kpi-dashboard-design

Overview of kpi-dashboard-design skill

What the kpi-dashboard-design skill does

The kpi-dashboard-design skill helps you design KPI dashboards that are decision-oriented instead of merely data-dense. It is built for planning dashboard structure, selecting the right metrics, matching visual patterns to KPI types, and organizing views for executives, managers, or operational teams. If your real job is “turn a business question into a dashboard people can trust and act on,” this skill is a strong fit.

Who should use this skill

Best-fit users include:

• analysts shaping stakeholder dashboards
• product and operations teams defining what should be monitored
• founders or managers building executive KPI views
• designers or developers who need a better metric framework before implementation
• AI users who want better dashboard recommendations than a generic “make me a dashboard” prompt

The main job-to-be-done

Most dashboard work fails before any chart is drawn: the team has not agreed on KPI definitions, audience, update cadence, or what decisions the dashboard should support. The kpi-dashboard-design skill is most useful when you need help with:

• KPI selection and prioritization
• executive vs tactical vs operational dashboard structure
• visualization choices for trends, targets, comparisons, and alerts
• metric governance so numbers do not contradict each other
• real-time monitoring layout patterns

Why use this instead of a normal prompt

A generic prompt may suggest attractive charts. The kpi-dashboard-design skill gives you a more disciplined frame: KPI hierarchy, SMART KPI thinking, audience-specific dashboard design, and monitoring patterns. That usually leads to better metric selection and fewer “nice dashboard, wrong decisions” outcomes.

What this skill does not do

This is not a data connector, BI tool plugin, or automatic dashboard generator. It will not fetch live data, validate SQL, or replace implementation work in Tableau, Power BI, Looker, Grafana, or custom apps. Use it to improve dashboard design decisions before or during build, not as an end-to-end analytics platform.

How to Use kpi-dashboard-design skill

Install context for kpi-dashboard-design

This skill lives in the wshobson/agents repository under plugins/business-analytics/skills/kpi-dashboard-design. If your skill runner supports remote installs, use the install flow your environment expects for GitHub-hosted skills. A common pattern is:

npx skills add https://github.com/wshobson/agents --skill kpi-dashboard-design

If your environment does not support direct install, open the source at:
https://github.com/wshobson/agents/tree/main/plugins/business-analytics/skills/kpi-dashboard-design

In this case, the repository evidence shows the skill is essentially contained in SKILL.md, so there are no extra helper files you need to inspect first.

Read this file first

Start with:

• plugins/business-analytics/skills/kpi-dashboard-design/SKILL.md

Because there are no resources/, rules/, or reference files for this skill, most of the usable guidance is in that one document. Read it once before prompting so you understand its framing around KPI levels, SMART KPIs, dashboard hierarchy, and monitoring patterns.

What input the skill needs from you

The kpi-dashboard-design usage quality depends heavily on the brief you provide. Give the skill:

• audience: executive, manager, team lead, operator
• business domain: SaaS, ecommerce, support, finance, product, operations
• dashboard purpose: monitoring, strategic review, diagnosing problems, team management
• decisions the dashboard should support
• candidate metrics and their definitions
• update frequency: real-time, daily, weekly, monthly
• constraints: screen size, tool, data freshness, stakeholder preferences
• known pain points: conflicting numbers, too many charts, unclear ownership, alert fatigue

Without these inputs, the output will stay generic.

Turn a rough goal into a usable prompt

Weak prompt:

• “Design a KPI dashboard for my company.”

Stronger prompt:

• “Use the kpi-dashboard-design skill to propose an executive SaaS dashboard for a B2B subscription business. Audience is CEO and VP Finance. The dashboard should support monthly planning and early risk detection. Core metrics available are MRR, net revenue retention, gross churn, CAC payback, pipeline coverage, burn multiple, and logo churn. We want one summary page plus drill-down ideas. Highlight which 5 KPIs deserve headline placement, what visual should be used for each, what targets and comparisons to show, and what definitions must be standardized to avoid contradictions.”

That stronger version improves output because it gives audience, cadence, decision context, metric inventory, and scope.

Best prompt structure for kpi-dashboard-design for Data Visualization

For kpi-dashboard-design for Data Visualization, use this structure:

1. Context: company/team/function
2. Audience: who reads the dashboard
3. Decisions: what actions it should drive
4. Metrics: available KPIs and formulas if known
5. Cadence: real-time, daily, weekly, monthly
6. Output request: layout, KPI ranking, chart recommendations, drill-downs, alerts
7. Constraints: tool limits, screen space, data quality, stakeholder habits

This produces much better design advice than asking only for chart suggestions.

Practical workflow that works

A reliable way to use the kpi-dashboard-design skill:

1. Ask it to classify your dashboard as strategic, tactical, or operational.
2. Have it narrow the KPI set to the smallest credible headline set.
3. Ask for metric definitions and governance risks.
4. Ask for a page hierarchy: summary, drill-downs, exceptions, detail views.
5. Ask for chart types and visual encoding guidance.
6. Ask what should be real-time versus periodic.
7. Ask it to critique your current dashboard or wireframe.

This sequence helps prevent overloading the first answer with too many mixed goals.

What good outputs should look like

Useful kpi-dashboard-design guide outputs usually include:

• a clear dashboard audience and purpose
• 4 to 6 top-level KPIs for executive views rather than 15+
• rationale for each KPI
• a proposed layout hierarchy
• chart recommendations tied to the metric type
• target, trend, benchmark, and variance suggestions
• warnings about metric definition conflicts
• recommendations for alerts or thresholds where relevant

If the output is only a list of charts, prompt again with more business context.

