How to Measure Content Performance: The Metrics That Matter

The reason most content reporting feels useless is that it collapses those layers into one number. A dashboard showing 40,000 impressions and 900 sessions tells you nothing about whether the problem is that you published too little, reached the wrong people, wrote a weak hook, or built a landing page nobody converts on.

Separate the layers and each number starts pointing at a specific decision.

Content performance measurement is the practice of tracking a piece of content through distribution, awareness, engagement, traffic and demand, so that a change in results can be traced to the layer that caused it.

Why do most content dashboards fail?

Most content dashboards fail because they measure output at the wrong unit and outcomes at the wrong depth.

The wrong unit is the individual post. A single LinkedIn post's performance is dominated by factors you do not control: the time it went out, whether a large account commented early, what else was in the feed. Scoring individual posts trains you to chase format tricks.

The wrong depth is stopping at traffic. Traffic is a middle layer. It sits above awareness and below demand, and on its own it cannot tell you whether you attracted buyers or students.

There is a third failure nobody puts on a dashboard: the distribution never happened. The webinar was recorded, the clips were planned, then week three arrived and nobody posted anything. You will never see that unless you measure distribution as its own layer. A content distribution audit usually surfaces it within twenty minutes.

What should you measure at the distribution layer?

At the distribution layer, measure whether the content you planned to publish was actually published, and how many separate distributions each source asset produced.

The metrics that matter here:

1. Distributions published per source asset. One webinar becoming 14 published items is a very different investment outcome from the same webinar becoming 3.
2. Plan completion rate. Of the distributions you scheduled, what percentage went live? Below 70% and your measurement of every layer above is measuring a plan you did not execute.
3. Channel coverage. How many distinct channels did the asset reach? LinkedIn only, or LinkedIn plus newsletter plus a resurfaced post six weeks later?
4. Time to last touch. The date of the final distribution minus the publish date. If that number is under 10 days, your content is dying in week one.

The source of these numbers is your own scheduler, calendar or plan document, not an analytics tool. Nobody sells you this metric, which is exactly why it goes unmeasured.

How this layer gets misread: teams count planned distributions instead of published ones. A plan with 20 items and 6 published is a plan that failed, but it reports as a productive month if you only count the plan. Count what went live.

What should you measure at the awareness layer?

At the awareness layer, measure how many people had the opportunity to see the content: impressions, reach, video views, email opens and search impressions.

Where each number comes from:

• Impressions and reach: LinkedIn analytics, X analytics, YouTube Studio
• Email opens: your email platform, with the caveat below
• Search impressions: Google Search Console
• Video views: the platform, using its own definition of a view

How this layer gets misread: impressions get treated as a result. They are not a result, they are a denominator. An impression means your content occupied screen space for a moment. Its only real use is as the bottom half of a rate: engagement rate, click-through rate, view-through rate.

Two specific traps. First, impressions and reach are different: 40,000 impressions might be 9,000 people seeing something four times. Report reach when the platform gives it to you. Second, email open rates have been unreliable since mail privacy features started pre-loading tracking pixels, so treat open rate as a directional signal between your own campaigns, never as an absolute.

What should you measure at the engagement layer?

At the engagement layer, measure who did something deliberate: reactions, comments, shares, saves, watch time, email clicks and scroll depth.

Engagement matters for two reasons that have nothing to do with vanity. It is your fastest read on whether the message landed, and on most social platforms it is the input to further distribution, so engagement in the first hour buys you awareness in hour six.

The metrics worth keeping:

• Engagement rate (interactions divided by impressions), because raw counts scale with audience size and tell you nothing about quality
• Comment-to-reaction ratio, because comments cost more effort than a like and signal a nerve you hit
• Shares and saves, the strongest signals available on most platforms
• Average watch time or percentage viewed for video, which is far more honest than view count

How this layer gets misread: high engagement on content that reaches the wrong audience. A post about marketing process that gets 200 reactions from other marketers, when you sell to heads of RevOps, is a well-performing post and a commercially irrelevant one. Before you celebrate an engagement number, look at who engaged. On LinkedIn you can check the job titles of the people who reacted, and it takes two minutes.

What should you measure at the traffic layer?

At the traffic layer, measure how many people left the platform and arrived on a page you control: sessions, new users, entry pages, source and medium, and on-page behaviour.

Where the numbers come from: your analytics platform (GA4, Plausible, Fathom Analytics or similar) for sessions and sources, Search Console for organic clicks and queries, and UTM parameters for anything you posted deliberately.

The traffic metrics that earn their place:

• Sessions and new users per entry page
• Click-through rate from each distribution channel, which requires UTMs
• Organic clicks and average position per URL from Search Console
• Engaged sessions or bounce rate, read as a hook check on the page, not as a quality score

How this layer gets misread: teams treat their analytics source report as the truth. It is a best effort. A meaningful share of real social traffic never identifies itself, which is the subject of the next section. If you report LinkedIn traffic straight from a channel report and then conclude that LinkedIn does not drive traffic, you have made a decision on a number you know to be incomplete.

