Most dashboards are pretty and pointless. They lag by weeks, hide the real levers, and collapse when the one analyst who built them goes on vacation. This playbook fixes that. In two weeks you will define the few metrics that move the business, wire real data to a live dashboard, embed it where people work, and build a lightweight governance loop so it stays trusted.

Who this is for and what you will build

This guide is for product teams, marketing leaders, operators, agencies, and founders who need a single source of truth. You will build one live dashboard that shows your North Star metric and 3 to 5 supporting KPIs, refreshed automatically, and visible to the right people. You can use Google Looker Studio, Microsoft Power BI, or another BI tool. The principles are identical.

Outcomes you can expect:

• 1 page that leaders can read in 3 minutes.
• Metrics that tie to revenue, retention, or risk.
• A refresh schedule with alerts when data breaks.
• A quarterly review that retires stale metrics and adds new ones deliberately.

Step 1: Align on a North Star and 3 to 5 KPIs

Dashboards work only when they mirror strategy. Pick one North Star metric that represents customer value and business value. Examples: weekly active teams, orders delivered on time, qualified opportunities created, or repeat purchase rate. Then pick a tiny set of supporting KPIs that explain movement in the North Star. Use the Balanced Scorecard or OKRs to ensure the set is complete but not bloated.

Checklist

• One North Star metric, plain language definition.
• 3 to 5 KPIs that are leading indicators you can influence inside a quarter.
• Written metric dictionary that includes name, owner, formula, source, and warning ranges.
• Tie each KPI to a single OKR or strategic theme.

Anti patterns

• Vanity metrics that go up with spend but not with value.
• Too many KPIs. If you cannot read it in 3 minutes, it is not a KPI set.
• No owners. A metric without an owner will drift.

Step 2: Choose sources and wire data the right way

Pick the smallest set of systems that cover your KPIs. Typical combos:

• Acquisition and engagement: GA4 for web and app events, product analytics like Amplitude or Mixpanel.
• Revenue: Stripe or your payment processor. For B2B, pipeline and closed won from HubSpot or Salesforce, plus invoices from QuickBooks or Xero.
• Operations: Project data from Asana or Jira, uptime from your status tool, shipping or fulfillment from your logistics system.

Connection patterns

• Looker Studio connects directly to Google sources, BigQuery, and many partner connectors. Use native connectors when possible. Use a warehouse like BigQuery if you need to model across systems.
• Power BI connects to hundreds of sources and can refresh on a schedule. Use gateway connections for on prem or files on shared drives.
• If you run product analytics, consider exporting raw events to your warehouse. This future proofs your dashboard and keeps your model transparent.

Minimal viable data model

• Date table with complete calendar and working days.
• Fact tables per domain: sessions or events, orders, subscriptions, deals.
• Dimension tables: channel, campaign, product, region, account tier.

Step 3: Define calculations once and reuse everywhere

Create calculated fields or SQL views so formulas are consistent.

Examples

• Conversion rate: conversions / sessions.
• Activation rate: activated_accounts / new_accounts.
• On-time delivery: on_time_orders / total_orders.
• Net revenue retention: (starting_MRR + expansion - contraction - churn) / starting_MRR.

Guardrails

• Decide on attribution. If you use last non direct click in marketing, write it down.
• Use explicit status fields in your CRM to avoid counting draft or test records.
• Round for display, not storage. Keep raw precision in the model.

Step 4: Build the dashboard for decisions, not decoration

Layout

• Top row: North Star metric with trend and target. Add a short sentence that explains the change since last period.
• Second row: 3 to 5 KPIs with spark lines and target bands.
• Third row: one diagnostic table or chart per KPI that lets you answer one why question without leaving the page.

Design rules

• Use consistent time frames. Default to trailing 28 days or quarter to date depending on your business.
• Put targets on the chart. People need to see the gap.
• Limit color. Use position and labels first. Use color only for alerts or status.
• Add a simple legend that states data currency, refresh cadence, last refresh time.

Accessibility

• Use at least 12 to 14 px labels in embedded contexts.
• Ensure keyboard navigation and descriptive alt text on embedded frames where possible.

Step 5: Refresh, permissions, and embed where people work

Refresh

• Cloud sources can refresh on a schedule. Set daily refresh for operational metrics and hourly only when you truly need it.
• Add alerts for refresh failures. Post to Slack or email the data owner.

Permissions

• Share by role. Executives get read only. Owners get edit and model access.
• Hide raw PII. Aggregate or hash IDs before they reach the dashboard.

Embed

• Put the dashboard in the tools people open every day. Examples: Notion, Confluence, SharePoint, your intranet, or a team homepage.
• If your audience is external, use a public embed option, but only for data you intend to share with the world.

Step 6: QA and governance that keeps trust high

Monthly QA

• Spot check totals against source systems.
• Scan for flat lines that imply broken tracking.
• Review outliers and annotate known events like launches or outages.

