So You Can Budget, Compare and Save.
We all have struggled to find exactly how much custom software applications cost to build, maintain, and enhance over their life. Accounting needs to know so they can budget accordingly, HR needs to know so they can assemble the team together, Management wants to know the Return On Investment (ROI) before embarking on implementation. Some applications are easy to calculate and others are not so straightforward. We all want our software application to be designed, developed, and deployed on time and under budget. Exactly how do you calculate total long-term cost of ownership or TCO? Do you have to stage “The Price is Right” for applications? Not really, the formula is very simple. When practiced every time it will help you budget, compare different alternatives, and save while creating successful software applications that exceed your customer’s expectations.
First let me tell you a time-tested empirical formula(1),
Long-term TCO = (Fx + Lr) * [1 + (Roi/Qq)]
(Note) 1: Empirical formulas are not proven scientifically, but they can be accurately applied to most scenarios.
And I will now explain the 4 simple variables it uses:
1. FIXED COSTS (Fx)
Start with the technical specifications and better yet, sit with the technical lead or architect to find out what are the fixed costs. Look at the deployment diagram and find out the cost of each box it shows. Now consider the cost of operating systems they will run on, and cost of all the tools that will be installed. Here are some pointers as to what a typical project may incur as fixed cost:
Design & Development Tools
Most recurring costs can be converted to a fixed cost by multiplying per cycle cost with number of total expected cycles over the life of the application.
Adding up all the values above will give you a total dollar amount, which is your Fx in the formula above.
2. LABOR COSTS (Lr)
Your project plan should have a section on time estimates. Again, your team-lead or architect can tell you more accurately how long is it going to take to build your application. Consider all roles and responsibilities from systems analyst gathering requirements, engineers developing it, to QA testing it and everyone else in-between. Estimate all their hours and put it in following three buckets:
Your Own Employees
Offsite & Offshore Consultants
It is better to multiply each individual’s required hours and rate, but for large teams you can use averages. Adding up all three buckets will give you a total dollar amount, which is your Lr in the formula above.
3. RETURN ON INVESTMENT (Roi)
This is very important and sometimes tricky variable in finding out the TCO. First of all you need to dollarize the benefits of the application you are building. That means translate the increased efficiency, potential growth, or reduction in cost etc. to dollar amount saved or earned per month. Now estimate how many months will it take you to recover the cost of building the application at that rate. I say estimate because it is a catch-22 situation otherwise, since you are trying to figure out the TCO in the first place. For a number of custom software applications Roi is around 12 months. You can plug-in a few different values for this variable and see where do you feel most comfortable.
4. QUALITY QUOTIENT (Qq)
This is where science meets art. We are calculating long-term cost of ownership of a custom software application, which depends on factors related to quality of the application. If the application has fewer bugs, the QA to Engineering to QA to Deploy cycle would be short. If application is documented well, future enhancements will be easier and answering questions will be quicker. Well, you see they all affect the long-term cost of running the application. To measure what such unknowns will cost us in dollar terms for the life of the application, I find it most effective to put quality related issues in following four basic buckets and rate them on a scale of 1 to 4:
You can put a number between 1 and 4 for each of them based on your prior experience with same team, or software. If you don’t have past data, select a number for each that you want your application to have. You can even have your own buckets of four most important factors. Adding these four will give you the last variable Qq needed for the formula.
Though the formula asks for Roi in months needed to recover the cost, the TCO is for the life of the application.
If you got the idea for formula to find TCO of a custom software application, lets do an exercise with your numbers into right column on the worksheet below:
Our sample data Your data
Fx: $120,000 _________________
Lr: $300,000 _________________
Roi: 12 _________________
Qq: 14 _________________
Substituting values in TCO = (Fx + Lr) * [1 + (Roi/Qq)]
TCO = 780k
Thus Total Cost of Ownership for our sample application is $780k over its anticipated life (around 10 years). This figure really helps budget, compare, and save on custom application development.
About the Author
Mahesh Lalwani is Founder & CEO of Pacview, your partner in custom software & speech applications. Pacview has worked with dozens of companies in Healthcare, Financial, Technology & Telecommunications, Education and Entertainment industries, helping them create successful software applications that exceed customer’s expectations. Through Pacview’s expertise in design and development our clients enjoy the benefits of building value through successful software applications. Visit our site at www.pacview.com for more information.
Note: You’re welcome to “reprint” this article as long as it remains complete and unaltered (including the “About the Author” info at the end), and you sent a copy of your reprint to [email protected]