The cadence that turns marketing from a list of requests into a sequence of removed constraints.
Written for Teams working hard on a backlog assembled from whoever asked most recently.
The bottleneck
Most marketing backlogs are prioritised by who asked loudest and most recently. That produces constant activity across many small items and no measurable movement on the thing actually limiting growth.
A fixed cycle forces one question at the start: what is the single biggest constraint right now, on evidence. Everything else waits. That is uncomfortable and it is where the compounding comes from.
Why it happens, five levels down
- Effort is high and results are flat.Why? Work is spread across many low-impact items.
- Priorities are set by request volume.Why? There is no evidence-based prioritisation method.
- Nobody ranks constraints by impact.Why? Nobody measures constraints in the first place.
- Measurement is not part of planning.Why? Planning happens in meetings, not against data.
- There is no operating rhythm.Root cause. Without a cycle, the loudest request wins every time.
The diagnostic
Run this before planning any work. It tells you whether this is your constraint or whether something else is.
- List everything currently in progress. Rank by likely revenue impact. Note how many are low.
- Ask what the single biggest constraint is. If answers differ, that is the first problem.
- Check whether any work in the last quarter was measured against a baseline.
- Identify who can say no to a request. If nobody can, the cycle will not hold.
- Check whether learnings from finished work are recorded anywhere.
- Count how many things started in the last quarter versus how many finished.
The build sequence
| Stage | Work | Timing |
|---|---|---|
| Week 1 | Diagnose. Establish the current constraint from evidence, not opinion. | |
| Week 2 | Prioritise. Choose one constraint, record the baseline, write down what you are not doing. | |
| Weeks 3–4 | Execute. Work on the chosen constraint only. | |
| Week 5 | Test. Measure against the baseline. Be willing to conclude it did not work. | |
| Week 6 | Report, document the learning, and choose the next constraint. |
KPI watchlist
| Metric | Source | Read it as |
|---|---|---|
| Movement on the chosen KPI | Whichever system owns it | Whether the cycle worked |
| Learnings recorded | Cycle log | Whether the team compounds |
| Backlog quality | Backlog review | Whether prioritisation improved |
| Cycles completed on time | Project record | Whether the rhythm holds |
Record a baseline and a date range for each before any work begins. Reconstructing a baseline afterwards is usually impossible.
Four ways this goes wrong
- Running two constraints at once. You learn nothing about either.
- Extending a cycle because the work is not finished. Cut scope instead.
- Skipping week five because the result looks poor. That is the most valuable week.
- Letting an urgent request restart the cycle. Log it for the next one.
Pick one constraint and hold it for six weeks
The discipline is refusing everything else, and that is where the compounding is.
