Custom Business Dashboards: What Should You Build First?

Most businesses don't need a bigger report. They need a quicker answer to a question they ask every week. A good first dashboard is built around that question, not around every number your software can export. This guide walks through how to scope version one so it actually gets used, and what to leave for later. ## Start with one recurring decision Before choosing charts, write down the decision the dashboard should support and how often you make it. For example: - "Every Monday, we decide which open quotes to chase." - "At month-end, we check whether cash coming in covers what's going out." - "Each morning, the office manager assigns today's jobs." If you can't name the decision, the dashboard will turn into a wall of numbers nobody acts on. One clear decision gives you a test for every metric: does it change what we do? ## Choose five metrics or fewer Five isn't a magic number, but it forces useful trade-offs. Aim for a mix: - **One or two outcome metrics** that show results, such as revenue booked this week or cases signed this month. - **Two or three leading metrics** that move earlier and are easier to influence, such as open quotes, average response time to new inquiries, or overdue invoices. - **One health check**, such as when the data was last updated, so everyone knows the numbers are current. Write a one-line definition for each metric. "Lead," "active client" and "revenue" often mean different things to sales, operations and accounting. Agreeing on definitions up front prevents most dashboard arguments later. **Illustrative examples (not client data):** | Business | Decision | Possible first metrics | | --- | --- | --- | | Home services company | Which quotes to follow up this week | Open quotes, quote-to-job rate, average job value, overdue invoices, weekly revenue | | Marketing agency | Which accounts need attention | Retainer revenue, hours used vs. budget, overdue deliverables, upcoming renewals | | Law firm | Where intake is getting stuck | New inquiries, time to first response, consultations booked, missing intake documents | ## Check your data sources before you design anything The fastest way to stall a dashboard project is to discover halfway through that the data isn't where you thought. For each metric, note: 1. **Where it lives today.** Your CRM, accounting software, a spreadsheet, an e-commerce platform, or someone's head. 2. **How to get it out.** An API or built-in connector is ideal. A scheduled export can work. Manual copy-and-paste usually means the dashboard won't stay current. 3. **How clean it is.** Duplicate customers, missing dates and inconsistent categories need fixing at the source, or the dashboard will faithfully display the mess. 4. **How fresh it needs to be.** Most management dashboards are fine with daily or hourly updates. Near-real-time data adds cost and complexity, and your tools may cap refreshes anyway. For example, Microsoft's documentation states that Power BI limits semantic models on shared capacity to eight scheduled refreshes a day, with up to 48 on Premium, PPU or Fabric capacity. Starting with one or two data sources keeps version one manageable. You can connect more once the first view is trusted. ## Decide who needs access, and to what List who will use the dashboard and what each person should see. An owner may want everything; a team lead may only need their own pipeline. If the dashboard will show salaries, client details or other sensitive data, plan access controls from the start rather than adding them later. ## A practical version one A first release that gets used often looks like this: - One dashboard page focused on one decision - Up to two data sources - Five key metrics with written definitions - A visible "last updated" time - Basic sign-in for the people who need it That's deliberately small. It's also roughly the scope of DartSol's [Custom Dashboards starter package](/custom-dashboard), because a focused first version is easier to check against your real numbers and quicker to improve. ## What to leave for version two Once people rely on the first view, these are common next steps: - **Drill-downs** from a summary number to the records behind it - **Targets and alerts**, such as a notice when overdue invoices pass a threshold - **More data sources**, like your marketing or support tools - **Forecasts**, once you have enough clean history to make them meaningful Adding these later, based on how the dashboard is actually used, avoids building features nobody opens. ## Common mistakes to avoid - **Tracking what's easy instead of what matters.** Page views and follower counts are simple to pull but rarely change a weekly decision. - **Skipping definitions.** If two people calculate a number differently, they'll stop trusting the dashboard. - **Relying on manual updates.** If someone has to paste numbers in every Friday, the dashboard will go stale the first busy week. - **Designing before checking access.** Confirm you can connect to each data source before investing in layout. ## Next step If you'd like help scoping your first dashboard, [tell us what you track today](/custom-dashboard#discuss) and which decision you want to make faster. We'll reply by email with questions and a written quote for the agreed scope. --- **Sources** - Microsoft Learn, ["Data refresh in Power BI"](https://learn.microsoft.com/en-us/power-bi/connect-data/refresh-data) *The cover image is a conceptual illustration, not a screenshot of a client project.*