Build a Four-Touchpoint Weekly Rhythm From a Messy Backlog
Objective: Given a disorganized 12-item experiment backlog with no owners or scores, build the actual weekly rhythm artifact: an async-scored backlog doc, a Friday review agenda template, and a portfolio-rebalance checklist, following the lesson's four-touchpoint structure.
You've joined the growth team at Awfis Space Solutions, the NSE-listed Indian flexible-workspace company, three weeks after a reorg. The old team ran no fixed cadence; experiments got discussed 'whenever someone had time.' Your first task is standing up the four-touchpoint rhythm from the lesson before the team's next planning cycle.
Score the 12-item backlog with ICE, build the Monday async doc, the Friday review agenda, and the biweekly portfolio-rebalance checklist.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, familiar, sorting and formulas handle ICE math without extra tooling
Free, shareable, comment threads support async pre-reads
The process
2 steps
Step 01 of 02
The lesson's live-prioritization failure mode wastes meeting time on math one person could do alone; scoring happens before Monday's 10-minute confirmation, not during it.
Given 12 unscored backlog items (headline, funnel stage, rough effort), which 3 to 5 make this week's cut once ICE-scored?
Procedure
- Import the 12 items with a blank ICE score column each
- Score Impact, Confidence, Ease for each row (1-10), most teams score independently then average
- Sort descending by ICE product, top 5 become Monday's confirmed list
- Flag the bottom 3 as 'parking lot', explicitly deferred, not silently dropped
Awfis Growth Backlog, Week of Aug 24 (ICE-scored, top 5) 1. Onboarding tour skip-rate fix, ICE 486 (I8 C9 E6.75) 2. Coworking-plan comparison page, ICE 420 3. Referral incentive reminder email, ICE 392 4. Day-pass checkout autofill, ICE 336 5. Meeting-room search filters, ICE 315 Parking lot (deferred, not dropped): virtual tour embed, pricing FAQ redesign, Slack integration prompt
Healthy
5 items enter Monday's meeting pre-scored, discussion takes 10 minutes confirming, not debating math
Unhealthy
12 unscored items get argued over live, the meeting runs 40 minutes and nothing ships that week
What this means
Async scoring turns the Monday meeting into a confirmation, not a negotiation, freeing the room's attention for the one or two genuinely ambiguous calls.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Monday meetings run long and re-litigate scores every week | Move ICE scoring to an async doc with a Sunday-night deadline | 30 min |
Step 02 of 02
Balfour's format spends roughly 40 of 45 minutes interrogating what experiments taught the team, each owner presents one learning in under 5 minutes.
Given 3 experiments finishing this week (onboarding tour, referral email, day-pass checkout), what does the Friday agenda look like?
Procedure
- Create one section per finishing experiment: hypothesis, result, is-it-valid check, learning
- Send the doc Thursday afternoon, 24 hours before Friday's meeting
- Require owners to fill their section before the meeting starts
- Reserve meeting time for the 3-question sequence per experiment, not for reading the doc aloud
Friday Review Agenda, Aug 28 (45 min) 1. Onboarding tour skip-rate fix (owner: Rahul, 5 min + 10 min discussion) 2. Referral incentive reminder email (owner: Divya, 5 min + 10 min discussion) 3. Day-pass checkout autofill (owner: Karan, 5 min + 10 min discussion) 4. 5-min buffer for a cross-cutting theme if one emerges
Healthy
Every owner arrives having already written their section, discussion time exceeds presentation time
Unhealthy
The doc gets opened for the first time in the room, presentation eats the whole 45 minutes
What this means
The pre-read requirement is what converts a status meeting into a decision meeting; skipping it collapses the format back into round-robin updates.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Owners read their section out loud instead of discussing it | Enforce the 24-hour pre-read deadline, open with discussion, not presentation | 5 min |
Final deliverable
Three artifacts: an ICE-scored backlog doc, a Friday review agenda template, and a biweekly portfolio-rebalance checklist with at least 3 questions.
See a reference example
Portfolio Rebalance Checklist, Awfis Growth (biweekly) 1. Which funnel stage got the most experiments the last 2 weeks? Is that intentional or just where it's easy to test? 2. Any experiment still 'in progress' after 3 review cycles? Kill or timebox it. 3. Has a past 'sure thing' win stopped moving its metric? Re-test or retire the claim. 4. What's the one funnel leak nobody has proposed an experiment against yet?
Success criteria
You're done when you can:
- Backlog is scored and sorted, top 5 clearly separated from the parking lot
- Friday agenda follows the learning-per-owner, 3-question-discussion structure
- Portfolio checklist includes a zombie-initiative check and a funnel-coverage check