A strategy is not translated until people can tell what they must decide differently on Monday.
Why it matters when the plan changes
Launches are designed as communication events and measured on reach and comprehension, figures that describe whether a message arrived, not whether it was translated. Both can be excellent while nothing changes, because knowing the strategy and knowing what to decide differently are separate things. The second requires each person to work out the implication for their own decisions, which no cascade produces on its own. Sull, Homkes and Sull found only eleven per cent of managers believed priorities carried the resources needed for success, a gap a well-received launch leaves untouched.
The tension is that the translation is local and the message is central; a single articulation cannot tell a regional finance manager and a product lead what each should stop doing. Leaving it to them produces several private versions of the strategy, each defended in good faith, so a unanimously supported plan produces divergent behaviour. The same research found nine per cent of managers could rely on colleagues in other functions and units all the time.
In practice
A strategy is launched with a town hall, a cascade pack and a questionnaire showing ninety per cent comprehension. Six months later the initiative count has grown and three units are running incompatible interpretations. Everyone can state the strategy correctly, and nobody was told what to stop.
Evidence
Most managers do not believe stated priorities are properly resourced, which is what an unlanded translation produces.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)Reliance across units is far weaker than within a line, so a message travelling down each line lands differently in each.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
What it cannot tell you
A launch tells you a strategy has been communicated, not that it has been translated. It cannot show whether a regional finance manager or a product lead now knows what to stop doing. High comprehension scores can coexist with several private, sincerely held versions of the same strategy, each producing different decisions on Monday.
Questions
Awareness, which is necessary and not sufficient. People learn what the strategy is. Whether they know what to decide differently depends on a separate step that has to happen in each unit, and that step is usually assumed rather than designed.
Ask people separately what they will stop doing and who now decides a named contested call. Agreement means it has landed. Divergence means the organisation is carrying several private versions, each held sincerely, which is the normal state six months in.
Because each leader translates for their own unit from their own understanding, and the translations are never compared. Everyone is doing the work faithfully and the results differ, which is a structural property of cascading rather than a failure of anyone in the chain.
Translation sessions in which each unit works out what it stops, starts and decides differently, and those outputs compared across units. That comparison is the step that catches divergence, and it is almost never scheduled because the launch was treated as the end.
Within a quarter, while the interpretations are still forming and cheap to correct. After two quarters each unit has acted on its own reading and committed resources to it, and correcting it then means undoing completed work rather than clarifying a question.