48.9k Log in Get started
Dashboard

What goes in a content marketing dashboard in Metabase?

A content marketing dashboard measures the whole content program in one place: which clusters earn traffic, how much of that traffic becomes signups, how fast the team ships and refreshes pieces, and which pages are quietly decaying. It sits a level above the page-by-page view in a content performance dashboard — this is the program review, that is the URL-level diagnosis.

For: content leads, demand gen, and marketing ops. Grain: one row per page per day, rolled up to week or month. Source: web analytics export, CMS metadata, and a conversion events table.

What does a content marketing dashboard look like?

Here’s the layout this guide builds. Program-level numbers sit at the top — sessions, signups, and this month’s publishing output — so the weekly review starts with whether the program is growing. Traffic and conversion by cluster come next, because the cluster is the unit you make decisions about; velocity and decay sit at the bottom, where the “what should we work on” conversation happens.

Content marketing dashboard in Metabase showing traffic by cluster, content-attributed signups, publishing velocity, and decaying pages.
An example content marketing dashboard in Metabase, built from web analytics, CMS, and conversion data. Figures are illustrative.

Explore a live example

Click around a real Metabase dashboard below — filter it, hover the charts, and drill into the underlying questions to see how the cards are put together.

Which cards belong on a content marketing dashboard?

The eight below cover the three questions a content program answers for: is traffic growing, does it convert, and is the library being maintained.

  • Organic sessions and signups this month, with session-to-signup rate and this month’s publishing counts (numbers)
  • Organic sessions by content cluster, weekly (line)
  • Content-attributed signups and conversion rate per month (combo)
  • Sessions share by cluster, last 90 days (donut)
  • Signups by content cluster, last 90 days (row)
  • Content velocity — new versus refreshed pieces per month (stacked bar)
  • Decaying pages — largest organic traffic decline versus the prior quarter (row)
  • Top converting pieces — sessions, signups, and conversion rate per page (table)

What data does the dashboard need?

  • A sessions table from your analytics export — GA4’s BigQuery export or a warehouse-native tool — with session_id, landing_page_path, channel, and timestamp.
  • A conversion events table (signup_completed, demo requests) carrying the session_id so signups join back to the content session.
  • CMS metadata per piece: published_at, updated_at, author, and cluster — this is what powers the velocity chart.
  • A content_clusters mapping table of path pattern to cluster name, with an explicit “Unmatched” bucket.
  • Optionally, Search Console impressions and CTR per URL to separate “fewer clicks” from “fewer rankings” when a page decays.

How do you build it?

  1. Land sessions, conversions, and CMS metadata in one warehouse — the Google Analytics guide covers the GA4 BigQuery export, and most CMSs sync via their API or an ETL tool.
  2. Build the content_clusters mapping table and a shared Metabase model that tags every organic session with its cluster.
  3. Join conversions to sessions in that same model using your chosen attribution rule (session-scoped last touch is the usual start), so every card inherits one definition of “content-attributed.”
  4. Add the decay question: trailing 90 days of organic sessions per page versus the prior 90, filtered to pages above a minimum traffic floor and down more than 30%.
  5. Add filters for date range, content cluster, and channel, and schedule a monthly subscription to the content team’s channel.

Example card SQL

Organic sessions, signups, and conversion rate by cluster and month PostgreSQL 
WITH organic_sessions AS (
SELECT
  s.session_id,
  DATE_TRUNC('month', s.started_at)      AS month,
  c.cluster
FROM web_sessions s
JOIN content_clusters c
  ON s.landing_page_path LIKE c.path_pattern
WHERE s.channel = 'organic'
  AND s.started_at >= NOW() - INTERVAL '6 months'
),
signups AS (
SELECT DISTINCT session_id
FROM conversion_events
WHERE event_name = 'signup_completed'
)
SELECT
os.month,
os.cluster,
COUNT(*)                                  AS sessions,
COUNT(su.session_id)                      AS signups,
ROUND(
  100.0 * COUNT(su.session_id) / COUNT(*), 2
)                                         AS session_to_signup_pct
FROM organic_sessions os
LEFT JOIN signups su USING (session_id)
GROUP BY 1, 2
ORDER BY 1, 2;

Metrics

Organic traffic Conversion rate Click-through rate Engagement rate

Integrations

Google Analytics + Metabase Google Search Console + Metabase Ahrefs + Metabase HubSpot + Metabase

Dashboards

Content performance Organic search performance Newsletter growth Marketing funnel

FAQ

How is this different from a content performance or organic search dashboard?
Scope. A content performance dashboard is page-level — rankings, CTR, and engagement for individual URLs — and an organic search performance dashboard covers the whole search channel, content or not. This dashboard measures the content program: whether clusters are growing, whether publishing pace holds, and whether any of it turns into signups. In practice you drill from here into the page-level view when a cluster moves and you want to know which URLs did it.
How do I attribute signups to content?
Decide the model before you build a single card, because every number downstream depends on it. The simplest defensible version is session-scoped last touch: a signup counts for content if the session it happened in — or the last non-direct session before it — landed on a content URL. First-touch flatters top-of-funnel guides; last-touch flatters comparison pages near the decision. Whichever you pick, compute it in one shared model and reuse that model in every card, so the combo chart and the top-pieces table can never disagree about what “content-attributed” means.
How should I group pages into content clusters?
A mapping table beats regex in the query. Create a small content_clusters table (or Metabase model) with a path pattern and a cluster name per row — /guides/% → Guides, /blog/% → Blog — and join sessions to it. When URLs don't encode the cluster, add a cluster field in the CMS and sync it instead. Keep an explicit “Unmatched” bucket and put it on the dashboard: if it grows, your mapping has drifted, and a silent catch-all is how a cluster chart quietly stops meaning anything.
How do I detect content decay without false alarms?
Compare each page's trailing 90 days of organic sessions to the 90 days before, and flag pages down more than ~30% with enough baseline traffic to matter — a floor of a few hundred sessions keeps small pages from dominating the list. The two classic false alarms are seasonality (compare to the same period last year for seasonal topics) and cannibalization, where a newer page of yours absorbed the traffic — which is a routing decision, not decay. Review the flagged list monthly and route real decay into the refresh queue that your velocity chart tracks.
Why don't my numbers match the GA4 interface?
Usually thresholding, sampling, or attribution defaults. The GA4 UI applies data thresholds that hide small rows, and explorations can sample; the BigQuery export gives you raw events with neither. GA4's UI also applies its own attribution model to conversions, which will not match a session-scoped join you build in SQL. Pick the warehouse export as the source of truth, document the known gaps, and stop reconciling to the last percent — the trend and the ranking are what the dashboard is for. See the Google Analytics integration guide for the export setup.
Should content velocity count updates or only new posts?
Track both, as separate series in one stacked bar. On a mature library, refreshing a decayed piece routinely beats publishing a new one — the page has age and links a new URL lacks — so a velocity chart that only counts net-new posts pushes the team toward the lower-return work. Counting refreshes needs an honest definition, though: a meaningful revision recorded in the CMS (a substantial updated_at change, a changed word count, an editor flag), not a typo fix. If refreshes suddenly dwarf new posts, check the definition before praising the strategy.
What's a good session-to-signup rate for content?
There is no universal benchmark worth chasing — informational blog traffic converting at 0.5% can be healthy while a comparison page at 2% is underperforming, because intent differs by an order of magnitude across clusters. That is why this dashboard breaks conversion rate out by cluster instead of reporting one blended number: the blended rate moves whenever the traffic mix shifts, even if no page got better or worse. Compare each cluster to its own history, and treat mix shift as its own finding.