Validated learning is the use of feedback to inform and improve future actions. It is important to improve an organization's ability to undergo enterprise Agile transformation. This improvement is the primary responsibility of the transformation rollout team at any given level or area of the business. Those teams decide what changes should go on their transformation backlogs and what the priorities should be. They decide where good Agile patterns need to be applied in regard to each change, how and for what purpose, and what benefits may be expected from doing so.
Each change on the backlog is given acceptance criteria by the transformation rollout team. These criteria stipulate the conditions that must be satisfied in order for the change to be implemented. Different change items may have different acceptance criteria even when the same patterns are being applied. For example, if the limited WIP pattern is being coached to a particular delivery team, a suitable criterion might be that only three items can be in progress at a time. Other teams may be allocated different WIP limits. The more appropriate and better focused the acceptance criteria, the greater the chances are that the benefits of the pattern will be leveraged.
The transformation rollout team must, therefore, determine whether or not the acceptance criteria are suitable for the intended target. This evaluation should be done on an ongoing basis and at least once every transformation heartbeat for each change that is being rolled out. This determination is made on the basis of actionable metrics:
If the metrics indicate that the expected benefits have been achieved, then the change may be retired.
If the benefits have not been achieved, regardless of whether or not the acceptance criteria have been met, then further work remains to be done. The rollout team can replan the transformation backlog and the acceptance criteria for relevant changes may be improved.
Note that if the benefits have not been attained even though the acceptance criteria have been satisfied, then the change should still be retired. The lesson is that the acceptance criteria were inappropriate for the change target. A more appropriate change, with improved acceptance criteria and possibly implementing different patterns or practices, should be added to the backlog and prioritized.
Validated Rollout Strategy
When a transformation backlog is first created, none of the change items will yet have been applied. The transformation rollout team may not yet have had the opportunity to coach any good practices at all within the organization. No hard and direct experience of doing so may exist. This means that a transformation strategy may have to be initially formulated on the basis of:
Experiences gained elsewhere.
The transformation rollout team's understanding of the organization's present state.
The vision for change that the transformation sponsors have articulated.
The team should expect and plan to have valuable feedback no later than the completion of the first transformation heartbeat. Actionable metrics should have been elicited at that point, and the content of the transformation backlog may be revised on the strength of the lessons learned. This means that new changes can be added, their acceptance criteria can be shaped by recent experience, and different patterns may be exploited in different ways. Any entries that are no longer appropriate can be challenged and potentially removed. Other parts of the enterprise can perhaps be engaged in the change process, and the priorities given to each item in the transformation backlog can be reconsidered.
A validated transformational rollout strategy does not have to be a document. It can simply be the collected and shared experience of the rollout team throughout the Agile transformation. The important thing is that it must be evaluated in terms of actionable metrics at least once every transformation heartbeat, and that it informs the revision of the transformation backlog. The validation or revision of the transformational strategy provides closure to each and every cycle of the validated learning loop.