Some things in life are inevitable: death, taxes, and obsolete software. For extreme examples of obsolete software see this article by Kenneth Corbin: U.S. CIO aims to cut legacy spending, proposes IT modernization. While technologies fall out of favor, companies change, their markets grow, they are familiar with their software and may hang onto it long past its prime. Though there are many reasons not to change enterprise software, possibly the biggest is inertia. If it isn’t broken, why fix it?
The cost of not replacing old software must be estimated to answer the above question, especially for systems that have been in place for more than ten years. This cost is often much higher than expected and is incurred every year. Start the process by estimating the common costs of obsolete software listed below:
- Revenue. Would new software enable the selling of new services or products that can’t be provided with the current system? That missed revenue is a cost.
- Profit. Would new software increase profit by:
- Allowing more to be charged for existing services, e.g. because they are done faster or with greater precision?
- Allow costs to be cut, e.g. by reducing process steps?
- Growth. Is organizational growth being constrained by obsolete software? For example, is the company unable to tackle markets in new locations because of software limits? See an example of how software limitations played a significant part in the failure of Target’s expansion in Canada, ultimately costing them over $7 billion.
- Retention. Will the new system better retain customers, e.g. with faster service? If your market is online, would the new software reduce abandoned shopping carts?
- Information quality & currency. Today, business runs on information which must be comprehensive and current. Obsolete systems don’t capture information in the detail required and are slow in providing it to other systems. For example, a batch job may only run at night, yet the business needs the information in real time.
- Support costs. Complex systems cost more to maintain than simpler systems, e.g. needing more admins, developers and help desk resources. Newer systems tend to cost significantly less to run because many of those complexities have been eliminated with better software. What would the annual cost difference be?
- Training costs. The cost of training new users on obsolete systems can be high. For example, we were recently talking to a client who still used green screens. They complained that because the old software was so different to what was on the market today, it took more than four weeks to get new employees productive on those old but critical systems. With new employees every month, this is an ongoing cost for them.
- Business risk. Old software often runs on obsolete hardware. If that hardware fails, everything stops. You may struggle to find replacement hardware and technical skills for the repairs. Estimate the cost to the business of losing that system, and multiply it by the probability of the system failing.
- Integration costs. It will often be very expensive to integrate obsolete software into more modern systems. This cost is incurred either by having to pay more to write integration code or have data moved manually, e.g. via spreadsheets. Estimate the annual costs of integrating the old system with other new systems that are being deployed in the organization.
- Software maintenance. After about five years most companies will have spent more on software maintenance than they originally paid for the software. Also, many organizations pay for maintenance that never will be used. Estimate the annual costs for software maintenance, and subtract the maintenance costs for the replacement software or cloud service.
- Time & effort. Will the new software enable existing work to be done faster and with less effort?
- Productivity & efficiency. New software will usually enhance productivity & efficiency because of things like better user interfaces or improved workflows. Examine the anticipated productivity and efficiency gains from the new software, and estimate the annual cost of not having them.
- Processes. If the new software supports improved processes, what would these be worth?
- Quality. Would the new software allow earlier detection and correction of defects in your system or production? What would that add to the bottom line? What effect would that improvement in quality have on reducing customer churn?
- Reporting. Modern software usually has far better reporting capabilities than older software:
- Reducing the work needed to generate reports.
- Making reports available sooner, e.g. by not needing to wait for batch processes to complete or eliminating manual data massaging.
- Providing information that could not be reported on in the old system.
To estimate the cost of not replacing obsolete software, estimate the annual amounts for each item above that applies, along with those from any other situations particular to your organization. The total of all those items is an estimate of the annual cost of not replacing the software. The bigger this number, the more urgent replacing the obsolete software becomes.
By failing to examine the cost of not replacing obsolete software, you may be leaving a lot of money on the table. At a minimum, software that is older than ten years should go through this exercise, which should then be repeated every three years. The upside of replacing obsolete software is that some of the savings can find their way back into your pocket in the form of bonuses or promotions. You have a vested interest in taking a close look at that old software.