
• Is it possible to have a “mixed” SaaS model ?

Mostly these are asked by startups that are struggling for growth or profitability and having difficulty actually achieving my Top Ten Dos and Don’ts of SaaS in practice.

To answer these questions accurately, it’s essential to have a strong understanding of how SaaS creates economic value over in-house, licensed or home grown software.  If you can create value, then the model is appropriate.  If you can’t, then in-house software is an equal or better choice.  And, as a SaaS vendor you will have no competitive advantage.

Be warned, that there will be few fluffy marketing tips in this post and the series to come, and you are likely to encounter some very specific economics terminology.  This is SaaS microeconomics 101! It’s going to get a little heavy.

The only difference between software and software-as-a-service is that SaaS is delivered over a standards-based network called the Internet.  Therefore, all new economic value and competitive advantage must flow from this difference.  SaaS economic value over software comes in two Web-enabled flavors:

1) A lower cost structure from economies-of-scale that derive from aggregating customers via the Web onto a single, vertically integrated infrastructure i.e., hardware, software, maintenance, etc. This cost savings is generally passed on to the customer through a low subscription price and is generally referred to as the SaaS lower total cost of ownership (TCO)

2) Reengineering business processes by leveraging network automation e.g., online trial, integrating local offices, support chat, etc. and network effects, e.g., crowdsourcing, support forums, revenue-sharing monetization, etc.

Both of these sources of value can create sustainable competitive advantage for the SaaS vendor, 1) is a low cost advantage and 2) is a source of product differentiation.

However, when the nature of your customers, application or technology limit your ability to create value from either of these sources, then you have reached the boundaries of the SaaS model for your business, because beyond these limits there will be no competitive advantage over in-house software. Read more »

In many ways, marketing a software-as-a-service (SaaS) application is more like marketing packaged software, computer hardware or consumer electronics than enterprise software.  Failure to make this paradigm shift has meant the death of many a SaaS startup.  The reasons are simple.  First and foremost, enterprise software is usually delivered in an unfinished state.  The so-called product is delivered and then configured, customized, integrated and QA-ed onsite to deliver a unique solution–a product of one.  This is more akin to the artisan products of a cottage industry than to manufactured commodities.

The fact that SaaS is a commodity delivered via the Web entails a shift in business model that affects everything from product design to organization design.  Below is a summary of characteristics that contrast the traditional enterprise software business model to the new SaaS business model.

In the world of SaaS multi-tenant architectures and bargain basement prices, the entire business model hinges on having a single commodity sold at high volumes.  Moreover, SaaS is marketed and delivered primarily within a single channel, the Internet.  This creates incredibly tight coupling between the product, business strategy and operations.  In particular, there is an unusual itermingling of the product itself with the other 3P’s of marketing:  price, promotion and place.  For example, a change in pricing model will usually entail simultaneous changes to both your Web ordering code and your license management code.  And, search engine optimization (SEO) is likely to impact how your product is designed and delivered over the Web, not just your marketing website and landing pages.

I was recently asked the question of what is the ideal cost structure for an on-demand software business.  There are very few benchmarks out there of successful mature SaaS companies, so I’d like to propose the following as the cost structure to which on-demand software companies should aspire.  Whether you can actually achieve it in your business is likely to be a result of market demands, technology and sadly enough company culture (as many SaaS business are still saddled with enterprise habits)

Below is what I would characterize as the typical enterprise software company cost structure.  This model is no accident, and it has proved to be immensely profitable over the last 20 years (just ask Oracle and Microsoft).  It is characterized by drivers that work synergistically to create the whole: perpetual license pricing, feature competition, solution selling and customization.  All of these characteristics (and their ensuing complexity and costs) derive from the underlying buyer belief that the system will deliver some level of competitive advantage.  While this may have been true 20 years ago, and may still be true for some fundamental business processes, it is patently untrue for 90% of most IT infrastructure today.  Hence the rise of SaaS and Open Source.

Unfortunately, the preceding paradigm is a self-reinforcing business model that naturally evolves toward this equilibrium.  It is incredibly difficult to break out of economically and culturally.  Below is what I am proposing to be the natural equilibrium cost structure of a well run on-demand software business.

