On this page
Start with the decision, then the audience
Ask what decision the dashboard should help someone make. “Which channel gets more budget next month” is a decision. “Show me everything” is not. Each metric on the page should be there because a change in it would trigger a conversation or an action.
Then pick one audience. An owner wants revenue, cost per customer and trend. A paid media specialist wants spend, cost per result and the campaigns driving them. Build a core view for one and a shorter summary for the other, not one page for all.
Write the decision at the top of the page in a sentence. “Use this to decide where to move next month’s budget” keeps people honest about what to add, and gives you a reason to delete a chart that does not help.
Choose five to eight numbers
A workable set for a business running ads: spend, conversions, cost per conversion, revenue, return on ad spend or blended CAC, and one site metric such as conversion rate. Add organic clicks from Search Console if SEO matters. Define each one in writing, because “conversion” means different things in GA4 and an ad platform.
If you are unsure which to pick, marketing KPIs maps metrics to goals. Examples that work as templates are a one-page weekly view, a channel comparison table, and a funnel strip from sessions to leads to customers.
Add context to every number: a comparison period, a target and a plain colour rule for when it is off. A figure with no comparison forces the reader to guess whether it is good, and most will not bother.
Is your tracking counting every conversion?
Runs tracking health checks on your pixels and tags. Sends a weekly performance review.
Build it in a free tool first
Looker Studio is a free option and connects to GA4, Google Ads and Search Console directly. Meta, LinkedIn and TikTok data usually needs a connector or a scheduled export into a Google Sheet, which Looker Studio can then read. A spreadsheet with a pivot table is a legitimate first dashboard.
Keep the page simple: totals as scorecards at the top, a trend line for the primary KPI, one table of channels or campaigns, and a date range with a comparison period. Use consistent UTM names so channels group correctly, which the UTM builder helps with.
Put definitions in a hidden tab or a footnote. When someone asks what “conversion” means on the page, the answer should be a click away, and it should be the same answer as last month.
Know what each source is counting
Platform numbers and GA4 will not match. Each platform credits itself using its own attribution window and view-through rules, and GA4 uses its own model. Show both and label them, or pick GA4 or your store as the source of truth for revenue and use platform data for spend and delivery.
The reasons are covered in cross-channel reporting. A blended view of total revenue over total spend, as in blended ROAS, avoids the double counting.
Document known gaps on the page itself. A one-line note such as “Meta conversions include view-through, GA4 does not” prevents a recurring argument and builds trust in the rest of the numbers.
Dashboard layouts worth copying
The weekly executive view has six scorecards across the top: spend, revenue, conversions, cost per conversion, blended ROAS and new customers. Under them sits a trend line for the primary KPI with last period overlaid, and a short table of channels.
The paid media view is a table of campaigns with spend, results, cost per result and change from last week, sorted by spend. Add a column for status or issues so a disapproved ad or limited budget is visible.
The funnel view moves from sessions to engaged sessions, to leads or add-to-carts, to customers, with the rate between each step. It shows where people leave, which is where you work next. If you want an SEO tab, add clicks, impressions and top queries from Search Console.
Start with the weekly executive view. Add the others only when someone asks a question the first one cannot answer. A dashboard that grows by request stays useful, and one built to be complete on day one rarely gets used.
Keep it useful
Set a refresh schedule, assign an owner and look at it on a fixed day. Retire any panel nobody has opened in a month. People ask whether ChatGPT can create a dashboard. It can help write formulas, queries and layout ideas, but it does not connect to your accounts to build one that refreshes itself.
If you would rather get the weekly readout than maintain the page, Sprites reads your connected channels and sends a performance review on a schedule, with changes proposed as approval cards.
Questions
How can I create a dashboard?
Decide the decisions it supports, pick the audience, choose five to eight metrics, connect your data sources in a tool such as Looker Studio or a spreadsheet, and lay them out on one page. Set a refresh schedule.
What are some KPI dashboards?
Common ones are a weekly performance view, a channel comparison, a lead funnel, an ecommerce revenue and CAC view, and an SEO view of clicks and rankings. Each is built around a goal.
Can ChatGPT create a dashboard?
It can draft the metric list, spreadsheet formulas and queries. It does not pull your live data into a self-refreshing dashboard. You still connect sources in a dashboard tool.
What are the four types of dashboards?
Usually listed as strategic, operational, analytical and tactical. For marketing, the practical split is an executive summary, a channel view and a campaign-level working view.