IT metrics: 5 measurement mistakes to avoid

Are you maximizing ROI, increasing revenue, and adding customers? Then your IT metrics should not be all about speed. Consider these data measurement pitfalls – and how to avoid them
177 readers like this
AIOps misunderstandings, explained

In a world fueled by data, enterprises that aren’t measuring and analyzing their performance are doomed to fall behind their competitors. Without the consistent refinement enabled by accurate, insightful measurements, companies are likely to stagnate and miss opportunities for growth.

However, not all technology metrics are created equal, and a poor measurement strategy can be nearly as problematic as having no strategy at all. Even the best tools are useless in the hands of an inept craftsman, which is why modern enterprises must develop a well-considered measurement approach tied to business outcomes.

5 IT metrics mistakes and how to avoid them

Here are five common mistakes to avoid when plotting metrics for your technology or product engineering strategy:

1. Over- or under-measuring

When designing your technology metrics program, bear in mind the resources at your disposal and your organization’s capacity to enact change based on data insights. Setting too many KPIs will not only dilute the impact of the measurements that matter, but it will make it impossible to react to each data point. Over-measuring can also place an unnecessary burden on your tech/engineering teams, preventing them from focusing on the measurements and tasks that matter.

On the other hand, collecting too few measurements could lead to biased or distorted results. Zeroing in on just a couple of performance metrics leads to a lack of context, meaning the insights provided by these data points could lead your enterprise astray. Finding the right scope of measurements for your team means finding a number that is manageable but that provides a balanced view of your innovation apparatus.

[ Read also: OKRs vs. KPIs: what's the difference? and How to explain OKRs in plain English. ]

2. Prioritizing speed over effectiveness

At the end of the day, your enterprise’s performance is about business outcomes. Are you maximizing return on investment, increasing revenues, and growing your customer base? Tech metrics should be designed to track progress towards these outcomes, yet too often measurements are centered around speed. Sure, there is value in measuring time to market, but that value is muted when the feature being brought to market is flawed or underwhelming.

Instead of measuring quantity-based metrics like velocity or lines of code, focus instead on quality; a good metric to assess performance quality is time spent on unplanned work or fixing previous work. 

A good metric to assess performance quality is time spent on unplanned work or fixing previous work.

3. Zooming in too closely on local performance

While this issue is particularly relevant for large global enterprises, it can also be applied to even the smallest startups. At the end of the day, the most important metric is customer satisfaction; enterprises should maintain a high-level view of this by tracking whether their service is functioning and whether their customers are suffering due to any errors or outages. Zooming in too closely – whether that’s on individual performance instead of team performance, or on the success of a single office as opposed to the enterprise as a whole – often leads teams to miss the forest for the trees. Stay focused on the ultimate objective and measure accordingly.

[ Read also: 4 books to boost your data storytelling skills. ]

4. Failing to connect output to input

It makes sense that dev teams focus their attention on outputs; measuring the amount, quality, and speed of features and tools created. However, it’s important to place these metrics in the context of the inputs that preceded them. How much was invested in achieving a certain outcome? How much developer attention did it require, and how many company resources were used to reach the goal? Measuring inputs enables teams to identify areas of waste throughout the development of process, refining the system in its entirety and improving ROI.

5. Settling into stagnant metrics

The goal of any good metrics strategy is to use data to evaluate and refine internal processes. This same goal must be applied to the metrics strategy itself: Without reviewing the usefulness of your measurements on a regular basis, you’re likely to stagnate with a process that isn’t meeting your needs. Enterprises should aim to review their measurement approaches on a quarterly basis, taking a good look at what metrics need to be tweaked or scrapped in service of more useful information.

We are living in the golden age of data-driven engineering. Enterprises must make the most of the resources at their disposal to stay at the bleeding edge of innovation and avoid falling behind an ever-growing number of competitors.

[ Get exercises and approaches that make disparate teams stronger. Read the digital transformation ebook: Transformation Takes Practice. ]

Jiani Zhang is President of the Alliance and Industrial Solution Unit at Persistent Systems, where she works closely with IBM and Red Hat to develop solutions for clients. Prior to this role, Jiani was the General Manager of the Industrial Sector for Persistent Systems.

Social Media Share Icons