You are here

Making the case for continuous delivery: How formulas can show the value

public://pictures/Andrew-Phillips .jpg .png
Andrew Phillips, VP of DevOps Strategy, XebiaLabs

When building a business case for adopting continuous delivery, there are many different benefits you can use. Increased competitiveness, enhanced goodwill, and improved revenue will be at the top of the list, but such gains are relatively intangible. You need to quantify your case by looking at how you can speed delivery, reduce wasted time, cut operating costs, and decrease production incidents and defects. You should not ignore the intangibles, but your business case for continuous delivery will have more impact if you back it up with some hard numbers.

The good news is you should already have all the data you need. The challenge is to create a formula that allows you to plug in the relevant numbers and discover exactly what the bottom-line advantage of continuous delivery is going to be.

[ Learn how value stream mapping can improve your DevOps workflow. Download this GigaOm Research Byte report today. ]

Calculating the cost of production incidents

When you're looking at adopting continuous delivery (CD) to reduce production incidents, you have to consider three things:

  • A reduction in the number of incidents and their severity due to standardized environment provisioning
  • Less manual effort required for provisioning
  • Less time spent waiting for developers and testers

The average cost of data center downtime is $5,600 per minute, and the average downtime is 90 minutes, according to a Ponemon Institute study, so that's more than half a million dollars per incident.

Go back to your root cause analysis data and work out how many of your production incidents were caused by provisioning errors. You can make the calculation to find the potential revenue losses from your own business with a simple formula, as outlined in this Evolven blog.

LOST REVENUE = (GR/TH) x I x H GR = gross yearly revenue TH = total yearly business hours I = percentage impact H =number of hours of outage

Apply this formula and you can illustrate the potential revenue savings when you reduce outages with CD.

[ Get Report: The Journey to Enterprise‐Scale DevOps: Becoming DevOps Determined ]

Savings in time and resources

Looking beyond the impact of the outage, you can also factor in potential savings in terms of resources and the time you gain from avoiding incidents. That increase in productivity applies to everything, from the manual build process to deployment failures.

Analyze the number of deployment failures you've had, work out the average cost per failure, and then look at the frequency of your failures. You can cut that cost with CD, and you can also factor in the regained time that would have been spent diagnosing and remediating.

It's harder to cost your manual testing, but a shift to API-based testing can speed the process, extend coverage, and reduce manual errors and inconsistencies. How many of your defects could have been caught and fixed earlier? Automated testing frees up developers to focus on improving the product and testers to focus on finding defects earlier. The cost benefit becomes more apparent over time, as the number of builds goes up, because overhead increases with manual testing as you add resources.

Down the pipeline

There are important benefits further down your delivery pipeline. With the right of kind of analytics you can calculate the cost of implementing features that your end customers don't care about. Only 33% of new features have a positive impact on outcomes, 33% have no impact, and the final 33% actually have a negative impact, according to a Forrester webinar.

Define your desired results, deliver in small, frequent releases, and use the feedback loop to analyze the reception. You don't want to rely on user feedback alone, so plug some metrics into the platform to automatically gather evidence. By examining what is being used the most and what isn't being used at all, you can effectively prioritize your efforts.

All the time that's saved can be used to increase revenue. You can take the same formula and apply it to showing the benefit of speedier delivery. Plug in the numbers for what a product makes and work out how much extra revenue would have been generated if it were delivered two weeks earlier.

The final consideration here concerns customer experience. If you can improve the customer experience score, even by a single point, the increase in revenue can be dramatic. It could be worth $175 million annually to a wireless provider, $65 million for a hotel, or $61 million for a TV provider, according to Forrester.

Every business will have its own set of standards and risk levels, and you'll need to tweak your calculations accordingly. But you can use this formula as a solid foundation for working out the tangible benefits of CD, significantly strengthening your business case.

[ Webinar: Paving the Last Mile to Production: Putting the "O" in DevOps ]