Outcome-First Behavioral Intervention
Fix the outcome, diagnose pressures, and test interventions around it
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 97%
This method begins with a measurable behavioral outcome rather than a favored solution. The team observes what people actually do, identifies forces that promote the behavior and pressures that inhibit it, and then develops interventions without privileging engineering, marketing, or operations. Each intervention is treated as an experiment orbiting a fixed outcome. For example, a long checkout may be an engineering problem, but it may instead be a perceived-wait problem that information or distraction can solve more cheaply. Results determine the next move: retain an effective intervention, modify an uncertain one, or test a different mechanism. This produces practical behavioral change while reducing the risk of optimizing the wrong variable.
Origin
Extracted from Masters in Business through Matt Wallaert's discussion of behavioral scientists as outcome-focused interventionists.
Core principles
- 01Treat the desired outcome as the fixed point
- 02Stay solution-agnostic and problem-focused
- 03Observe actual behavior before proposing solutions
- 04Distinguish motivating pressures from inhibiting pressures
- 05Test interventions instead of defending one discipline's preferred tool
How to run it
- 1
Fix the outcome
Define the specific observable behavior or experience that should change. Keep that result fixed while remaining flexible about how to reach it.
Pro tip Express the outcome as something that can be measured before and after an intervention.
Watch out Do not define the outcome as implementing a preferred feature.
- 2
Observe existing behavior
Watch what people do in the real environment rather than relying only on stated preferences or departmental assumptions.
Pro tip Look for moments where behavior diverges from what users say they want.
Watch out A plausible story about behavior is not evidence.
- 3
Map behavioral pressures
List the forces promoting the desired behavior and the inhibiting pressures making it difficult, risky, ambiguous, or unpleasant.
Pro tip Test whether motivation already exists before trying to increase it.
Watch out Teams frequently mistake an inhibiting-pressure problem for a lack of desire.
- 4
Generate agnostic interventions
Consider changes to information, environment, process, incentives, tools, and experience. Select ideas based on their mechanism, not departmental ownership.
Pro tip Include low-cost perceptual interventions alongside structural changes.
Watch out Do not let the available tool dictate the diagnosis.
- 5
Run a bounded test
Introduce one intervention at a scale that permits useful measurement without unnecessary exposure. Compare the resulting behavior with the baseline.
Pro tip Use the smallest test capable of disproving the intervention.
Watch out Avoid changing several mechanisms at once if you need to identify causality.
- 6
Pivot around the result
Preserve the outcome while iterating on the intervention. Continue testing until the mechanism and behavioral effect are sufficiently clear.
Pro tip Record failed tests as evidence about the problem's boundaries.
Watch out Do not protect an intervention merely because a team invested in it.
In the wild
Customers report that a five-minute checkout feels excessively long. Rather than immediately rebuilding the point-of-sale system or adding counters, the store displays an accurate estimate of the remaining wait and tests whether reduced ambiguity improves satisfaction.
→ The store can improve perceived waiting and customer satisfaction without necessarily reducing the physical wait.
Microsoft observed unexpectedly low school search traffic. Classroom visits showed that children were already curious; teachers were inhibited by advertising, tracking, privacy, and adult-content concerns. Bing removed those barriers with an ad-free, untracked school version and locked safe search.
→ School searches increased after inhibiting pressures were removed.
Common mistakes
Treating every problem as a motivation gap
People may already want the outcome but face practical, emotional, or structural barriers. More encouragement will not remove those barriers.
Using the department's default tool
Engineers, marketers, and operators naturally propose solutions from their own disciplines. That can prematurely narrow the intervention space.
Optimizing the wrong variable
A team may shorten an objective wait when dissatisfaction is actually driven by ambiguity or how the wait feels.
Is it for you?
Best for
Teams changing customer, employee, or user behavior through products, policies, or experiences.
Not ideal for
Situations where the desired outcome cannot be observed or ethically influenced.
From the transcript
“the fixed point is the outcome, and then you pivot around experiments that help you test around that outcome.”
“There's nobody who's solution agnostic and his problem focused, right”
“one of the tenants of behavior science, go look at how people are behaving.”
From the episode
Matt Wallaert Is on a 'Chief Behavioral Officer' Mission
Matt Wallaert