The Simple Marketing Dashboard: 8 Metrics That Actually Matter

Build a marketing dashboard that fits one screen and reads in 30 seconds. The eight metrics that drive decisions, plus templates and the metrics to delete.

The average marketing dashboard fails the most basic test of a useful tool: nobody opens it. Pull up the average B2B SaaS Looker Studio template and you'll find forty-seven widgets across six tabs, half of them showing 'bounce rate by browser version' or 'sessions per user by city,' and exactly zero answering the question the founder actually asks on Monday morning: did last week make us money, and what should we change?

A useful marketing dashboard does the opposite. It fits one screen. It reads in 30 seconds. It surfaces the small number of metrics that lead to decisions and ruthlessly hides everything else. This guide walks through the eight metrics that belong on every marketing dashboard regardless of business model, the metrics to delete, and how to wire it up in Looker Studio, Sheets or an AI tool without spending a week.

Why most marketing dashboards fail

Three pathologies kill dashboards. They all stem from the same root cause: dashboards get built by people optimizing for completeness rather than decisions.

Vanity metric overload

Total page views, total users, total sessions across 'all time' or 'last 90 days'. These numbers always go up because the website has existed longer. They tell you nothing about whether last week was good or bad.

No comparison period

A widget showing '12,438 sessions' is useless without context. Is that good? Bad? Normal? Every number on a dashboard needs a comparison: vs last week, vs last month, vs the same period last year. Without a delta, the number is decoration.

No path to action

If a metric moves up or down, what do you do? If the answer is 'nothing' or 'I'm not sure,' that metric doesn't belong on the dashboard. It belongs in a monthly deep-dive report read once and filed away.

The eight metrics that belong on every dashboard

These eight cover acquisition, behavior and outcomes. Together they answer: are we getting enough traffic, from the right places, and is it turning into money?

1. Weekly sessions with week-over-week delta

Total sessions in the last 7 days vs the previous 7 days, shown as a number plus a percentage change. Compare week-over-week, not day-over-day — daily numbers are too noisy to drive decisions, and most business cycles are weekly. Add a sparkline of the last 12 weeks underneath so you spot trend reversals at a glance.

What it tells you: whether the top of your funnel is growing, shrinking or flat. If sessions are falling for three weeks straight without an obvious cause, that's the alarm — investigate before conversion rate masks the problem.

2. Channel mix

Sessions broken down by acquisition channel: Organic Search, Paid Search, Direct, Social, Email, Referral. Show as a stacked bar across the last 8 weeks. Don't show absolute numbers per channel — show the share of total.

What it tells you: how dependent you are on a single channel. If 70% of your traffic is one paid campaign, you're one budget cut away from a crisis. If organic is shrinking while paid grows to compensate, your cost of acquisition is about to rise.

Watch out for: the 'Direct' bucket inflating. Usually that's GA4 failing to attribute properly because UTM parameters are missing on email or paid traffic. If Direct is more than 25% of sessions, fix tracking before reading anything else.

3. Weekly conversions

The single conversion event that matters most to your business: purchases for ecommerce, demo bookings for B2B, signups for SaaS, qualified leads for service businesses. One number, one delta. Resist the urge to put five conversion events on the dashboard — pick the one that pays the bills.

What it tells you: whether the marketing engine is producing the outcome that matters. Sessions can lie. Conversions can't.

4. Conversion rate

Conversions divided by sessions, as a percentage, with week-over-week delta. Crucial because it separates two very different problems. Falling sessions but stable CVR means a traffic problem (acquisition channel issue). Stable sessions but falling CVR means a site problem (broken checkout, regression, pricing perception).

Industry benchmark: 1-3% for ecommerce, 2-5% for B2B SaaS demos, 8-15% for ungated content downloads. But your own historical baseline matters more than any benchmark.

5. Top 5 landing pages with deltas

The five most visited landing pages of the week, with sessions, conversion rate and delta vs previous week. Landing pages are where users arrive — not 'all pages viewed.' This is where you spot a blog post going viral, a product page suddenly outperforming, or a page that quietly stopped working.

Pair this widget with a 'pages with biggest CVR drop' table updated weekly. That's where the easy wins hide.

6. ROAS or CAC (if you run paid)

If you spend any money on ads, you need one of these two on the dashboard. Ecommerce typically tracks ROAS (revenue / ad spend). Lead-gen tracks CAC (total spend / new customers). Both need a target line so you instantly see whether you're above or below the profitability threshold.

Without this metric, you're flying blind on the most expensive part of your marketing. With it, you catch failing campaigns within days instead of months.

7. Engaged sessions rate

GA4 replaced bounce rate with 'engagement rate': the share of sessions that lasted at least 10 seconds, triggered a conversion, or had two or more page views. It's a more honest measure of content quality. A landing page with 80% engagement rate is doing its job; one with 30% needs investigation.

