Enterprise organizations have tons of data locked away in their various systems and applications however they struggle to gain insightful metrics from that data. Many fail to use their customer and business data to shape the decision-making processes within their organization.
A majority of organizations focus on industry standards or generalized operational metrics but these metrics are often not good enough to help with understanding what’s going on and aligning people to what the target behavior is. Metrics need to correlate meaningful business outcomes.
Organizations need to identify the desired target state that can be easily measured and tracked. Organizations need key performance indicators (KPIs) that determine what is important and provide meaningful insight for decision-makers. They speak to the direction your workforce, leadership, customers, and other stakeholders are taking and they drive the behavior of the organization or the system.
In this blog, you’ll learn what a metric is, how it can drive the desired outcome and how to pick the right ones for your business.
What is a metric and how do we choose one?
A metric is a measure that can be determined or computed. Organizations use metrics to assess the state of a function or process; this could be the conversion of quotes to orders or tracking the progress of a specific quality, such as efficiency of an application. Metrics should be used to inform key stakeholders surrounding a business, such as investors, customers, and all types of employees, including executives and middle managers.
In the software industry, metrics quantify a product or a project — which can help the organizations understand the software performance, its quality, or the productivity and efficiency of development teams. API teams need to track different performance indicators for each API, while infrastructure teams will have completely different metrics from the product and engineering teams.
When monitoring an API, teams monitor metrics based on the position of the API in its product lifecycle. For example, teams will focus on usability and design metrics for a recently-deployed API, then they will focus on reuse of that API. For an API that’s been well-adopted, they will focus on feature metrics and adding more features.
Many organizations struggle to choose appropriate metrics to track. While there is no rule or a particular mantra to choosing a metric, Sudheendra Chilappagari has these easy guidelines for how to identify the right metrics:
- Clearly understood: When defining a metric, ensure that it is easily understood by everyone in the organization. Teams should be able to comprehend what each metric signifies and how it aligns with the organization’s strategy or goals.
- Easily measurable: Every metric should be easily monitored and measured. If it can’t be measured, then it is impossible to understand the behavior it drives.
- Readily accessible: Metrics should be easily accessed. We should be able to compute them easily and make them available in enterprise tools like Google Analytics, Amplitude, Looker, Slack, and others.
- Lead to actionable insights: Good metrics always lead to insights (which lead to action and results). Analyzing them informs decisions like reducing onboarding hurdles, building features that lead to value-added usage, or personalizing user experiences to grow engagement.
- Avoid vanity metrics: Frequently, we come across feel-good metrics like total requests or number of users. Measuring totals gives you only the half picture. To get the full picture, you need the context or the business outcome it drives. For example, you need to measure things like conversions (%), growth (change), customer engagement, and customer satisfaction. Totals = vanity metrics.
How do metrics drive behavior?
Metrics are one of the most efficient ways to determine whether your team, project, product, and your organization are performing as desired. Metrics and measurements can forecast problems and proactively help resolve issues early on.
While the company vision and the employee expectations are usually high-level, metrics give concrete instructions and expectations guidelines. Therefore, they drive behavior and culture within an organization.
When people understand what an attainable performance metric is intended to measure, they will work to achieve the desired result. If the metric is beyond people’s capability to influence or is just outright impossible to attain, then you can expect negative behavior.
Metrics and KPIs are an effective means for measuring the performance of an organization, project, or system against the overall goal and objectives, as well as defining the behaviors. If metrics are not thought through carefully, they may encourage teams to do the wrong thing to achieve a KPI or they can sometimes be contradictory to other teams’ objectives.
The problem with how we choose metrics
A metric should always be tied to a business outcome — but we often see metrics that don’t connect to any outcome. They will simply display what is happening without any behavioral context. Some call these vanity metrics because they might make you look good, but there’s no depth to them. Organizations looking at vanity metrics may follow a process such as this:
- The leadership will come up with a business objective and translate that to a metric that can be measured.
