Obtaining Enterprise Architecture Metrics - Part 1

For most EA organizations there is a huge struggle to justify their existence. There is quite a bit of criticism on the practice from folks asking, Does Enterprise Architecture Matter?  

The common quote I hear from our EA customers about opposition is simply "Why do we need you" or "So what we are disorganized, it has worked in the past". Debating those points are challenging with out the information to quantify an EA function.

After hearing those statements and questions then there is a mad rush to get the "ROI of EA". The are articles out there they give prescriptive guidance to obtain metrics such as this article. The problem with articles like this is that they make two mistakes. They are unable to clearly Quantify and Qualify the metrics.  See below:

image

So maybe then you are thinking, well let's align this with the next big EA program such as "future state architecture" processes or strategic programs like what Pankaj Arora talks about with his post on EA ROI. But this approach is flawed because the ownership and accountability of the actual implementation of those ideas are by other teams. You will not get the credit for that great new Online Banking System but rather the e-commerce team.

For the most part EA organizations do not have this quantifiable information. Since they do not, EA's equip themselves with what I call EA Analyst Battle Cards such as the one below: 

image

This was a compile from IASA organization in 2007 from the leading EA analyst communities.  

This is a mistake to carry around EA Analyst Battle Cards because they do not tell your EA customers (e.g., your CIO or CFO):

  • Your EA Value
  • Current State of Your Enterprise
  • The Future Value of EA

What this does show your customers is the potential, that's it. Gartner, MIT and such provide great data points what's possible but the circumstances in which these studies derived could be drastically different than what your enterprise looks like.

So what are some of the things you need to keep in mind wen trying to obtain the "right" metrics:

  1. Obtain the Information
    • Know what you can get - Try to avoid a "fools errand run" by knowing what types of metrics you can reliably obtain. If your current processes require guess work in metrics, skip it for now.
    • Get the Tools - To obtain metrics you need tools to collect and store the information. Once the basics are covered, consider automation into the EA process and into specific tooling. The tool sets must allow transparency of the EA process. This allows for self-service metrics from stakeholders.
  2. Process
    • Get a Process - The process must be able to Quantify and Qualify the metrics.
    • Repeatability - All metrics gathering processes need to be repeatable. If not then revise the process. This will ensure the reliability and the accuracy of gathering your metrics.
  3. Communicate
    • Understand your audience - Not everyone is looking for the same metrics, to demonstrate your value you should be able to communicate what they are concerned about.

          image

    • Traceable - Ensuring that the metrics that you are gathering "matter". Meaning that  the metrics you obtain should align with the organizations strategic imperatives. Additionally, demonstrating the validity of the metrics are aligned is crucial.
    • Relatable Information - To successfully communicate value avoid abstract metrics (e.g., Improved interoperability of systems by 40%). Provide information in terms they understand and can "touch and feel". To do this relate these metrics back to strategic programs or projects to demonstrate the tangible value.

 

In summary, the key to success with obtaining the right EA metrics is to have a clearly defined process to obtain the metrics, the tools to get the information and understand what are the right metrics you should obtain.

Remember, numbers are not everything. In many cases value is not always equated to numbers. Perception is very powerful. For example: "Jim was key to the success of this project, he saved the project team from making the wrong decisions in the architecture design process". This is very powerful, and it is difficult to put numbers around it.

In part 2 I will discuss some ways to obtain the right metrics.

 

Tags: Enterprise Architecture