Why does LinkedIn break content attribution?

LinkedIn breaks content attribution because a large share of clicks that start on LinkedIn arrive at your site with no referrer, so your analytics tool files them as direct traffic.

This happens through several routes, all of them ordinary user behaviour:

1. Someone reads your post on the LinkedIn mobile app, and the in-app browser or the handoff to the system browser strips the referrer.
2. Someone copies your link and pastes it into Slack, WhatsApp or an email to a colleague. That colleague clicks. No referrer, and the original post gets no credit.
3. Someone sees the post, does not click, and searches your company name two days later. That visit is attributed to organic search or direct.
4. Someone opens the link from a saved post or notification surface that sends no referrer information.

This is usually called dark social, and the honest position is that you cannot fully solve it. You can reduce it:

• Put UTM parameters on every link you post yourself, so at least the links you control identify themselves.
• Use a distinct landing path or offer per channel when the decision matters enough to justify the setup.
• Watch direct traffic as a correlated signal. If direct traffic and branded search rise in the same week you pushed hard on LinkedIn, that is evidence, even though it is not proof.
• Ask. A single question on your demo form, along the lines of how did you first hear about us, produces messier but more useful data than any channel report.

What you should not do is quietly assume the gap is zero. Report LinkedIn's contribution as a range, say which part is measured and which is inferred, and move on.

What should you measure at the demand layer?

At the demand layer, measure the actions that indicate commercial intent: newsletter subscriptions, gated asset downloads, demo requests, trial starts, pricing page visits and sales conversations that reference a specific piece of content.

Where the numbers come from: your forms and CRM, not your analytics tool. Analytics can tell you a conversion event fired. Only the CRM can tell you whether that person was a fit.

Metrics worth tracking:

• Conversions per entry page, and conversion rate against sessions on that page
• New subscribers attributable to a specific asset or campaign
• Pipeline-relevant conversions, filtered to your ICP rather than counted raw
• Content mentioned unprompted in sales calls, which is qualitative and worth logging anyway

How this layer gets misread: counting all conversions equally. Fifty newsletter subscribers who will never buy and two demo requests from target accounts are not comparable, and a dashboard that adds them together will point you at the wrong content. Segment by fit before you compare.

Which content metrics actually matter?

The table below maps each metric to its layer, its source, and the specific question it answers.

Two notes on using this table. Do not track all sixteen from day one: pick one metric per layer, get it reliable, then add. And record them per source asset rather than per post, so each row corresponds to the thing you actually invested in.

Why is revenue the sixth layer most teams cannot close?

Revenue is the sixth layer, and most small B2B teams cannot close it honestly because B2B buying involves multiple people, multiple sessions and multiple months, while their tracking captures one visit from one browser.

The specific blockers:

• The buying committee. The person who read your guide is often not the person who filled in the form.
• Cookie lifetimes and consent. A 90-day sales cycle outlives a lot of tracking.
• Dark social, as described above.
• First touch versus last touch. Both are wrong, in opposite directions. First touch overcredits awareness content, last touch overcredits pricing pages and branded search.

If you have a CRM, defined lifecycle stages, decent form hygiene and enough deal volume for patterns to appear, you can build a defensible revenue view. If you close 20 deals a year, you cannot, because the sample is too small for any model to mean anything.

The honest move is to measure demand rigorously, treat revenue as a directional annual check rather than a monthly metric, and say out loud which parts are measured and which are inferred. Calculating content marketing ROI honestly is mostly an exercise in being explicit about that boundary.

How often should you review each layer?

Review each layer on the cadence at which it can actually change.

Weekly revenue reviews produce noise. Quarterly distribution reviews mean you find out in month three that nothing shipped in month one.

What does this look like for a small team?

For a team of one to three marketers, the whole system fits in one spreadsheet with one row per source asset and roughly eight columns: asset name, publish date, distributions published, last touch date, total reach, total off-platform clicks, conversions, and ICP-qualified conversions.

Fill it in every Friday. After a quarter you will have something no dashboard gives you: a ranked list of which source assets returned most per unit of work.

This is the measurement model Distful is built around, because a tool that plans a multi-week distribution campaign is the only thing that can also tell you how many of those distributions actually shipped. The two Distful-coined metrics that come out of it, Distribution Yield and Distribution Multiplier, are simply the distribution and outcome layers divided by each other.

Where to start this week

Pick your last three source assets: a webinar, a guide, a customer story, whatever you actually produced.

For each one, write down two numbers. How many separate distributions did it produce, and how many off-platform clicks and conversions did it generate in total. Most teams find the first number is between 1 and 3, which explains the second number immediately.

If the first number is low, the fix is not better measurement, it is a real content distribution plan with more touches per asset and a longer tail. If you want a concrete version of that, repurposing a single webinar across three weeks is the easiest place to see the difference a distribution layer makes.

Then repeat next quarter with the same two columns. Comparable numbers over time beat comprehensive numbers once.