Lift failure
Lift failure is what happens when a team's attempts to get more users to do a specific behavior keep coming up empty across genuinely different interventions. A single failed attempt is normal; interventions miss for all kinds of reasons.
Definition
Accumulated lift failures across diverse interventions, where the share of signups performing the candidate behavior doesn't move — pointing to the team's intervention-design capacity as the failure rather than the proxy.
The diagnosis points at intervention design
Lift checks that an intervention designed to amplify a candidate behavior actually increased the share of the cohort performing it. When lift doesn't move across multiple diverse interventions, the proxy itself stays unindicted. What's failing is the team's ability to design interventions that effectively serve the underlying intent. The behavior the team is trying to amplify may still correlate with the commercial outcome; the team just hasn't found ways to get more users to do it.
The conclusion is operational rather than strategic. Keep the proxy; change how you're trying to move it.
Single failures don't qualify
A single lift miss is just a single-test failure. The charitable read on any one test is intervention-quality variance — the intervention may have been incomplete, miscalibrated, or applied to the wrong population. Lift failure is the pattern that emerges only after multiple diverse interventions have all failed to produce lift. Diversity matters here: if every intervention is a variation on the same underlying idea, multiple failures read as one design failing rather than as accumulated evidence about the team's capacity.
Distinct from hold failure
Lift failure and hold failure point at different things. Lift failure indicates the team can't reliably amplify the behavior; the proxy stays in the running. Hold failure indicates the behavior amplifies but the conversion correlation doesn't follow — the original correlation reflected which users did the behavior organically rather than a causal property driving the outcome — and triggers proxy failure as the strategic conclusion. Better intervention design can recover from lift failure; hold failure isn't a design problem.
Example
A team treats "completed first project" as a candidate behavior for their project-management tool. They run an onboarding redesign, then a guided-setup wizard, then a contextual prompt during the third session, then an email nudge sequence. Across all four, the share of signups completing a first project doesn't meaningfully change. The candidate behavior remains unindicted — users who complete a first project still convert at much higher rates than those who don't. What's indicted is the team's intervention-design approach. The interventions may all operate on surface layers of the user's experience while the actual barriers to first-project completion sit in starting state conditions the interventions don't address.