Startup Metrics That Matter (And How to Track Them in DenchClaw)

Most startup dashboards are full of metrics that feel good and don't tell you anything useful. Page views. Registered users. Engagement rate. These are rearview mirror metrics — they describe what happened, not whether your business is healthy.

The metrics that actually matter are much fewer, much harder to fake, and usually more uncomfortable to look at.

The Metrics That Actually Matter at Early Stage#

1. Retention (W1 / M1 / M3)#

Nothing reveals product-market fit faster than retention. Specifically:

• Week 1 retention: What percentage of users return in the first week?
• Month 1 retention: Still using at 30 days?
• Month 3 retention: Still using at 90 days?

A product with 40% M3 retention is a product people have built a habit around. That's real. A product with 3% M3 retention — even if it has 10,000 registered users — is not working.

How to track in DenchClaw:

-- Users who activated in month X and returned in month X+3
SELECT 
  COUNT(DISTINCT activated.id) as cohort_size,
  COUNT(DISTINCT retained.id) as retained_count,
  ROUND(100.0 * COUNT(DISTINCT retained.id) / COUNT(DISTINCT activated.id), 1) as retention_rate
FROM v_people activated
LEFT JOIN v_people retained ON retained.id = activated.id
  AND retained."Last Active" >= activated."First Active" + INTERVAL '90 days'
WHERE activated."First Active" >= '2026-01-01'
  AND activated."First Active" < '2026-02-01'

2. Activation Rate#

Of users who sign up, what percentage actually experience your product's core value? This is the percentage who complete your "aha moment."

For DenchClaw, activation is: user has created at least one CRM entry and sent at least one message to the agent. Until they've done both, they haven't experienced what DenchClaw actually is.

Most products have terrible activation rates — often under 20%. Improving activation is almost always higher leverage than improving acquisition.

3. Revenue / ARR#

For B2B products: annual recurring revenue and its growth rate. For consumer: daily or monthly revenue.

The most honest version of this metric: net new MRR. This is (new MRR added) minus (churned MRR). A business where new MRR = churned MRR is running on a treadmill.

4. NPS (Used Honestly)#

Net Promoter Score asks: how likely are you to recommend this product to a friend? Score 9-10 are Promoters; 1-6 are Detractors.

NPS is only useful if you actually collect it from users you haven't specifically selected to be happy. Survey every user, not just your happy cohort.

For DenchClaw: we ask every user who's been active for 30 days. The honest feedback from that cohort is much more valuable than cherry-picked testimonials.

5. Time to Value#

How long does it take a new user to get value from your product? Ideally, this is measured in minutes, not days.

Every minute you add to Time to Value is users you lose. For DenchClaw, the goal is: from npx denchclaw to a working AI CRM in under 10 minutes.

Metrics That Are Usually Vanity#

Total registered users — includes everyone who signed up once and never came back Page views — unless your business model is ads, who cares Social media followers — doesn't convert to users reliably App store downloads — means nothing without retention data Press mentions — nice, doesn't pay the bills

The tell for a vanity metric: would you still track it if no one else would ever see it? If the answer is no, it's for signaling, not learning.

Setting Up a Metrics Dashboard in DenchClaw#

Here's a practical setup for tracking early-stage metrics using DenchClaw:

People object fields to add:

• First Active (date)
• Last Active (date)
• Activation Date (date — when they hit your activation event)
• Is Activated (boolean)
• NPS Score (number 1-10)
• NPS Collected Date (date)
• Revenue MRR (number)
• Source (text)

With these fields, you can query everything that matters:

-- Ask DenchClaw:
"What is our 30-day retention rate this month vs. last month?"
"How many users activated this week?"
"What is our average NPS by acquisition source?"
"Show me our net new MRR for the last 6 months"

Or build a metrics dashboard Dench App (like the one in our analytics widget guide) that calculates these automatically.

The Honest Metrics Conversation#

The most useful thing you can do with your metrics is have the hard conversation they're pointing at. If your Week 1 retention is 20%, don't optimize your onboarding email sequence — talk to the 80% who left.

If your activation rate is 15%, don't A/B test button colors — figure out why 85% of users who were motivated enough to sign up didn't find value in your product.

Metrics are useful as a pointing device. The answer to what they're pointing at comes from users.

Frequently Asked Questions#

What's a good Week 1 retention rate for early-stage B2B SaaS?#

Benchmarks vary by category. For productivity tools, 25-40% W1 retention is acceptable at early stage. Above 50% is strong. Below 20% indicates the product isn't delivering on its promise quickly enough.

Should I track daily active users or monthly?#

Depends on your product's expected use frequency. A daily communication tool should be tracked as DAU. A quarterly reporting tool makes more sense as WAU or MAU. The right metric matches how often users should naturally use the product if it's working.

How do I sync real user activity data into DenchClaw?#

If you have a web product, add a webhook that fires when users hit key events (activation, login, subscription). Have the webhook call dench.agent.run("Update user X's Last Active to today") or use a direct DuckDB write via your backend.

Is it worth using a separate analytics tool (Mixpanel, Amplitude) alongside DenchClaw?#

Yes, for granular product analytics. Use Mixpanel or PostHog for event-level product analytics and user flows. Use DenchClaw for relationship tracking, investor/customer context, and synthesizing patterns across your CRM. They serve different purposes.

Ready to try DenchClaw? Install in one command: npx denchclaw. Full setup guide →