Quarterly governance

• Retire one metric. Add one new metric with intent.
• Reconfirm targets and make them visible.
• Update the metric dictionary. Reassign owners if needed.

Incident playbook

• If a data source fails, freeze the dashboard with a visible banner.
• Publish a short note with what failed, interim process, and estimated fix date.

Two tool recipes you can follow today

A) Build in Looker Studio

1. Create a data source for GA4, Google Sheets, or BigQuery. Use your warehouse when you blend multiple systems.
2. Add scorecards for North Star and KPIs. Add target fields and display as reference lines on charts.
3. Add filters for date, channel, or product line. Hide advanced filters from exec views.
4. Enable scheduled email delivery on Monday mornings to a small distribution list.
5. Embed the dashboard in your wiki or site using the provided iframe. Set access to your domain or specific groups.

Tip: Prefer BigQuery as the source when you need row level joins, custom attribution, or faster performance over very large datasets.

B) Build in Power BI

1. Connect to data sources. For files or on-prem systems, configure a gateway so scheduled refresh works.
2. Model relationships. Create a Calendar table and mark it as a date table. Define measures in DAX for core KPIs.
3. Build the layout: North Star at top, KPIs second, diagnostics third.
4. Configure scheduled refresh and failure notifications.
5. Embed in your portal or share as an app for controlled distribution.

Templates you can copy

1) Metric dictionary (paste into your doc tool)

• Metric name:
• Owner:
• Purpose:
• Formula:
• Source systems:
• Data currency and refresh:
• Target and threshold bands:
• Caveats:

2) KPI target bands

• Green: at or above target
• Yellow: within 10 percent of target
• Red: below target

3) UTM conventions for dashboard links

Use ?utm_source=dashboard&utm_medium=internal&utm_campaign=exec-digest on links you put inside the dashboard so you can see usage in your analytics.

4) Quarterly review agenda

• Did the North Star move in the desired direction
• Which KPIs explained the movement
• What decisions will we make this quarter
• Which metric do we retire and which do we add

14 day sprint plan

Day 1 to 2 – Align on North Star and KPIs. Draft the metric dictionary.

Day 3 to 4 – Inventory sources. Decide warehouse or direct connectors. Create service accounts and access.

Day 5 – Build minimal data model. Create calculated fields or SQL views.

Day 6 – First draft dashboard with layout rules.

Day 7 – QA against sources. Fix mismatches.

Day 8 – Add targets and annotations. Write the one sentence narrative under the North Star.

Day 9 – Set scheduled refresh and alerts. Document the refresh window.

Day 10 – Embed in wiki or portal. Pilot with a small group.

Day 11 – Collect feedback. Remove charts that add no decision value.

Day 12 – Security review. Confirm no PII is exposed.

Day 13 – Add email delivery of a weekly snapshot.

Day 14 – Executive readout. Commit to the quarterly governance cadence.

Frequently asked edge cases

We have multiple products with different cadences Create one executive page with rollups and separate product tabs for deep dives. Do not mix incompatible time frames on the same chart.

Our GA4 event names changed last quarter Document the version change in the metric dictionary and adjust your model so the new and old events map to the same concept.

Finance closes monthly and product is daily Publish two speeds. Daily operational dashboard and a month close view that reconciles revenue and cost.

Our data volume is huge Move repeated calculations to the warehouse. Pre aggregate by day or week. Keep your BI tool visuals simple.

Internal resources from WebDev & Design

• Strategy and measurement intersects with trends: Digital marketing trends for 2026.
• Technical hygiene matters to data quality: 14 signs your website is not ready for 2026.
• For teams rolling this out to physical locations, pair with: Google Business Profile tips, tricks to rank and scale.

Sources

• Looker Studio – About connectors and data sources: https://docs.cloud.google.com/looker/docs/studio/about-connectors-data-sources-and-credentials
• Looker Studio – Embed a report: https://docs.cloud.google.com/looker/docs/studio/embed-a-report
• Power BI – Scheduled refresh overview: https://learn.microsoft.com/en-us/power-bi/connect-data/refresh-scheduled-refresh
• Power BI – Publish to web and secure embed: https://learn.microsoft.com/en-us/power-bi/collaborate-share/service-publish-to-web and https://learn.microsoft.com/en-us/power-bi/collaborate-share/service-embed-secure
• GA4 – Recommended events: https://developers.google.com/analytics/devguides/collection/ga4/reference/events
• GA4 – BigQuery Export overview and limits: https://support.google.com/analytics/answer/9358801 and schema: https://support.google.com/analytics/answer/7029846
• Amplitude – North Star Framework: https://amplitude.com/books/north-star/about-north-star-framework
• Google re:Work – Set goals with OKRs: https://rework.withgoogle.com/intl/en/guides/set-goals-with-okrs
• HBS Online – Balanced Scorecard overview: https://online.hbs.edu/blog/post/balanced-scorecard