| name | understand-identify-execute-framework |
| description | A 3-stage product development process designed to increase shipping velocity and success rates by de-risking ideas before building. Use this when facing flat metrics after major launches, entering a new problem space, or planning high-velocity growth sprints to ensure work is grounded in first principles rather than justification bias. |
Understand, Identify, Execute (UIE)
The "Understand, Identify, Execute" framework is a method for de-risking product development by treating "understanding" as a first-class citizen in the roadmap. Rather than shipping a volume of guesses, teams allocate specific bandwidth to uncovering root causes and funnels before committing to full execution.
The Anti-Pattern: Identify, Justify, Execute
Many teams fall into the trap of deciding what they want to build first (Identify), then pulling data to support the decision (Justify), and sinking months into building it (Execute). This often results in "flat metrics" upon launch because the work wasn't grounded in a real user pain point or technical reality.
The UIE Process
1. Understand (First Principles)
Create an "intentional affordance" in your sprint for work that doesn't involve shipping code. This is a line item on the roadmap, not background work.
- Goal: De-risk the project and close knowledge gaps.
- Tactics:
- Instrumentation: If you don't have logging for an 8-step signup funnel, the "Understand" work is adding that logging.
- Funnel Analysis: Mapping exactly where users drop off.
- User Archetypes: Identify the "Adjacent User"—the person just outside your current user base who is struggling to use the product.
- Cross-functional Input: Ask Engineering, "What do we need to understand about the codebase to scale this?" and Data Science, "What proxy metrics are we missing?"
2. Identify (Targeted Opportunity)
Based on the "Understand" phase, select the specific lever that will drive the most impact.
- Compare what you know with high confidence (data-backed) against what you are interested in (hypothesis-based).
- Focus on low-effort, high-likelihood wins discovered during the research.
3. Execute (Parallel Pathing)
Run execution and understanding in parallel.
- The Portfolio Approach: A roadmap should not be 100% execution.
- Initial Mix: In a new problem space, aim for 60% Execution / 40% Understand work.
- Mature Mix: As confidence grows, shift to 85% Execution / 15% Understand work.
- Feedback Loop: Use the insights from this sprint’s "Understand" work to plan the next sprint’s "Execution" work.
Examples
Example 1: Instagram Signup Flow
- Context: The team wanted to improve the signup rate but had no visibility into the 8-step funnel.
- Understand: The team spent a sprint doing nothing but instrumenting every step of the funnel and growth marketing schemas.
- Identify: The data showed a massive drop-off between account creation and friend discovery.
- Execute: Launched a "connections pivot" that prioritized human-to-human connections over celebrity recommendations, doubling retention over 18 months.
Example 2: Instacart Reordering
- Context: Retention was lower than expected for long-term users.
- Understand: PMs dogfooded the app as "Adjacent Users" (busy parents) and looked at cohort data. They found that after 5 orders, 90% of a user's cart was the same as previous weeks.
- Identify: The product made reordering difficult (requiring 7+ clicks to find previous items).
- Execute: Shipped a streamlined "Buy it Again" feature and "1-Click Reorder" for staples like milk and eggs, significantly increasing repeat order frequency.
Common Pitfalls
- Building to Justify: Never pull data specifically to "win" an argument for a feature you’ve already decided to build.
- Skipping the Affordance: Assuming "Understand" work happens in the background. If it isn't a line item on the sprint board, it won't get done.
- Analysis Paralysis: Spending 100% of time on "Understand" work. You must always be shipping something to keep the feedback loop alive.
- Ignoring the Adjacent User: Designing only for "Power Users." Always test your flow as a brand-new user with no search history or pre-existing "graph" to see where the experience breaks.
