3 ways to retool your architecture to reduce costs

It's been two years since our organization started our Technology Business Management (TBM) journey. In How to measure the cost of running applications, I talked about what we got out of the implementation as early adopters by leveraging Apptio's SaaS solution.

Because it provides cost transparency, an application's total cost of ownership (TCO) has been the shining star that helps the platform and application owners get the most value from their TBM implementation.

[ Learn how IT modernization can help alleviate technical debt. ]

As we've continued our adoption and expanded our TBM operation to the entire IBM CIO Hybrid Cloud Platform organization, we've learned more than we expected. Here are three of the most important lessons learned.​

1. Reduce costs by decommissioning servers

When you have one team developing and operating one application deployed on one server, it's not difficult to figure out the infrastructure, labor, and software costs of developing the application. However, when you have thousands of applications, it becomes a real mess to trace every penny flowing into your application's TCO.

Consider situations where applications are in different lifecycle stages (for example, some in development and others in production) and have different business criticality (some are more important to the business than others), they have different asset classifications (such as freeze, invest, or sustain), and there are internal billing and chargebacks involved. In these cases, identifying a common allocation rule that the platform owner and application owner agree upon is significant.

One question to answer is whether to allocate cost for applications you phase out or cancel. My answer is yes if the application is still actively connected to the server. Whether it's because no one remembered to decommission it from the server or the application team wants to keep the server available for future apps, at the end of the day, it accrued cost to the company. If these costs aren't included in your reporting, no one cares about them. By showing costs associated with phased-out and canceled applications, the application owner can reduce costs by decommissioning the server if it makes sense.

[ Managed services vs. hosted services vs. cloud services: What's the difference? ]

2. Identify cost-saving opportunities by analyzing scenarios associated with sunsetting applications

There's a common misunderstanding that if an application's TCO is $1,000, you will save $1,000 by eliminating it. The reality is far more complicated. For platforms, especially hybrid cloud, some costs may go away immediately, while some will be redistributed to new application portfolios. The labor costs associated with the application will remain unless you actually eliminate that labor; otherwise, those costs will be redistributed across the other applications the team supports. In other words, labor costs will increase for those other applications. For software license costs, sunsetting one application may or may not reduce it immediately.

Therefore, your first step is mapping the application's TCO. The next step is identifying fixed vs. variable costs; this will contribute to scenario analysis and support business leaders with decision-making.

3. Avoid costs by understanding excess capacity

There are two views of measuring excess capacity. First, to decrease unused reservations by application owners, you must allocate total platform cost by total reservations. However, for platform owners to understand excess capacity compared to the total built capacity, you must allocate the total platform cost by the total built capacity.

Having visibility into how much capacity is consumed compared to total built capacity will help platform owners decide whether they need to add fixed assets. If they do, how much to provision should not only be based on technical needs but also on the financial impact.

Conclusion

The TBM journey is not about implementing a tool to generate reports; it's about transforming the culture from making decisions based on financial-driven needs to considering business-driven needs, as well as continually providing business insights to drive business decisions.

[ Hybrid cloud and Kubernetes: A guide to successful architecture ] 

This originally appeared on Hybrid Cloud How-tos and is republished with permission.