Use it as a quality signal on the landing pages widget, not as a standalone number.

8. Revenue (or pipeline) per channel

The last and most powerful widget: revenue broken down by acquisition channel for the week. For SaaS or B2B replace revenue with pipeline created or qualified opportunities. This is the metric that ends arguments about which channel deserves more budget.

If you can only build one revenue widget, this is it. Channel-level revenue is what justifies the next quarter's marketing investment.

The metrics to delete from your dashboard today

An honest dashboard is defined by what it leaves out. These metrics belong in a quarterly review document, not on the screen you check weekly.

• Total users (all time) — only goes up, tells you nothing about now.
• Bounce rate — replaced by engagement rate in GA4 and was always misinterpreted.
• Pages per session — almost never correlates with revenue.
• Average session duration — heavily skewed by outliers and bot traffic.
• Sessions by browser, OS or screen resolution — engineering debug data, not marketing.
• Geographic breakdown beyond top 3 countries — actionable only when expanding internationally.
• Real-time users — psychologically addictive, operationally useless.
• Demographic charts (age / gender) — sample-based, unreliable, rarely drive a decision.

How to build it: three options

Option 1: Looker Studio (free, 2 hours)

Looker Studio (formerly Data Studio) connects to GA4 natively. Start from the official 'GA4 Acquisition' template, delete two-thirds of the widgets, and add the eight above. Set the date range filter to 'last 7 days vs previous 7 days' as the default. Schedule a weekly email export every Monday at 9am so the report shows up before you remember to look.

Pitfalls: Looker Studio's GA4 connector samples data above certain thresholds and is occasionally slow. For a dashboard refreshed once a week this is fine; for hourly monitoring it's not.

Option 2: Google Sheets (free, 4 hours)

If you want full control and don't mind a quarterly maintenance pass, the GA4 Sheets add-on pulls raw data into a spreadsheet you can shape however you want. Better for finance-style dashboards that mix marketing and revenue data from your billing system.

Option 3: AI analytics tool (60 seconds, paid)

If building and maintaining a dashboard isn't your job, an AI analytics tool delivers the same eight metrics in plain English, daily, without you opening anything. The trade-off is monthly cost (typically €20-€80 for a single property) vs the hours saved building, maintaining and actually reading the dashboard.

Cadence: when to look at the dashboard

Even a perfect dashboard fails if it's checked at the wrong time. The pattern that works for most owner-operated businesses:

• Monday morning, 10 minutes: scan all eight metrics, note anything down more than 15% vs last week. Pick one investigation for the week.
• Friday afternoon, 5 minutes: confirm the week closed roughly where it should. Adjust paid budgets for the weekend if needed.
• First Monday of the month, 30 minutes: pull the same dashboard at month view, identify the one structural change to make next month.
• Quarterly: full deep-dive with cohort analysis, LTV by channel, content performance — the report nobody reads weekly but matters quarterly.

Common mistakes when building or reading the dashboard

Even with the right metrics, three mistakes undo the value.

First, reacting to one-week noise. A 12% drop in conversion rate over one week is often statistical noise, especially below 50,000 sessions. Treat one-week moves as signals to investigate, not decisions to act on. Two consecutive weeks of movement is when to actually change something.

Second, mixing branded and non-branded organic. If branded search (people Googling your company name) is in the same Organic bucket as non-branded SEO, you can't tell whether your SEO is working or your brand awareness is growing. Segment them with a regex filter on Search Console queries.

Third, ignoring attribution windows. GA4 defaults to 'data-driven attribution' with a 30-day window. Email campaigns and long B2B sales cycles look weak in shorter windows. Make sure your dashboard's attribution model matches your actual sales cycle.

If you're using AI: ask the dashboard, don't build it

The eight-metric dashboard is the right answer for a manually built report. An AI copilot inverts the model: instead of building a fixed dashboard and reading it on a schedule, you ask the questions the dashboard would have answered, in plain English, on demand. "How did last week compare to last week?" replaces opening Looker Studio. "Which channel drove the most revenue, and how does that compare to spend?" replaces the ROAS widget. The eight metrics still drive the questions, but the chat replaces the canvas.

Metriko answers the eight-metric questions automatically every Monday and lets you follow up in plain English when something looks off. No widgets to maintain, no Looker Studio to wrestle.

Keep reading

Want to go deeper? These related guides build on what you just read: [how to measure marketing ROI](/guides/how-to-measure-marketing-roi), [where your customers come from](/guides/where-do-customers-come-from), [ecommerce KPIs explained](/guides/ecommerce-kpis-explained), [AI for Google Analytics](/guides/ai-for-google-analytics).