AI Skill Report Card

Pivoting and Reframing Ideas

A-86·Oct 2, 2026·Source: Extension-page
14 / 15

When an idea is stuck, don't brainstorm more of the same — change the lens. Pick one technique below and run it in 10-15 minutes:

Problem: "Our habit-tracking app has stalled at 2% conversion."

Quick reframe (Reframing Matrix - 4 lenses):
- Product: What if tracking happened passively (sensors) instead of manual logging?
- Planning: What if we sold to employers instead of consumers?
- Potential: What if the data became a certifiable wellness credential?
- People: What if coaches used it, not end-users directly?

Pick the lens with the most energy and prototype that framing first.
Recommendation▾
Add a 'bad output' example contrasting a shallow/ineffective reframe vs. a good one to reinforce the quality bar
14 / 15

Progress:

  • Step 1: Name the exact point of being "stuck" (no ideas vs. idea exists but isn't working vs. idea works but market rejects it)
  • Step 2: Choose the matching technique (see decision table)
  • Step 3: Generate 5-8 raw variants using that technique — quantity over quality
  • Step 4: Stress-test top 2-3 against real constraints (budget, users, data)
  • Step 5: Commit to one direction and define the smallest validation test

Step 1 — Diagnose the stuck point. This determines which tool to reach for:

SituationUse this
No ideas at all, blank pageAI generator / Random Idea prompt, SCAMPER
Idea exists but feels flat or obviousReframing Matrix, Metaphorical Thinking
Idea exists but can't find the angleReverse Brainstorming (invert the goal)
Idea is validated but market/traction is wrongFormal Pivot (Ries's 10 types)
Too many directions, need to narrowOpposite Thinking Canvas + scoring

Step 2 — Apply the technique.

Reframing Matrix — Split the problem into 4 lenses and brainstorm each independently:

  • Product: features, specs, mechanics
  • Planning: process, timing, sequencing, logistics
  • Potential: future possibility, scale, adjacent markets
  • People: who's involved, who benefits, who decides

Reverse Brainstorming — Ask "How could we guarantee this fails / makes things worse?" List those failure modes, then invert each into a solution. Surfaces blind spots normal brainstorming misses.

SCAMPER — Run the idea through each filter: Substitute, Combine, Adapt, Modify, Put to another use, Eliminate, Reverse.

Metaphorical Thinking — Force an analogy: "If this product were a restaurant/sport/relationship, what would it look like?" Extract the structural insight, not the literal metaphor.

Formal Pivot (for validated-but-failing ideas) — Identify which of Ries's pivot types applies:

  • Zoom-in (one feature becomes the whole product)
  • Zoom-out (product becomes one feature of something bigger)
  • Customer segment (same product, different buyer)
  • Customer need (same buyer, different problem solved)
  • Platform pivot (app → platform or vice versa)
  • Business architecture (high margin/low volume ↔ low margin/high volume)
  • Value capture (change monetization mechanism)
  • Engine of growth (viral ↔ paid ↔ sticky)
  • Channel pivot (change distribution/sales channel)
  • Technology pivot (same solution, different tech)

Step 3 — Generate volume before judging. Aim for 5-8 variants minimum per technique before evaluating any of them. Premature filtering kills the technique's value.

Step 4 — Stress-test survivors against: Can we build/test this in 2 weeks? Does it require a customer we don't have access to? Does it contradict known constraints (budget, regulation, tech)?

Step 5 — Commit and define a falsifiable test. A reframe isn't done until there's a concrete next action: a prototype, a 10-customer interview script, or a landing page test.

Recommendation▾
The decision table could include a brief note on how to handle ambiguous cases where multiple situations apply simultaneously
17 / 20

Example 1 — Reverse Brainstorming Input: "Our onboarding has a 60% drop-off and we don't know why." Output:

  • Inverted question: "How would we design onboarding to maximize drop-off?"
  • Failure list: too many steps, ask for payment first, no visible progress, generic copy, no quick win
  • Flipped solutions: cut to 3 steps, delay payment ask, add progress bar, personalize first message, surface a "quick win" action in under 60 seconds

Example 2 — Formal Pivot Diagnosis Input: "We built a scheduling tool for freelance tutors, but tutors aren't paying — though they love referring parents." Output: This matches a Customer Segment pivot — same core product, same problem (scheduling + trust), but the paying customer should be parents/schools, not tutors. Next test: rebuild pricing page targeting parents, run 20 cold outreach messages to school admins.

Example 3 — Reframing Matrix Input: "Our B2B analytics dashboard isn't getting daily active use." Output:

  • Product: turn it into a Slack bot that pushes insights instead of a dashboard users must visit
  • Planning: send a weekly digest instead of real-time data
  • Potential: white-label it for consultants to resell to their clients
  • People: target the exec who wants a summary, not the analyst who wants raw data → Selected: Product reframe (push-based Slack bot), validated with 5 existing customers in one week.
Recommendation▾
Consider trimming the technique descriptions slightly (SCAMPER, Metaphorical Thinking) since Claude likely knows these frameworks already
  • Diagnose before generating. "No ideas" and "wrong idea" need different tools — don't brainstorm when you should pivot, and don't pivot when you just need more raw ideas.
  • Separate generation from judgment. Produce all variants first; evaluating mid-stream kills volume and biases toward the obvious.
  • Ground pivots in signal, not mood. A pivot should respond to specific customer feedback or data, not just fatigue with the current idea.
  • Reframing is a small shift, not a reset. If the output looks like an entirely unrelated idea, that's a new concept, not a reframe — label it honestly.
  • Always end with a test. A reframed concept without a next validation step is just a new opinion.
  • Use AI generators for volume, humans for judgment. AI tools are excellent at producing raw variants fast; the stress-test and commit steps need domain judgment.
  • Treating every stuck moment as a "pivot" when it's actually a lack of initial ideas — wastes the heavier pivot framework on a simple brainstorming gap.
  • Running a reframing technique once and taking the first output instead of generating a full set (5-8 variants) before picking.
  • Pivoting based on internal boredom with an idea rather than external evidence (customer behavior, market data).
  • Confusing "reverse brainstorming" with just thinking negatively — the technique requires the explicit invert-back-to-solution step, not just listing problems.
  • Skipping the stress-test step and committing to a reframed idea that ignores known hard constraints (budget, timeline, regulation).
  • Using a drastic pivot type (e.g., full business architecture change) when a lighter reframe (e.g., messaging or audience shift) would have solved the actual blocker.
0
Grade A-AI Skill Framework
Scorecard
Criteria Breakdown
Quick Start
14/15
Workflow
14/15
Examples
17/20
Completeness
18/20
Format
15/15
Conciseness
13/15