Cisco Blogs
Share

Seven Software Business Models – Part 1

- February 23, 2015 - 2 Comments

Many years ago I found myself talking to a venture capitalist about the differences between SaaS, outsourcing, ASPs, MSPs, online applications; etc. Also I noticed that my Stanford students had little understanding of the economics of software, so I developed the idea of seven business models to cover everything in the software business, remove the buzzwords and replace them with economic models.

In my previous post, I talked about the Seven Ways to Move to the Cloud. In the second issue (there’s a lot here), I’ll break this into two separate posts, discussing models one through four here, and models five through seven in the next issue publishing on Monday, March 2.

CLOUD 7 business models

Note the dollar numbers used throughout are intended to be relatively representative.

 

Model 1: Traditional (On Premises)

The traditional, on-premises Business Model One has been used for many years by vendors such as SAP, Oracle, and Microsoft. In this model, the customer pays a one-time, perpetual license fee, which is the right to use the software.  For enterprise business applications like those from Oracle or SAP it can be about $4,000 per user.

In addition to the license fee, the vendor will charge for support and maintenance, which is typically a derivative of the license fee. Vendors charge between 20% and 30%, so if you use 20% then on the $4,000 perpetual license fee the customer will spend an additional support fee at $800 per user, per year.

If you’re a student of Oracle prior to the Sun acquisition, you’ll know Oracle was doing about $15 billion worth of revenue, and $12 billion was a recurring maintenance revenue stream. Profitability in these models can be north of 90%. They are amazing business models if you can get them.

 

 Model Two: Open Source

Model Two is the open source model, meaning the user gets the software for free. However, the vendor charges for support and maintenance as the mechanism to monetize the open source. The only reason I put $1,600 in our example above is just to make the point that you could charge more than you do in Model One.

There are actually very few examples of success with this model, although some of you may be thinking about the single example, which is Red Hat. Red Hat’s been enormously successful with the open source model. Red Hat monetizes your subscription to the Red Hat network, a subscription to the support and maintenance of the Red Hat software.

 Cost to Manage Software is 4x the Purchase Price Per Year

Before moving on to Model Three, it’s important for those of you who have never managed traditional applications to realize that buying the software is not really the end of what you pay for.

Not only do you have the cost of the hardware and software but more importantly, the cost to manage the security, availability, performance, and change of the application and everything underneath, such as the computing storage environment, the data center, and the network.

As a rule of thumb, the cost to manage software is about four times the purchase price of that software per year. Therefore, if you pay $4,000 for the software, you’re going to spend $16,000 per user per year to manage the security, availability, performance, and change management required for the entire infrastructure running under the application. This roughly works out to be about $1,300 per user, per month.

For those of you who are having a hard time believing this, just go talk to any CIO of any major corporation and they’ll tell you they spend 80-85% percent of their budget managing the existing portfolio of applications and software.

 

 Model Three: Outsourcing

As a result of this cost to manage software, Model Three came into existence. The original company in this model is a company called EDS. Outsourcing companies came in and said, “We’ll manage your mess for less.” Go and buy the software from the software vendor, pay them the support and maintenance fees, but we’ll manage it for less than what you’ll spend doing it yourself, meaning less than $1,300 per user, per month in our example.

As you can see in the model, the outsourcer can do that at home, meaning at EDS’s data centers, or at customers, meaning the outsourcer will take over your data centers.  This model took an enormous leap in the late 1990s, because of the advent of the Indian outsourcers like Tata, Infosys; etc. The reason for this is Model Three is powered by human labor. This means that the management of the security, availability and performance of the software is done by people.  Therefore you can reduce cost by moving to low labor rate countries.

However, there is a floor to the cost.

 

Model Four: Hybrid

Model Four is the model originated a little over ten years ago at Oracle. Since then, many existing traditional software companies have adapted this model. Basically, you still say, “Please pay me for the software and the support and maintenance, but I – the creator of the software – will manage the security, availability, performance of the software at significantly lower cost than you can by yourself. In the example, I show this number at $150 per user, per month.

The question you might be asking yourself is, “how is this lower cost possible?”

The answer is there’s something very fundamental at work here, which from an economic point-of-view is driving everything going on in cloud computing.

… the answer will be revealed in my next post.  Visit my Cisco Blogs page to read more on Monday, March 2.

 

And For More Information

For more information, and many more examples of how businesses moved to the cloud, check out my book on Cloud Computing Fundamentals .  In fact, Seven Business Models is discussed in more detail in a TED-sized chapter in this book.

 

 

Tags:

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.

2 Comments

    Excellent Post. The advantage of simplification of an idea.

  1. The cloud computing concept is more appropriate to a business concept than a theoretical or applied computer science one. Is a paradox to name a technology stack with a business model name. Practically, the stack of technologies associated to the cloud computing concept implements a family of distributed computing paradigms. The "computing" construct from cloud computing concept switch the meaning of the concept from business to theoretical or computer science domain. Excellent post!

Share