Tracking · Dashboard & KPI Spec

Build a Win-Rate Trend Dashboard

Spec a win-rate trend dashboard that segments wins by rep, service, and deal size so the trend is actionable.

foundermanagerIntermediate⏱ 3-5 hours
When to use
Use when your win rate is moving but you don't know if it's a real signal or noise. Run before any sales process change so you have a baseline to measure against. Best when you have 12+ months of closed deal data.
The prompt
You are a RevOps lead spec'ing a win-rate trend dashboard. You separate signal from noise by segmenting wins, not averaging them.
Agency: [AGENCY_NAME] — [SERVICES] | BI: [BI_TOOL] | CRM: [CRM_TOOL] | Audience: [DASHBOARD_AUDIENCE] | Current win rate: [WIN_RATE]% | Historical data: [HISTORY_MONTHS] months | Deal size bands: [DEAL_BANDS]
Spec the win-rate trend dashboard widget-by-widget — chart type, metric, filter, refresh cadence, action threshold, owner. The dashboard must show trend over time AND segmentation (by rep, service, deal size, source) so the founder can locate the cause of any movement.

- Max 8 widgets
- Always show trend (weekly or monthly) AND a segmentation cut
- Include a moving average (smooth out noise)
- Compare current to last quarter and YoY
- No vanity metrics

Markdown table: Widget # | Widget Name | Chart Type | Metric | Filter | Refresh | Action Threshold | Owner. Add a 3-bullet 'how to tell signal from noise using this dashboard'.
Variables
  • [AGENCY_NAME] — Agency name
  • [SERVICES] — Services sold
  • [BI_TOOL] — BI tool
  • [CRM_TOOL] — CRM
  • [DASHBOARD_AUDIENCE] — Founder, sales mgr
  • [WIN_RATE] — Current win rate %
  • [HISTORY_MONTHS] — Months of historical data available
  • [DEAL_BANDS] — Deal size buckets (e.g., $5k MRR)
Example input
Agency: Northstack — SEO retainers + one-off audits | BI: Looker | CRM: HubSpot | Audience: founder | Win rate: 28% | History: 24 months | Bands: $5k MRR
Example output
| # | Widget | Chart | Metric | Filter | Refresh | Action Threshold | Owner |
|---|---|---|---|---|---|---|---|
| 1 | Win Rate (4-wk Moving Avg) | Line | Win % | Last 26 weeks | Weekly | 3-wk decline = investigate | Founder |
| 2 | Win Rate by Service | Line | Win % per service | Last 12mo | Monthly | Any drop >5pt = act | Manager |
| 3 | Win Rate by Deal Size Band | Bar | Win % per band | Last 6mo | Monthly | Big deals dropping = enterprise issue | Founder |
| 4 | Win Rate by Rep | Bar | Win % per rep | Last 6mo | Monthly | Bottom rep 40% = root-cause project | Founder |
| 8 | Sample Size per Segment | Bar | Deals closed | Per segment | Monthly | <10 deals = trend unreliable | Manager |

Signal vs noise: 1) Don't trust widget 1 alone — segment with widget 2-4. 2) Check widget 8 sample size before reacting. 3) If trend appears in only one segment, fix that segment — don't change the whole process.
Pro tips
  • Always pair a trend chart with sample size — a 'win rate drop' from 30 to 20% on 8 deals is noise.
  • Use a 4-week moving average instead of weekly raw — agency deal volume is too lumpy.
  • Track loss reasons rigidly (5 options max) or your root-cause analysis becomes vibes.
Works with
ClaudeChatGPTGemini
Done with prompts? Time to install the system
Book a STAOS call
Related prompts