A metric is a quantitative measurement that describes a phenomenon. In a digital media outlet, a metric may reflect how many people have viewed a news article, how long they spent on it, how many times a video has been played, what percentage of users arrived at a page from a newsletter, or how many readers started a subscription process.
The central idea is simple: a metric converts part of reality into a number. That number is not the complete reality, but it helps us observe it more objectively. The problem is that not all metrics serve the same purpose. Some help determine whether a goal is being achieved. Others help explain why something happens. Others, in fact, can be of little use or even misleading if interpreted out of context.
In a newsroom, measurement only makes sense when it answers a specific question. For example:
• Which sections generate the most loyalty?
• Which formats achieve deeper reading?
• Which types of content attract more new users?
• Which content contributes most to conversion or registration?
Without a clear question, metrics tend to become noise. They are checked out of habit, but they do not help decision-making.
KPIs and diagnostic metrics
Not all metrics have the same value. A fundamental distinction is the one between KPIs and diagnostic metrics.
A KPI (Key Performance Indicator) is a metric directly connected to a relevant objective. If the goal of a section is to increase the number of returning readers, then a metric related to recurrence or visit frequency may function as a KPI. If the goal is to increase subscription conversion, then the conversion rate or the volume of new subscribers attributable to certain content may be a KPI.
A diagnostic metric, by contrast, is not necessarily the main indicator of the objective, but it helps explain why the KPI goes up or down. For example, if conversion drops, it is useful to look at diagnostic metrics such as traffic source, device type, reading depth, behavior at the paywall, or audience composition. These metrics are not the final objective, but they help interpret the situation.
Put simply: the KPI tells you what is happening in relation to a goal; the diagnostic metric helps explain why it is happening.
A newsroom that does not yet distinguish between these two types of metrics tends to make two opposite mistakes. The first is focusing only on one main figure without understanding what drives it. The second is getting lost in many secondary figures without clearly identifying the real indicator of success.
Correlation and causation
One of the most common mistakes in data interpretation is confusing correlation with causation.
The fact that two variables move together does not automatically mean that one causes the other. For example, a certain section may publish many pieces with photo galleries and also obtain many page views. But this does not prove that the gallery causes the performance. It may be that the section covers highly demanded topics, has better homepage placement, or receives more traffic from search engines.
Correlation indicates that there is an association between two variables. Causation, by contrast, implies that one variable produces an effect on another. Demonstrating causation requires greater rigor: more context, proper comparisons, and often experimentation.
For a newsroom, this distinction is important because it prevents premature conclusions. An isolated data point may suggest a relationship, but it is not sufficient by itself to claim that an editorial decision caused a particular result.
Basic digital analytics vocabulary
Before working with data, it is useful to have a minimum shared vocabulary.
User: a person or browser uniquely identified within a measurement system. It does not always correspond to a clearly identifiable real person, but it serves as a basic unit of analysis.
Session: a set of interactions performed by a user within a given period. A session may include several page views, events, or actions.
Event: a specific action recorded by the analytics tool. For example, clicking a button, playing a video, or completing a form.
Source / medium: information about where traffic originates. For example, search engines, social networks, direct access, newsletters, or referrals from other sites.
Conversion: an action that an organization considers valuable. It may be a registration, a subscription, a download, or any behavior defined as a goal.
What makes a metric useful
A useful metric usually meets several conditions:
• it is connected to a real decision
• it is clearly understandable
• it can be measured with reasonable consistency
• it allows comparison
• it is not interpreted in isolation from context
The goal is not to measure more. The goal is to measure better.
Try it yourself
Your editor has just sent you this message after checking the analytics dashboard:
“Great news — our photo gallery from the city marathon got 85,000 page views yesterday. The climate investigation also did well: 12,000 views, but people spent an average of 6 minutes reading it. And the subscriber newsletter? Only 4,200 views — but 67 new subscriptions came in that same day.”
The editorial team is now debating which content type to prioritise next month.
Consider:
- What does each metric tell you — and what does it not tell you?
- If your goal is to grow subscriptions, which content would you argue deserves more investment? What data supports your reasoning?
- The editor says “the marathon gallery clearly drives subscriptions when we publish it.” Is this a valid conclusion? What’s missing?
- What single additional metric would you most want to see for each piece — and why?
There is no single correct answer. The goal is to practise asking the right question before drawing a conclusion from a number.