If your plans go stale before the sprint ends, you don’t need a louder roadmap—you need tarnplanen: a compact, living plan that updates when reality changes. This guide shows exactly how to run it without chaos.
What is tarnplanen?
Tarnplanen is an adaptive planning method: a single source of truth that is deliberately lightweight, continuously updated by evidence, and easy for the whole team to change. Think of it as a plan that behaves like a product—versioned, observed, and iterated.
Unlike static plans, tarnplanen makes change the default. Adjustments aren’t failures; they’re features.
Who needs tarnplanen (and who doesn’t)?
Use tarnplanen if your environment shifts weekly: fast-moving demand, supply variability, short feedback loops, or uncertain priorities. Great fits include product teams, growth marketing, e-commerce ops, logistics, and service operations.
Consider a conventional plan if your constraints are fixed for long periods (e.g., compliance projects with locked scope) where deviation is costly and rare.
Core principles
- Directional clarity: Align on outcomes and constraints—details emerge from evidence.
- Evidence before opinion: Real-time signals trigger reviews and updates.
- Team co-ownership: Those closest to the work can propose and ship plan changes.
- Small reversible steps: Prefer low-risk, high-learning moves you can roll back.
- Transparent history: A change log explains what changed, why, and what happened next.
The tarnplanen playbook (step-by-step)
- Define the 90-day outcome & guardrails. One paragraph on the outcome; list non-negotiables (budget bands, risk limits, service levels).
- Choose 5–7 signals with thresholds. Examples: conversion ±10%, lead time >15% above target, stockout rate >2%, churn >1.5× median, NPS dip ≥5.
- Set decision cadences. Daily 10-min stand-up for micro-tweaks; weekly 45-min review for direction changes.
- Assign roles. Monitors watch signals, Owners propose actions, Approvers green-light guardrail-sized changes, Documenters log context and outcomes.
- Publish the plan where work happens. One page. One link. No attachments that die in inboxes.
- Run monthly retros. Tighten thresholds, prune dead metrics, simplify rituals.
Governance: cadences, guardrails, and the change log
Cadences: Two rhythms keep tarnplanen stable—daily for “course corrections,” weekly for “direction changes.”
Guardrails: Pre-agree bands that don’t need a meeting (e.g., “Channel budget can move ±15% weekly if CAC improves”). Anything larger hits the weekly review.
Change log: Every update records signal → decision → owner → next check date. Over time, this becomes your institutional memory.
Signals, thresholds & actions (cheat sheet)
| Signal | Threshold | Default action |
|---|---|---|
| Conversion rate | ±10% vs baseline for 3 days | Shift spend 10–20% toward winning channel |
| Lead time | >15% above target | Switch to backup supplier; expedite priority orders |
| Inventory stockouts | >2% weekly | Raise reorder point; shorten review cycle |
| Customer churn | >1.5× weekly median | Trigger save offers; escalate root-cause fixes |
| NPS / CSAT | Drop ≥5 points | Spin up 1–2 experiments to address top complaint |
Recommended tooling stack
- Source of truth: Wiki/Notion/Confluence page for the plan + change log.
- Signals: Your analytics/BI dashboard with alerting (email/Slack) at thresholds.
- Execution: Task board (Jira, Linear, ClickUp) linked from each change log entry.
- Comms: Slack/Teams channel #tarnplanen for proposals and approvals.
Worked examples (3 scenarios)
1) E-commerce growth
Signal: Paid search CAC improves 18% week-over-week. Action: Reallocate 15% budget from display to search. Check: If blended CAC stays < target for 7 days, keep allocation; else revert.
2) Logistics delays
Signal: On-time delivery slips to 88% (target 96%). Action: Shift 20% of orders to regional carrier; prioritize high-LTV customers. Check: Review daily until ≥94% for 5 consecutive days.
3) SaaS churn spike
Signal: Weekly churn hits 1.7× median. Action: Launch in-app save flow + outreach to at-risk cohort. Check: If churn normalizes in 2 weeks, keep; otherwise escalate feature fix.
Copy-paste templates
Tarnplanen one-pager
Outcome (90 days): Guardrails (budget bands, risk, SLAs): Signals & thresholds (owners): Decision cadences (daily/weekly): Roles (Monitors | Owners | Approvers | Documenters): Change log URL:
Weekly review agenda
- What thresholds tripped? Why?
- Proposed changes (with smallest reversible option)
- Owner, start date, next check date
- Update change log; link tickets
Change log (example)
Date: 2025-08-27 Signal: CAC improved 18% WoW on paid search Decision: Move 15% budget from display → search Owner: M. Rahman Next check: 2025-09-03 Outcome (on check): Blended CAC -9%; keep allocation
FAQs
Is tarnplanen just “agile” with a new name?
No. Agile focuses on how you build. Tarnplanen focuses on how the plan itself evolves—the process that updates direction based on evidence.
How big should a tarnplanen be?
One page plus a living change log. If you can’t read it in five minutes, it won’t be used.
Won’t frequent changes cause chaos?
Guardrails prevent thrash. Small, reversible changes at a clear cadence create stability, not chaos.
What’s the fastest way to start?
Pick one KPI with a threshold, run the cadence for two weeks, and maintain a change log. Expand once it sticks.