Use cases where this skill is strongest

The skill is especially useful for:

• executive KPI dashboard design
• SaaS health dashboards
• operations monitoring boards
• product retention or cohort views
• redesigning dashboards with conflicting or overloaded metrics
• setting dashboard governance before BI implementation

It is less differentiated for purely decorative UI work or highly technical BI modeling tasks.

Common adoption blockers

Before you install or rely on kpi-dashboard-design, note the likely blockers:

• your team has not defined KPI formulas
• stakeholders want one dashboard for every audience
• no one has agreed on update frequency
• users ask for every available metric
• you expect the skill to build tool-specific dashboards automatically

The skill helps with dashboard thinking, but it cannot fix missing organizational alignment by itself.

kpi-dashboard-design skill FAQ

Is kpi-dashboard-design good for beginners?

Yes, especially if you understand your business context but are unsure how to structure a dashboard. It gives a better framework than starting from blank-page prompting. Beginners still need to supply business goals and metric definitions; otherwise the advice remains high level.

When is kpi-dashboard-design better than ordinary prompting?

It is better when the hard part is KPI choice and dashboard structure, not wording. If your challenge is deciding what belongs on the page, what should be headline vs drill-down, or how to separate executive and operational views, this skill is more useful than a plain “design a dashboard” request.

Can this skill help with existing dashboards?

Yes. A good use case is critique and redesign. Give it your current KPI list, audience, and problems such as contradictory metrics, cluttered layout, or lack of actionability. Ask it to identify what to remove, regroup, redefine, or escalate.

Does it work for real-time monitoring dashboards?

Yes, the source material explicitly covers real-time monitoring patterns. It is a good fit for operations, service health, or live business monitoring where signal clarity and threshold design matter. Be explicit about alerting needs and refresh cadence.

Is this skill tied to a specific BI tool?

No. The current repository structure shows only SKILL.md, with no tool-specific scripts or assets. That makes the skill portable, but it also means you should ask for output formatted for your target tool if you need implementation-ready guidance.

When should I not use kpi-dashboard-design?

Skip it when you need:

• SQL generation or metric pipeline debugging
• pixel-perfect product UI specs
• automated dashboard building inside a BI platform
• deep statistical analysis instead of dashboard structure

In those cases, pair it with other skills or workflows.

How to Improve kpi-dashboard-design skill

Give definitions, not just metric names

A major failure mode is assuming a metric name is enough. “Churn,” “active users,” or “conversion rate” can mean different things across teams. For better kpi-dashboard-design usage, provide:

• formula or business definition
• numerator and denominator
• time window
• segment scope
• owner of the metric

This helps the skill catch conflicts and recommend cleaner dashboard hierarchy.

Ask it to optimize for decisions

Do not ask only for “best KPIs.” Ask:

• what decision each KPI supports
• what threshold or target matters
• what action should follow if the KPI moves
• which metrics are lead vs lag indicators

This improves usefulness because the dashboard becomes action-oriented rather than descriptive.

Separate audiences early

One of the most common dashboard mistakes is mixing executive, manager, and operator needs in one view. Improve kpi-dashboard-design results by stating the primary audience and refusing mixed-scope outputs unless you explicitly want a dashboard suite.

Force prioritization

If you give 20 metrics, the model may try to preserve too many. Ask it to:

• rank all candidate KPIs
• keep only the top 5 for the main page
• move the rest to drill-downs
• explain why some should not be headline KPIs

This usually produces a much more realistic design.

Request critique after the first draft

A strong iteration pattern is:

1. get the first dashboard proposal
2. ask for weaknesses, blind spots, and governance risks
3. ask what could mislead executives
4. ask what should be removed
5. ask for a revised layout

The second pass is often where the kpi-dashboard-design skill becomes genuinely useful.

Provide constraints from the real environment

Better outputs come from concrete constraints such as:

• “single laptop screen for weekly exec review”
• “wallboard visible from 10 feet away”
• “must work in Power BI with limited custom visuals”
• “daily refresh, not real-time”
• “stakeholders distrust ratios without raw counts”

These details meaningfully change layout and visualization advice.

Compare headline KPIs with diagnostic metrics

If the first answer mixes summary and diagnostics, ask the skill to split them:

• headline KPIs: what leadership watches first
• diagnostic metrics: what explains movement
• operational monitors: what needs immediate response

This sharpening step makes dashboard hierarchy much stronger.

Use concrete examples in your prompt

Example improvement prompt:

• “Use kpi-dashboard-design to redesign our customer support dashboard. Audience is support leadership. The current dashboard has 18 charts and users ignore it. Available metrics are first response time, backlog, reopened tickets, CSAT, SLA breach rate, ticket volume by channel, and staffing coverage. We need one summary page for daily review and one drill-down page for queue diagnosis. Recommend what to keep, remove, and group, plus the best chart type for each surviving KPI.”

This works better because it defines audience, pain point, metric list, and required output.

Know the limits of the source material

This skill has useful conceptual coverage, but the repository evidence suggests it is document-based rather than backed by scripts, references, or hard decision rules. Treat it as a strong planning and framing tool, then validate implementation choices with your BI stack, data model, and stakeholders.

Pair kpi-dashboard-design with review from real users

The fastest way to improve output quality is to test the proposed dashboard with the people who will use it. Ask the skill to generate a review checklist covering:

• what decisions the dashboard supports
• whether each KPI is trusted
• whether any chart is hard to interpret
• what is missing for action
• what should move off the main page

That final review closes the gap between a good AI design recommendation and a dashboard people actually use.