Twelve Metrics to Monitor for SOA's ROI

Wednesday, I joked that Joe McKendrick had kicked off the quarterly “how to calculate SOA's ROI” discussion with his recent post on how to measure SOA's performance.

SOA expert and blogger David Linthicum and Dan Foody of Progress's SOA Infrastructure blog soon followed up McKendrick's post. I particularly liked Foody's take on the topic, which included a lot of excellent caveats about calculating SOA's ROI and what will and won't impress the business when it comes to SOA metrics.

Still, rather than rehash the nuances of their discussion, I thought it'd be fun to compile a list of the metrics experts say should be used to calculate SOA's ROI. You know - so you have something to look forward to this weekend. Because hyperlinking on something like this can be a nightmare, I'm going to include the metric's author in parenthesis.

Here goes:

1. Return on investment (ROI) per service (Mark Little, CTO of Red Hat). ROI per service can serve as an early indicator of your overall ROI. Little warns that all services won't generate a positive ROI, and some services may only perform a business task in conjunction with others, which means you'd have to look at the ROI across several services.
2. Revenue per service (Little). Not all services will generate revenue, of course.
3. Service growth rate/reuse (Little), or Number of New Services Generated and Used as a Percentage of Total Services (Dr. Jerry Smith, CTO of Symphony Services). This metric can help you ensure you're reusing services when possible and not developing redundant services. It's also useful in calculating ROI for a service.
4. Business agility (Little and Linthicum) or Mean Time to Service Development (Smith), measured in how long it takes for a service to go from the design stage to the deployment stage.
5. Mean Time to Service Change (Smith). The time it takes to change a service is another option for measuring agility.
6. Reliability (Little), measured in mean time to failure (MTTF) and mean time to recovery (MTTR). Smith calls this “Service availability,” but he adds another factor to measure as part of it: “Cost of Not Using or Stopping a Service." Smith notes that a well-designed SOA will have “low shutdown or switching costs.”
7. Service Vitality Index, (Smith) which “tracks the amount of revenue from new services over the last 12 months as a proportion of total service revenue.”

Leo Shuster, who is responsible for National City Bank's SOA, wrote an excellent piece on SOA ROI in February and, in addition to many of the above IT metrics, he suggested you also look at the following business and financial metrics:

• Efficiencies associated with service reuse
• Integration time savings
• Related opportunity costs
• Cost savings/avoidance
• Reduction in project and maintenance costs

He also offered this formula for calculating SOA's ROI:

SOA ROI ($) = Cost Savings/Efficiencies Achieved - All SOA-Related Investments

or

SOA ROI (%) = SOA ROI ($) All SOA-Related Investments

Finally, for more help on pinpointing how SOA affects integration, reuse and agility,  you might check out this short slide presentation, which was created by Linthicum. It includes charts showing how SOA can reduce integration costs and how to calculate the value of reuse and agility.