- Management will establish a target and a time frame for achieving it.
- Management will communicate only the target numbers to the teams leaving the behavioral context or the outcome behind.
- An individual or team performance is determined by whether the target is achieved or not.
So, the team members tend to do anything and everything possible in their capacity to achieve the target number rather than focusing on the goal behind the target. They may end up taking shortcuts or engaging in unethical practices.
If we carefully look at the above practice, it redefines the original intent of the metric that was tied to a business outcome. It does it by redefining the metrics as:
- A number or a target: A quantifiable metric makes it easy for the management to use it as the only means for communicating a goal. This lacks communicating the expected outcome or the expected behavior.
- A means to evaluate the performance: In the absence of a well-articulated outcome, it now becomes easy for the leaders to use that target metric to evaluate the performance of the team, individual, or system.
- A motivation: Using pure numerical metrics for communicating the target or evaluating the performance misleads the teams and encourages them to find shortcuts or incorrect practices to achieve it. They tend to forget about the intended outcome.
It’s quite normal to translate a complex attribute, such as a system behavior or a business outcome, to a mere metric. But most often, this comes at the cost of losing track of the real end goal and ends in a suboptimal result.
How should we use the API metrics?
Choosing the right indicators is not always straightforward. We have seen many product managers employ the wrong metrics or use so many that evaluating the data becomes a mammoth task. Some best practices to choosing the right metrics include:
- Connect the metrics to business or technical goals. There needs to be an outcome linked to every metric.
- Focus on tracking trends over absolute numbers because trends can provide feedback on whether the team is moving towards or away from the desired behavior or state.
- Establish reporting intervals that make sense for your strategy. Keep intervals short because it can generate lots of options to react and change.
- Don’t use vanity metrics that make you look good to others but do not help you make decisions or strategies.
- Reset and update the metrics when you believe they aren’t driving outcomes.
- Pick only the right set of metrics that can show you the progress you have made in terms of target behavior. Measuring everything is usually a waste of time.
- Select metrics that are quantifiable and qualitative. The selected metric should give a clear picture of the performance of a system or an application.
- Don’t just limit the metrics to financial or customer indicators to show the value. These are important but not sufficient. Often non-financial assets such as product, process, and people indicators are primary value drivers.
Developing suitable metrics is only the first step. To improve long-term organizational performance, metrics must be used appropriately to drive behavioral changes.
First, accountability must be established. There should be an owner that is responsible for the results of each indicator. After reviewing the results, these individuals must be able to easily identify what conclusions to draw. Simply knowing whether or not you beat last year’s numbers is not good enough. Appropriate and challenging, yet realistic targets must be set to ensure that you are working towards optimal performance. Along with that, understanding what led to the situation is important to help with proper decision-making.
When reviewing current performance against the target, it is important to keep the context in mind while highlighting both the positive and negative results. If useful metrics were selected, then positive performance results will indicate that the team has adopted the behavioral changes you were trying to drive towards. Moreover, if there are negative results, it will be easy to determine what actions and behavioral changes are needed to improve the performance.
To facilitate visibility and use of the metrics, a simple, easy-to-interpret dashboard can be developed. Metrics and associated dashboards should be refreshed appropriately and results should be relayed to the entire team. Read this blog about building a KPI dashboard to gain more insight.
For anyone developing and working with APIs, it is critical to track the correct API metrics. Most companies would not launch a new functionality or a product without having the correct instrumentation to evaluate the behavior or the performance.
Similarly, you wouldn’t want to launch a new API without a way to the instrument and track the correct API metrics. There can be different ways of looking at the same underlying metric depending on which team is looking at it. Having both qualitative and quantitative data is key to planning and driving business outcomes. Hence, it becomes even more important in deciding on what metrics to compile.
Be sure to register for MuleSoft Transform in AMER-friendly timezones on December 7 or APAC-friendly timelines on December 9 to discover more about building a successful API-led connectivity strategy.