Six of them, and the test for each is the same: does it cost less to keep than to abandon? Practices that fail that test stop around week nine and nobody announces it.
Most execution advice is a list of things that are true and unsustainable. The useful filter is not whether a practice is good but whether it survives a bad month, because every plan meets one.
The six
One name per commitment. Including shared work. A commitment owned by a team is owned by nobody, and the research is consistent that cross-team commitments are where execution breaks. If three people genuinely share it, one of them owns it and the other two are named as contributors.
A measure that can go down. Activity measures only climb. Meetings held, documents produced, hours spent: all of them go up regardless of whether anything happened. A measure that can fall is the only kind that tells you something you did not already assume.
Fortnightly, not quarterly. The single highest-return change available. A stall caught in week two is a conversation. The same stall caught in week eleven is a post mortem, because the options that would have fixed it have expired.
Write the decision, not the discussion. Most meeting notes record what was said. The useful record is what was decided, who owns it, and by when. Everything else is recoverable; a decision nobody wrote down is not.
Fewer commitments than feels comfortable. Three to five at company level. The constraint is not ambition, it is recall. If people cannot name the priorities without looking, the priorities do not govern what gets worked on, and whatever is loudest does.
Collect progress where the work happens. Chasing updates is the most reliable way to kill an execution system. The moment updating the plan is a separate task from doing the work, the plan goes stale and the review becomes archaeology.
What to drop
| Common practice | Why it fails |
|---|---|
| The monthly reporting pack | Costs a day to assemble, tells the reader what they already knew a fortnight ago |
| Cascading every goal to every level | Produces commitments nobody chose and everybody inherited |
| Scoring confidence out of ten | Confidence is not a measure, it is a mood, and it is reported upward optimistically |
| The annual offsite as the only review | One plan, one review, eleven months of drift |
| RAG status without a rule | Amber means whatever the owner needs it to mean this week |
The RAG one is worth naming specifically. Traffic light status is useful only when the thresholds are written down and applied by the number rather than the owner. Without that rule, everything is amber until it is suddenly red, which is the week before the quarter ends.
The sequencing most teams get wrong
Teams usually start with the tooling and end with the rhythm. It works far better in reverse.
Run a fortnightly fifteen minute review on a spreadsheet for six weeks first. You will discover which commitments nobody can report on, which owners did not know they were owners, and which measures are activity in disguise. Fix that, and the tooling decision becomes obvious because you know what you need it to do.
Start with the tool and you will configure your existing problems faster.
Where software earns its place
Once the rhythm exists, the thing that breaks it is collection. Someone has to chase fifteen owners every fortnight, and that person burns out or deprioritises it, and the system decays without a decision ever being made to stop.
That is the job Empiraa GPS is built for: owners and measures in one place, progress gathered rather than chased, and a stalled commitment visible while the response is still a choice. Advisory and consulting firms running this for clients use GPS for Professional Services, where each client gets their own workspace and the firm sees every engagement from one screen.
Related reading: why strategy execution fails, and what a strategy execution solution actually has to do.
Strategy execution has a definition on this site, and strategy execution software covers the category and who should not buy one.
