Sample marketing dashboard: four worked examples and the build order

A marketing dashboard works when it is built for one audience and one decision. Four standard views cover almost every team: a monthly executive page, a weekly demand operating view, a channel view for the person running that channel, and a trading view for stores. Each needs different tiles, a different cadence and a different level of detail, which is why the single dashboard that tries to serve everyone gets opened once and then ignored. The four worked examples below give the tiles, the cuts, the data source for each number and the decision the view exists to support. All figures named are illustrative.

This page covers the dashboards. For what to automate and what to write by hand, see automated SEO reporting, for the client facing version see SEO client reporting, and for the metric definitions see SEO KPIs.

Example one: the monthly executive page

Audience: the leadership team. Decision: whether the plan is on track and whether to move budget. One page, five minutes to read, same shape every month.

Row	Tiles	Source
Outcome	Revenue or new customers against target, pipeline created, blended acquisition cost	CRM or order system, finance for spend
Funnel	Qualified leads, opportunities, win rate, average deal or order value	CRM
Demand sources	Contribution by channel group: organic, paid, email, partners, direct, events	Analytics joined to CRM, one agreed attribution rule
Leading indicators	Non brand organic sessions, branded search volume, list growth, share of voice	Analytics, Search Console, email platform
Commentary	Three sentences: what moved, why, what we are doing next	Written by a person every month

The commentary is the part that makes this view worth building. A dashboard with no written interpretation puts the work of understanding onto the busiest people in the company.

Example two: the weekly demand operating view

Audience: the marketing team in its weekly meeting. Decision: what to change this week. More detail, shorter horizon, trend lines rather than single numbers.

• Top row. Qualified leads this week against the weekly run rate needed to hit the quarter, and cost per qualified lead.
• Pipeline. Opportunities created, stage conversion rates, and the count of leads waiting on sales follow up.
• Campaigns in flight. One row per live campaign with spend, leads, cost per lead and a status flag.
• Website. Sessions and conversion rate for the five pages that generate most enquiries, plus any page whose conversion rate fell week on week.
• Content shipped. What published this week and what is due next week.
• Alerts. Tracking breaks, forms failing, ad accounts disapproved, pages dropping out of the index. Put this at the top during the first month, because that is when you find out how much of your data was broken.

Example three: the channel view

Audience: the specialist. Decision: which lever to pull inside the channel. One per channel, and the owner designs their own.

Channel	The tiles that matter
Organic search	Non brand clicks and impressions, top queries gained and lost, pages gaining and losing, index coverage, average position by page group, Core Web Vitals
Paid search and social	Spend, impressions, click through rate, cost per click, cost per qualified lead by campaign, creative level performance, share of budget pacing against plan
Email and lifecycle	Revenue from flows against campaigns, delivery rate, open and click rates by segment, unsubscribes, list growth, share of list that bought recently
Social and content	Reach, saves and shares rather than likes, follower growth, clicks to the site, which posts produced enquiries

Example four: the e-commerce trading view

Audience: the trading team, read daily. Decision: what to merchandise, discount or restock today.

• Revenue today and this week against the same period last year, with the comparison period stated.
• Orders, conversion rate and average order value, split by device and by new against returning buyer.
• Traffic and revenue by channel, with paid spend and return on ad spend beside it.
• Top products by revenue, fastest movers, and products out of stock that still have traffic.
• Returns rate and the three products driving it.
• Feed and site health: disapproved products, broken product pages, site speed on the top templates.

Layout rules that decide whether it gets used

• Top left is the one number the audience cares about. Everything else supports it. Most dashboards bury the outcome under channel detail.
• Every tile needs a comparison. A number on its own means nothing. Show previous period, same period last year, or target, and label which one it is.
• State the source and the refresh time on the page. The first question in every meeting is where the number came from. Answer it on the dashboard.
• One definition per metric, written down. A link to a short metric dictionary stops the recurring argument about what counts as a lead.
• Cut the vanity tiles. Impressions, followers and total pageviews belong in the channel view if anywhere. They do not belong on the executive page.
• Design for the screen it is read on. If leadership reads it on a phone, the page has to work in one column.
• Delete what nobody looks at. Review the dashboard every quarter and remove any tile that has not changed a decision.

Where the numbers come from, and why they never match

A marketing dashboard pulls from at least five systems: the analytics platform, Search Console, each ad platform, the email platform and the CRM or order system. They count differently by design. Analytics counts sessions and applies its own attribution; ad platforms count conversions against the click or view they are credited with, including some that analytics assigns elsewhere; Search Console counts clicks from search results only. So pick one system as the source of truth for each metric, name it on the tile, and stop reconciling the rest. Revenue comes from the order system or finance. Leads come from the CRM. Search clicks come from Search Console. Spend comes from the ad platforms. For how the search side of this works in detail, see the Search Console guide and Google Analytics for SEO.

Build order

• Write down the decision the dashboard supports and the person who makes it.
• List the five to eight metrics that decision needs, and nothing more.
• Fix the definition and the source for each one in a shared note.
• Build it once by hand, in a spreadsheet, and use it for two weeks. This is where you find out which tiles nobody needs.
• Then automate it in Looker Studio, Power BI, Tableau or whatever the business already runs, connecting the sources you named.
• Add the commentary field and make one person responsible for filling it in before each review.

What people ask about sample marketing dashboard

What should go on a one page marketing dashboard?

The outcome against target, the funnel that produces it, contribution by channel group, two or three leading indicators, and written commentary. Five to eight tiles and one block of text, which is what people actually read.

Which tool should I build it in?

Use what your company already runs. Looker Studio is the common choice when the data sits in Google products, Power BI when the business runs on Microsoft, and a spreadsheet is the right first version of every dashboard because it is fast to change.

How often should a marketing dashboard refresh?

Match it to the cadence of the decision. Daily for trading and paid media, weekly for demand generation, monthly for the executive page. A daily refresh on a monthly decision invites people to react to noise.

How many metrics is too many?

If the audience cannot say what each tile is for, there are too many. Eight on an executive page and fifteen on a specialist view is a comfortable ceiling.

Should clients get the same dashboard the team uses?

Give clients the executive page plus the channel detail for what they are paying for, with the commentary written for them. Keep the operational alert view internal, because it is a working tool rather than a report.

The tile that is missing from most dashboards

Every view above measures people who already found you. There is now a step before the click that nothing on these dashboards counts. Buyers ask ChatGPT, Gemini, Perplexity, Claude or Google AI Overviews which company to use, and the reply names two or three businesses and cites the pages it drew on. Those five answer differently from each other, so being named by one tells you nothing about the others, and a buyer who reads an answer that does not include you never becomes a session in your analytics. It is countable in the same way as anything else on this page: a fixed question set, the share of answers naming you, and the sources cited. AI Knows Us produces those numbers so they can sit beside the rest. See how it works.