This SaaS business model is equally self-reinforcing and composed of a number of drivers that work together synergistically:  subscription-based pricing, product simplicity, and the continuous automation and integration of marketing, sales and service processes throughout the product and the company’s overall Web presence based on catalytic learning. By catalytic learning, I mean Read more »

Henry Ford once said: “Any customer can have a car painted any colour that he wants so long as it is black.” Then, in 1923 Alfred Sloan of General Motors came along and changed the rules of the game by offering a tremendous variety in colors and models. But, GM didn’t do it one customer at a time. GM redesigned its manufacturing line with the required flexibility to produce a multitude of models and colors without compromising the inherent economies of scale of Ford’s assembly line innovation—a practice that has evolved into the concepts of flexible manufacturing and mass customization.

The primary enabler of mass customization is the elimination of setup costs. Setup costs occur from the labor, time and tooling it takes to switch a production line from one product to the other. High setup costs encourage long production runs to cover the expense incurred in switching over. By reducing them, production runs can be shortened. If they are eliminated, production runs can be reduced to a single unit. That is you can make the variations A, A1, A2, … An of a product (think GM models) for the same costs as making n units of A (think Ford Model Ts). If you apply this idea to enterprise software, taking each customer installation as a “unit” and the associated, customer-specific implementation, configuration, customization, and ongoing maintenance time and effort as the setup costs, then the roadblocks to mass customization in the SaaS model become clear: eliminate, automate and generally squeeze the cost out of your ability to handle unique customer requirements.

This business need entails an architectural requirement that is as essential to an SaaS vendor’s success as system security and the scalable, single-instance, multi-tenant architectural imperative. It requires automated deployment that consumes minimal resources, extensive, easy-to-use, self-service configuration and complete interoperability built on open, standards-based APIs. It cannot be off-loaded to VARS or customers. This shifts the costs downstream and undermines competitive advantage, because from the customer’s perspective, total cost of ownership is not reduced relative to installed software.

If you buy into the idea that on-demand software is the next evolution of software into an industry characterized by mass commodity markets and interchangeable parts. Then, your next concern must be “How can I make money in a commodity market? How can I create a sustainable competitive advantage?” This post is the first in a series that attempt to address this question.

In a nutshell, competitive advantage for vendors choosing the SaaS model will not come as easily as it has in the past for enterprise licensed software vendors and it will not come from traditional methods such as creating wiz-bang features, protecting source code, obfuscating product information, or arbitrarily locking in customers with unique customizations. The main reason B2B SaaS providers are struggling today—while their B2C counterparts are thriving—is that as commodity suppliers, B2C SaaS vendors are much more in tune with the traditional sources of competitive advantage. B2B SaaS has as much in common with computer hardware, telecommunications, financial services, and consumer packaged goods as it does with the old craft world of enterprise software. Maintaining a cost advantage, cultivating brand loyalty, network effects, service quality, mass customization, reduced time-to-market and continuous disruptive innovation are the keys business success. And, they are much harder to achieve.

Simplicity and cost efficiency come first in SaaS

A word of warning before you attempt to be different, make sure you have your cost structure in line. If you are competing against traditional licensed software, lower cost is your competitive advantage. If you are competing against another SaaS player, product complexity and higher cost will eventually kill any other advantage you try to achieve. Why? Because there are (or will soon be) many other choices just like you. Change happens, consumers are fickle and expect their software for free, and business buyers have even less loyalty when it comes down to price.

There, I said it. Difficult customers with highly specific requirements who expect to derive competitive advantage from your offering need not apply. If your target market consists of a list of fewer than 500 control-freak customers with sophisticated IT organizations, and lots of unique requirements, then you are wasting your time thinking of the SaaS model as the underlying approach to your business model. At best, you are a “managed service” where the underlying cost structure is identical to customized, licensed software, i.e., it really doesn’t matter if you run it or they run it, it will take the same number of people, the same number of servers, and the same number of lines of code.

It is important to note that the term “commodity” refers to the fact that the SaaS model is a mass market, high volume business where by and large “one size (read code base) fits all.” We don’t have to look any further than Google search to demonstrate this. How much more “one size fits all” does it get! Type in what you want to find, push button. Also, I am not precluding the possibility that you can attempt to overlay another competitive advantage on top of the inherent cost advantage SaaS entails (see below). But, your efforts will be visible to all your competitors and often short-lived. The proposed idea is that cost efficiency trumps product differentiation, and you deviate from this principle at your own risk.

Fundamentally, the SaaS model is the further commodifation of software achieved by freeing it of physical distribution and increasing interoperability through interchangeable parts (services) to create a disruptive software delivery model with new economies-of-scale and business models that achieve a significant cost advantage—not a functional advantage—over traditional licensed, installed software. It is typicially characterized by Read more »