Most content dashboards measure activity, not results. They fill up with numbers that go up reliably and tell you almost nothing about whether the writing did its job. Here are eight metrics that connect content to outcomes, how to measure each, and what to actually do with the reading. Then four metrics you can stop tracking today.

The 8 worth tracking

1. Organic clicks (not impressions)

Clicks from search tell you that people saw your result and chose it. Impressions only tell you that Google showed it somewhere, often at position 40 where nobody scrolls.

Measure: Google Search Console, Performance report, filtered to a page or folder. Watch clicks and average position together.

Act: if a page has high impressions and low clicks, the title and meta description are losing the auction. Rewrite the title to match the actual query intent and you can often double clicks without moving rank at all. If clicks and position are both rising, the page is earning trust; consider expanding it.

2. Assisted conversions

Very few people read a blog post and buy in the same session. Content usually does its work earlier, in research and consideration, then gets credit stolen by the last click (a branded search or a paid ad).

Measure: in GA4, look at conversion paths and the assisted-conversion view, or use a data-driven attribution model rather than last-click.

Act: find the posts that appear early in winning paths and protect them. These are your real demand generators even if last-click reporting calls them useless. Cutting a "low converting" post that assists 200 conversions a month is a common, expensive mistake.

3. Scroll depth and engaged time

Did people read, or did they land and leave? Engaged time and scroll depth separate a post that genuinely held attention from one that got an accidental click.

Measure: GA4 reports average engagement time per page; scroll tracking needs a custom event (fire at 25, 50, 75, 100 percent). Many analytics tools include scroll events out of the box.

Act: if readers consistently stop at 40 percent, something there is failing (a wall of text, a slow embed, a section that overpromises in the heading). Open the page, scroll to that point, and fix what you find. Long engaged time on a thin page is also a signal: people are hunting for an answer you buried.

4. Return readers

A first-time visitor might be an accident. Someone who comes back chose you. The ratio of returning to new readers is one of the better signals that your content has actual pull rather than just SEO luck.

Measure: GA4 new vs returning users, ideally segmented to your blog or resource section.

Act: if returns are low, you are renting traffic from search, not building an audience. Add reasons to come back: a newsletter, a series, updated evergreen pieces. If returns are healthy, you have something worth turning into a subscription relationship.

5. Branded search lift

When content works at scale, more people start searching for you by name. Branded search is a slow but honest measure of brand awareness that content created.

Measure: Search Console, filter queries containing your brand name. Track the monthly trend over quarters, not weeks.

Act: rising branded search usually trails a period of strong publishing or a piece that got shared widely. Correlate the lift with what you shipped and do more of whatever preceded it. Flat branded search despite heavy output means your work is answering questions but not making an impression.

6. AI citations and referral mentions

A growing share of research now happens inside AI assistants and AI overviews. Being cited or quoted there is the new equivalent of ranking, and it often happens without a click.

Measure: check whether assistants cite your pages for relevant prompts; watch for referral traffic from AI products in your analytics; track brand mentions in generated answers. Some SEO tools now report this directly, and Austen scores content for GEO (how likely it is to be surfaced and cited by AI) alongside traditional SEO.

Act: content that gets cited tends to be clearly structured, factually specific, and easy to extract a quote from. If you are invisible in AI answers, tighten your headings, add concrete data, and make claims quotable in a single sentence.

7. Email signups from content

A signup is a small, real commitment and a permission to keep talking. Attributing signups to the content that produced them shows which topics convert curiosity into a relationship.

Measure: tag signup forms by page or by content cluster. Even a simple hidden field recording the source URL gives you per-post signup counts.

Act: double down on the topics that drive signups, not just traffic. A post with 2,000 visits and 80 signups is usually worth more than one with 20,000 visits and 5. Move your best-converting offer onto your highest-traffic posts.

8. Pipeline influenced

For B2B especially, the question executives actually ask is whether content touched deals that matter. Pipeline influenced ties reading behaviour to opportunities and revenue.

Measure: connect your CRM to web analytics so you can see which accounts and contacts consumed content before and during a deal. Report the pipeline value of deals that touched content at least once.

Act: show this number to anyone who thinks content is a cost centre. Use it to decide which funnel stage needs more material. If lots of early reading never reaches pipeline, your middle-funnel content is the gap.

The 4 to stop tracking

Raw pageviews without intent

Pageviews count visits, not value. A viral post about an unrelated topic can flood your dashboard with traffic that never buys, subscribes, or returns. Reporting total pageviews as a success metric encourages chasing volume for its own sake. Keep pageviews as a denominator (for conversion rates), not as a headline.

Social impressions

Platforms count an impression when a post scrolls past the screen, sometimes for a fraction of a second. Impressions are cheap, inflatable, and almost completely disconnected from outcomes. If you must measure social, use saves, shares, and clicks through to owned properties, which require an actual decision.

Word count

Long content sometimes ranks well, so people concluded that length causes ranking. It does not. Length correlates with thoroughness, and thoroughness is what helps. Tracking word count as a goal produces padded 2,500-word posts that answer a question that needed 600 words. Measure whether the piece fully answers the query, then use exactly as many words as that takes.

Bounce rate (as commonly misread)

A high bounce rate is treated as failure, but a single-page visit is often a success. Someone searches "what time does the shop open", lands on your hours page, gets the answer, and leaves happy. That is a bounce and a win. GA4 already de-emphasised bounce rate in favour of engagement for this reason. If you still look at it, only do so on pages where a second action is genuinely expected.

A practical takeaway: pick the two metrics from the first list that map to your current goal (often assisted conversions and email signups for early-stage teams, pipeline influenced and branded search for established ones), put them on a single dashboard, and delete the four vanity metrics from every report you send. A smaller set of honest numbers will change what you publish faster than a crowded dashboard ever has.

8 Content Metrics That Matter (and 4 to Stop Tracking)