Weekly KPI digest from your SaaS stack

Seed-to-Series B founders and ops leads at 5-100 person SaaS startups who pull numbers from 5+ disconnected tools every week · Assemble in ~20 min; first full report in <10 min once data exports are ready.

Founder opsEdit flowon X● golden 91

step 1connector

Pull raw metric exports from each SaaS tool

⚡ CRM (HubSpot / Salesforce / Sheet)

step 2ready

Define which KPIs matter and set targets

analytics8 alts

step 3ready

Analyse metrics — variance, trend, anomaly

data-analysis3 alts

step 4ready

Generate structured performance report

performance-report20 alts

step 5ready

Build interactive HTML dashboard

dashboard-builder3 alts

step 6ready

Write the narrative digest for stakeholders

weekly-report8 alts

step 7connector

Distribute digest to team and investors

⚡ bring your own tool

Did this recipe work for you?

What you get

▸Structured KPI report with variance commentary per metric cluster
▸Single-file interactive HTML dashboard (shareable, no login)
▸Narrative digest ready to paste into email, Slack, or Notion
▸Reusable KPI taxonomy + target baselines for future periods

tools: SaaS exports free (native) · optional Composio/Zapier connector from $0-20/mo

tokens: ~$1-3 per weekly report run

≈ $4-12/mo at weekly cadence; under $1 for monthly-only runs

What stays yours

· Review variance notes and confirm the 'why' before sharing — Claude flags anomalies but cannot know about one-off events (a promo, an outage, a sales push) that explain them
· Approve and send the digest; automatic distribution to investors should have a human gate
· Tune the KPI taxonomy each quarter as the business stage and metric priorities shift

✓ Last verified 2026-06-11claude-sonnet-4-6 (subagent, test-mode fixtures)

✓ step 1 · Pull raw metric exports from each SaaS tool

✓ step 2 · Define which KPIs matter and set targets

✓ step 3 · Analyse metrics — variance, trend, anomaly

✓ step 4 · Generate structured performance report

✓ step 5 · Build interactive HTML dashboard

✓ step 6 · Write the narrative digest for stakeholders

✓ step 7 · Distribute digest to team and investors

view sample outputs from this run

step 2 · Define which KPIs matter and set targets

# KPI Taxonomy — Formly (B2B SaaS, ~$180k ARR stage)

Applied skill: analytics
Source: raw-metric-exports.md

---

## Taxonomy Design Principles

At the $150k–$300k ARR stage a company needs to distinguish three layers:
1. **North Star** — one metric that best proxies whether the product is creating durable value
2. **Leading indicators** — move before revenue moves; weekly cadence
3. **Lagging indicators** — confirm the story; monthly cadence

Anything not on this list is either logged for operational use or dropped to reduce noise.

---

## North Star Metric

| Metric | Definition | Why |
|-…

step 3 · Analyse metrics — variance, trend, anomaly

# Metric Analysis — Week of 2026-06-02 to 2026-06-08

Applied skill: data-analysis
Source: raw-metric-exports.md, kpi-taxonomy.md, metric-targets.md

---

## Summary Scorecard

| ID | KPI | Actual | Target | Status | WoW Δ | MoM Δ |
|----|-----|--------|--------|--------|--------|--------|
| R1 | Net New MRR | $2,780 | $2,300 | 🟢 GREEN | +$890 (+47%) | — |
| R2 | Gross Churn Rate | 4.2% | <5.0% | 🟢 GREEN | -3.7pp | — |
| R3 | New Paying Customers | 14 | 13 | 🟢 GREEN | +3 (+27%) | — |
| A1 | Trial Starts | 312 | 290 | 🟢 GREEN | +51 (+20%) | +13.2% vs May avg |
| A2 | Trial→Paid Conv. | 18.6% | …

step 4 · Generate structured performance report

# Structured KPI Report — Formly
## Week of 2026-06-02 to 2026-06-08

Applied skill: performance-report
Source: metric-analysis.md, variance-notes.md, kpi-taxonomy.md

---

## Executive Snapshot

| | |
|--|--|
| **Reporting period** | 2026-06-02 to 2026-06-08 |
| **North Star: Net New MRR** | **$2,780** (+47% WoW) |
| **Total MRR (end of week)** | $15,240 |
| **ARR run rate** | $182,880 |
| **KPIs green / yellow / red** | 13 / 0 / 0 |
| **Primary risk** | Monthly gross churn growing faster than new revenue |

---

## Section 1: What Happened

### Revenue

Net New MRR came in at $2,780 — the hi…

The problem

When I run a 10-50 person startup, I want my metrics from Stripe, GA4, HubSpot, and Slack automatically pulled, narrated, and formatted into a weekly digest, so that I stop wasting 3-8 hours every Monday copy-pasting numbers into a Google Sheet and writing commentary nobody reads.

Cybernetic layers

Memorypartial

Period-over-period comparison requires storing prior-week numbers. Use a Google Sheet, local CSV, or Notion page as the running ledger. No harvested memory artifact covers multi-tool KPI history yet — this is the top supply frontier for this recipe.

Backgroundbyo

For fully automatic Monday-morning delivery, pair with a cron trigger (Claude Code CronCreate, GitHub Actions, or Zapier schedule). The recipe itself is batch-shaped and runs in one session; scheduling is external.

Harnessready

Claude Code is the runtime. Install artifact skills into ~/.claude/skills/ and invoke by name. Steps 1 and 7 are connector steps that route through your existing tool accounts.

Click any step to see what fills it — and the ranked alternatives you can swap in.