Why you need real-time Metrics as an Enterprise Architect

Posted by Joerg G. Beyer on March 14, 2016

Enterprise Architecture toolscontain static information about the Application Portfolio and dependencies. An EA tool is usually separated from operational data, which is used in operations and development. In the past this was not a tragedy, as development cycles have been pretty long. Sometimes you released only two releases per year. But the world has changed and will change more and more in the future. The business success is dependent on dynamic user behavior.

You collect a lot of user information in data lakes to identify user patterns, you ensure high availability of operations to keep users on your platform. All the information you collect can help to retain or win business.

There are a number of tools in the market that provide operational data about availability, user logins, incidents, application performance, application health to name just a few. The same is true for more business related data like transaction volumes of processes, process cycle times and bounce rates of a website. But there are literally no tools that combine such operational data with an application portfolio view in a business context. The challenges a modern Enterprise Architect or Application Manager has make it necessary to bring this data together in a business context.

An isolated view of data does not foster the collaboration need of Biz, Dev and Ops

In times where the waterfall model was the paradigm for developers everything was fine. There was a sequential process that took some weeks after the requirements had been defined by the business before it was released. This was an easy process to develop IT applications. Business described their requirements, Development delivered an application and Operations was involved in the release-to-production process. Everybody could live with his silo view. But we are in the 21st century where Business, Development and Operations (BizDevOps) have to work closely together and need a shared view on all aspects of an application. A single language and a joint view on real-time data can help to better understand each other.

Ad-hoc decision taking in a dynamic environment needs a variety of different information

We are more often facing an agile development approach where you have to take decisions in a very short period of time, where you deliver something within a Sprint. And to take meaningful decisions whether to fix a bug, enhance functionality or stabilize operations (you cannot do everything at once), you need more information than before. And this information needs to cover all aspects: business, development and operations. Some day one or the other aspect is more important and you need to judge instantly what to do. And you must do more A/B testing to identify the right approach without an intensive upfront planning and analysis. For these kind of decisions, you need to know a lot of different aspects such as “37 users are right now affected”, and “this application supports a core process” and “uses an outdated technology” and “has a risk value of 8.7”. 

New Call-to-action

Iterative development needs real-time information to identify change needs for the Architecture

In many cases it is meaningless to prepare a 5-year plan for an Architecture or Application Roadmap. Many changes occur in minutes based on the collected information of user behavior. An eCommerce shop has to adapt its use and look & feel within a very short period of time to not lose money. If users are clicking away from your website, you need to stop this right now. What are actual bounce rates, what is the browser performance, what are click rates for a call to action? This kind of information drives your change needs from an architecture point of view. If the Enterprise Architect takes his role seriously he needs to support the Application Manager and System Architect rather than developing every day a new helicopter view and a 5-year roadmap. 

LeanIX makes your life so much easier

The easy-to-use interface of leanIX gets more powerful with the combination of real-time information. The information is not limited to pre-defined metrics. You can define and populate your own EA metrics with the use of your operational data sources via the REST API. We provide every leanIX customer with a free version of metrics where they can define up to 3 metrics. This could be for example the application performance collected in AppDynamics, the bounce rates from Google Analytics, and the availability of an application from Pingdom. In addition to that you can define your own metrics and calculate a risk index based on certain criteria. Or you are analyzing the user behavior of an application by sum up the unique logins for a day.

This raises the bar for EA tools.

Subscribe to the LeanIX Blog and never miss a post again!