How to Plan Each Observation Around a Defined Target and Sample
Good observation starts before anyone watches the work. This playbook helps you define the behavior, the conditions where it should appear, the time window, and the sample so the record reflects performance rather than whatever happened to catch your attention.
Developing
Start here. Build the foundation.- 1
Before the observation, write one target behavior as a visible action. Replace broad labels like 'communication' with the concrete response you expect to see, such as how the person handles a pricing objection. The target is ready when a colleague could read it and watch for the same thing.
- 2
Add the condition that should call for the behavior. Write it as a when-then statement, such as 'when the customer raises a competitor, the rep asks what they value about that option.' This lets you tell the difference between a missed opportunity and a situation where the behavior was not needed.
Proficient
Build consistency and rhythm.- 3
Choose the observation window and the unit you will record before the session starts. Decide whether you will count each instance, use a time interval, or make one yes-or-no judgment for the window. When the work begins, your job is to record, not to redesign the method.
- 4
Plan more than one look when the skill matters. Schedule observations across different times, contexts, or task types so one memorable session cannot carry the whole rating. The sample is stronger when no single event can swing the picture on its own.
Mastered
Operate at the highest level.- 5
Turn the target, condition, window, recording unit, and sample into a one-page plan. Give it to another observer without a verbal walkthrough. If they capture the behavior you intended, the plan is reusable.
Common Pitfalls
Avoid the common failure modes.- Walking in with a topic like 'communication' instead of a target that can be watched. A broad topic gives every observer room to see something different.
- Deciding what counts while the action is already happening. That shifts the grain mid-session and makes the record harder to trust.
- Building the whole rating from the session that was easiest to schedule. Convenience is not the same as a representative sample.