Skip to main content

Documentation Index

Fetch the complete documentation index at: https://wiki.jamesvarga.com/llms.txt

Use this file to discover all available pages before exploring further.

​
What it is

The Founder’s 2×2 is a decision-making framework that maps two independent axes:
  • Generation axis — how a mind produces ideas. Linear / convergent (sequential, logical, narrowing) vs lateral / divergent (associative, reframing, expanding).
  • Response axis — how a person treats someone else’s idea. “Yes-and” (accept and build) vs “yes-but” (qualify and challenge).
Crossing the two yields four operating modes. High-performing teams deliberately move between them by phase rather than living in one cell. The two axes are independent — you can be laterally generative while blocking everyone (“yes-but, here’s my better wild idea”) or linearly methodical while accepting and building (“yes-and so the next logical step is…”). The framework is a synthesis of validated components — Guilford’s divergent / convergent thinking, Keith Johnstone’s improv work on accepting and blocking, Amy Edmondson’s psychological safety, Charlan Nemeth’s dissent research, and the Double Diamond — not a single citable original model. It is offered as a working vocabulary for founders and operating teams to make mode-switching explicit and deliberate.

​
Who it is for

Founders, operators, and team leads who need to switch between exploration and execution within the same week — sometimes the same meeting. Particularly useful in early-stage commercial work where a missed mode-switch produces one of two recognisable failure modes:
  • Marching confidently off a cliff — flawless linear execution of a wrong premise. The team is internally agreeable and operationally disciplined but never red-teams its core hypothesis.
  • Serial pivoting — endless lateral reframing without commitment. The team confuses pivoting with whatever feels different this week.

​
Prerequisites

  • A team large enough to have meetings worth structuring (two or more).
  • A willingness to declare what mode a session is in, out loud.
  • Psychological safety as Edmondson defines it — not niceness, but a shared belief that interpersonal risk is safe. Without it, “yes-but” becomes “shut up” and “yes-and” becomes “please nobody disagree with me.”

​
The four modes

Yes-and (accept and build)Yes-but (qualify and challenge)
Lateral (divergent, expanding)Ideation / Explore — early discovery, MVP hypotheses, “what if we tried…” Defer judgment, build on every offer. Goal: maximum option set before any constraint applies.Provocation / Red-team — pre-mortems, “what would have to be true?” Challenge the framing itself, not the polish. Goal: surface the unstated assumption before it ships.
Linear (convergent, narrowing)Execution / Build — commit and ship the agreed plan. Disagree before, commit during. Goal: convert decision into output.Audit / Stress-test — due diligence, convergent selection, kill criteria, compliance review. Goal: pass-or-fail evaluation against documented criteria.

​
Steps

  1. Declare the mode out loud. Every working session has a stated mode. Anyone can name which cell the room is currently in. The mode is a choice, not a personality.
  2. Time-separate generation from evaluation. Never diverge and converge in the same conversation. Run silent or written idea generation first — this defeats production blocking, which research consistently identifies as the dominant reason real brainstorming groups underperform pooled individuals. Then a separate critical pass.
  3. In the critical pass, permit debate. Nemeth’s research, replicated in the US and France, found that explicit permission to debate and criticise produced more and better ideas than the classic “defer judgment” instruction. Update Osborn’s rules: defer judgment during generation, then explicitly licence challenge during evaluation.
  4. Engineer authentic dissent, not theatre. Use pre-mortems before major commitments. Find someone who actually disagrees rather than assigning a devil’s advocate — role-played dissent tends to bolster the original view, while real disagreement expands the search.
  5. Match the mode to the phase. In search and pre-PMF, weight toward Lateral + Yes-and, but gate it with scheduled pivot-or-persevere reviews (Linear + Yes-but). Post-PMF and in execution, weight toward Linear + Yes-and, but protect periodic red-team sessions (Lateral + Yes-but) so the team doesn’t march off a cliff.
  6. Call the switch. The founder’s job is to name the transition. “We’re switching modes — we have three options, now we’re going to test them against the kill criteria.” The transition itself is a leadership act.

​
What good looks like

  • Every meeting has a declared mode at the start.
  • Generation and evaluation never happen in the same conversation without an explicit switch.
  • Disagreement happens before decisions and stops after — Bezos’s “disagree and commit,” Andy Grove’s “constructive confrontation.”
  • Pivots happen once or twice in a startup’s life. The Startup Genome data (650+ startups) found that companies that pivot once or twice raise 2.5× more money, grow 3.6× faster, and are 52% less likely to scale prematurely than startups that pivot zero or more than twice — direct evidence that moderate mode-switching beats both extremes.

​
Thresholds that should change the approach

  • If pooled individual ideas consistently beat your group sessions → you have a production-blocking or evaluation-apprehension problem. Switch to brainwriting / nominal-group methods.
  • If no one ever disagrees in meetings → psychological-safety failure. Edmondson’s grid: high safety + low standards = Comfort Zone; low safety + high standards = Anxiety Zone. Only high + high = Learning Zone. The problem is candour, not ideas.
  • If you’re pivoting more than twice without new validated learning → serial-pivot mode (Startup Genome’s danger zone). Impose Steve Blank’s 72-hour rule and a convergent decision gate before any further pivot.
  • If execution is flawless but a core assumption has gone untested for a quarter or more → schedule a mandatory pre-mortem or red-team session before further spend.

​
Common failure modes

  • Treating “yes-and” as a universal rule rather than a phase-specific discipline. Even improv practitioners warn against this — the Upright Citizens Brigade tradition switches from “Yes And” to “If, Then” once a scene’s premise is established.
  • Mistaking psychological safety for niceness. A team that “never argues” is exhibiting a red flag, not health. Edmondson has spent a decade correcting this misreading.
  • Assigning a devil’s advocate instead of finding real disagreement. Role-played dissent tends to bolster the original view rather than challenge it.
  • Confusing pivoting with mode-switching. A pivot is a substantive change to a business-model component, not a change of mood.
  • Sticking in execution mode while the core thesis has gone untested. Marc Andreessen’s “buttoned-down team heading straight off a cliff.”

​
Example: applying it to a 5 Days to Scale sprint

A 5 Days to Scale sprint is largely a Linear + Yes-and exercise — commit to one target, one offer, one ask, and execute across five days. The framework prescribes how to enter and exit the sprint:
  • Before Day 1: a Lateral + Yes-but session that challenges the sprint’s premise. Is the selected segment actually winnable in five days? What would have to be true? If the answer is “the segment is wrong,” restart the selection rather than running a wasted sprint.
  • During Days 1-4: the sprint operates in Linear + Yes-and. Disagreement happens before, not during — the team commits to the plan and ships.
  • At Day 5: an explicit Linear + Yes-but pass against the success criteria. Advance the deal or formally qualify it out. Disqualification is a valid result.
This is the pattern across most frameworks on this wiki — each one assumes a particular mode and the value is in knowing when to enter and exit it.
Last modified on June 1, 2026