How to Track Your Key Sales KPIs in DenchClaw

Track sales KPIs in DenchClaw with live DuckDB dashboards. Monitor pipeline value, win rate, sales cycle length, and more—all from your local CRM.

Mark Rachapoom

March 26, 2026·8 min read

Tracking sales KPIs in most CRMs means waiting for a BI tool to refresh, exporting to a spreadsheet, or paying extra for a reporting add-on. In DenchClaw, your KPIs are SQL queries against a local DuckDB database that returns results in milliseconds. You can ask the AI agent for any metric, see it instantly, and save it as a live dashboard.

Here's how to set up the metrics that actually matter.

The Core Sales KPIs You Need#

Before building dashboards, agree on which numbers matter. The standard set for most sales teams:

Pipeline value — total $ value of open deals by stage
Win rate — closed-won deals as a percentage of all closed deals
Average deal size — mean value of closed-won deals
Sales cycle length — average days from lead to close
Activity metrics — calls made, emails sent, meetings booked per week
Conversion rates — lead → opportunity → close rates by stage
Revenue forecast — weighted pipeline by close probability

Step 1: Make Sure Your Deal Object Has the Right Fields#

Your KPIs are only as good as your underlying data. Check that your deals object has:

Deal Value (number) — in dollars or your currency
Stage (enum) — Discovery, Proposal, Negotiation, Closed Won, Closed Lost
Close Probability (number, 0-100) — for weighted pipeline
Created Date (date) — when the deal entered the pipeline
Close Date (date) — actual or expected close date
Closed At (date) — when it was marked Closed Won or Lost
Owner (text or relation to team members)

Ask DenchClaw to audit your current schema: "Check my deals object and tell me if I'm missing any fields needed to track win rate, average deal size, and sales cycle length."

Step 2: Query Your Pipeline Value#

This is your most-checked KPI. Run it on demand:

"What's my total pipeline value by stage this month?"

DenchClaw translates this to SQL:

SELECT "Stage", 
       COUNT(*) as deal_count,
       SUM(CAST("Deal Value" AS DECIMAL)) as total_value,
       AVG(CAST("Deal Value" AS DECIMAL)) as avg_value
FROM v_deals
WHERE "Stage" NOT IN ('Closed Won', 'Closed Lost')
GROUP BY "Stage"
ORDER BY total_value DESC

Results appear as a table and a bar chart. Ask for a weighted version:

"Show me weighted pipeline value—multiply deal value by close probability for each deal."

Step 3: Calculate Win Rate#

Win rate tells you how effective your sales process is. Ask:

"What's my win rate for deals closed this quarter?"

The agent queries:

SELECT 
    COUNT(CASE WHEN "Stage" = 'Closed Won' THEN 1 END) as won,
    COUNT(CASE WHEN "Stage" = 'Closed Lost' THEN 1 END) as lost,
    ROUND(100.0 * COUNT(CASE WHEN "Stage" = 'Closed Won' THEN 1 END) / 
          NULLIF(COUNT(CASE WHEN "Stage" IN ('Closed Won', 'Closed Lost') THEN 1 END), 0), 1) as win_rate_pct
FROM v_deals
WHERE "Closed At" >= DATE_TRUNC('quarter', CURRENT_DATE)

Track win rate over time: "Show me monthly win rate for the last 6 months as a line chart."

Step 4: Track Average Deal Size and Sales Cycle#

Average deal size helps you understand your revenue mix. Sales cycle tells you how long it takes to close:

"What's my average deal size and average sales cycle length for closed-won deals this year?"

For sales cycle length, DenchClaw calculates the difference between Created Date and Closed At in days, then averages across won deals.

These two metrics together tell you a lot. If deal size is going up but sales cycle is also lengthening, you might be moving upmarket faster than your process can handle.

Step 5: Build a Live KPI Dashboard#

Instead of asking these questions one at a time, build a dashboard that shows all your KPIs at once:

"Build a sales KPI dashboard app that shows: total pipeline value (number card), 
win rate this quarter (number card with comparison to last quarter), 
average deal size (number card), deals by stage (bar chart), 
and revenue closed per month this year (line chart). 
Pull all data from my deals object in DuckDB."

DenchClaw creates a .dench.app with a polished UI that queries your live DuckDB. It appears in your workspace sidebar as a tab you can open anytime. The data refreshes every time you open it.

Step 6: Set Up Activity Tracking KPIs#

Pipeline KPIs tell you about outcomes. Activity KPIs tell you about inputs—and inputs are what you can actually control day-to-day.

Create an activities object if you don't have one, then ask:

"Show me my activity metrics for this week: emails sent, calls logged, 
meetings booked, and LinkedIn outreach sent."

The agent queries your activities object and returns counts by type. Compare week-over-week: "Show me my weekly activity counts for the last 4 weeks as a table."

Step 7: Track Stage Conversion Rates#

This is your funnel analysis. Where are deals getting stuck?

"Show me conversion rates between each pipeline stage: 
what percentage of Discovery deals move to Proposal, 
Proposal to Negotiation, Negotiation to Close?"

To answer this properly, you need stage transition history. If your deals object has a Stage field but no history, ask DenchClaw to start logging stage changes:

"Whenever a deal's Stage field changes, log the previous stage, new stage, deal ID, and timestamp to a stage_history table."

Set this up now and you'll have meaningful funnel data within a few weeks.

Step 8: Set Up a Revenue Forecast#

Weighted pipeline is your most honest revenue forecast:

"Build a revenue forecast for this quarter. For each open deal, 
multiply Deal Value by Close Probability. Sum those for each month 
based on Expected Close Date. Show as a bar chart with this month, 
next month, and the month after."

For a more conservative forecast, use stage-based probabilities instead of deal-level ones:

"What's my revenue forecast if I use these probabilities: 
Discovery 10%, Proposal 30%, Negotiation 60%, Negotiation 80%?"

Step 9: Create a Weekly KPI Report#

Automate your weekly metrics review:

"Every Monday at 8am, generate a weekly KPI summary and send it to me 
on Telegram. Include: new deals this week, total pipeline value, 
win rate MTD, deals closing this week, and any deals with no activity 
in 7+ days."

This becomes a cron job in DenchClaw. You'll get a Telegram message every Monday morning before you open your laptop with a full snapshot of your pipeline health.

Step 10: Compare Team Performance#

If you're tracking an owner field on your deals:

"Show me win rate, average deal size, and deals in pipeline by owner, 
sorted by total pipeline value descending."

This surfaces performance differences without requiring a separate BI tool or manager dashboard. Everything you need is a SQL query away.

Key Metrics Reference#

Quick reference for the most common sales SQL patterns in DenchClaw:

KPI	Query Pattern
Total pipeline	`SUM(Deal Value) WHERE Stage NOT IN ('Won', 'Lost')`
Win rate	`COUNT(Won) / COUNT(Won + Lost)`
Avg deal size	`AVG(Deal Value) WHERE Stage = 'Closed Won'`
Sales cycle	`AVG(Closed At - Created Date) WHERE Stage = 'Closed Won'`
Deals at risk	`Stage != 'Closed' AND Last Activity > 14 days ago`