How we built the
Idea → Impact Diagnostic

Since November 3, 2026, the Idea → Impact Diagnostic on this site has opened Version 2 (beta). Reports made earlier used Version 1. This page explains Version 2 first, then Version 1.

Version 2 (beta)

Version 2 asks 25 statements across five stages, plus four optional reflections. It helps you see where to look first, or tells you when no single stage stands out. It does not claim to find your one true constraint.

The pipeline assumption (“Fix upstream first”), the “friction upstream compounds downstream” wording and the “primary bottleneck” wording in the Version 1 section below describe version 1. Version 2 does not claim that one stage causes another.

How version 2 differs from version 1

• A fifth stage, Defining the Work. It asks whether the work is thought through and agreed before building, apart from whether you learn from customers.
• Twenty-five scored statements, five per stage. Thirteen use the same words as version 1. Twelve are new.
• The page names one stage under “Where to look first” only when it scores at least 0.6 below the next lowest. When two stages below the Strength zone are too close to separate, it shows both. In any other close result it says that no single stage stands out, or that there is little friction in any stage.
• Each statement you agreed with gets one action. Statements you only came close to agreeing with are listed, but get no action.
• Your role and company size are saved and shown on your result, with a note about comparing across roles and sizes.
• Four optional written reflections. They are never scored.

The five stages

Insight Generation asks whether your teams learn from customers and use what they learn, before building and while they build. It covers checking problems with customers, using evidence over opinion, reaching customers, getting feedback to the people who set priorities, and testing with customers during the build.

Strategic Alignment asks whether priorities hold steady and whether people know who decides. It covers priority changes, decision rights, outside approvals, consistent signals from leaders, and whether people can link their work to this year’s goals.

Defining the Work asks whether the work is thought through and agreed before building starts. It covers understanding the problem, comparing options, sharing one view of scope, agreeing a measure of success, and finding big obstacles early.

Execution Excellence asks whether decided work reaches customers smoothly and predictably. It covers the time from decision to customers, waiting on other teams, how much work is in flight, rework, and the effort to release.

Learning & Adaptation asks whether, after you ship, your teams learn from results and act on them. It covers how fast results are reviewed, whether the original idea is checked, acting on results, recording lessons and sharing bad news.

Defining the Work is the newest idea. It rests on feedback from one leader, so we treat it as a hypothesis. We have not yet tested whether leaders see it as its own stage.

Every statement is read against one meaning of “teams”: the teams that turn ideas into what your customers get, in the part of the organization you lead. A CEO answers for the whole organization. A functional leader answers for their own area.

How scoring works

Each statement uses the same six-point scale: Strongly disagree, Disagree, Somewhat disagree, Somewhat agree, Agree, Strongly agree. There is no neutral midpoint. Agreeing with a statement means the friction is present.

A stage score is 7 minus your average answer, shown to one decimal. A 6.0 means you strongly disagreed with every statement in the stage. A 1.0 means you strongly agreed with every one.

A stage needs at least 4 of its 5 answers to be scored. With 4 answers, the score is rounded down to one decimal, toward more friction.

Strength is a score of 4.8 or above. Attention is 3.6 to 4.7. Action Required is below 3.6. In terms of your average answer, Strength is 2.2 or lower, Attention is above 2.2 up to 3.4, and Action Required is above 3.4.

These cut-offs are judgment, chosen so that Action Required starts where answers lean toward agreeing. They are not validated.

How the page decides where to look first

The page names one stage under “Where to look first” only when it scores at least 0.6 below the next lowest stage. If exactly two stages below the Strength zone are less than 0.6 apart at the bottom, it shows both under that heading and says it cannot pick between them. In any other close result it says that no single stage stands out, and it never points you to a stage in the Strength zone. If every stage is in the Strength zone, it says there is little friction in any stage.

One answer moving one step changes a five-answer stage score by 0.2, so 0.6 is three steps. We chose 0.6 because 0.4, which is two steps, named a stage too often in our simulation when no stage was truly lower. The margin is judgment, not a validated cut-off.

How often can a named stage be noise? We tested the rule on simulated answers in which all five stages were set to be equally strong and each answer varied at random by about one scale step. About 1 in 5 of those results still named a single stage, and between 1 in 10 and 3 in 10 as we changed the assumptions. When one stage was set a full step worse than the others, the page named it about half the time and named a different stage about 2 times in 100. Real answers may behave differently. Treat a named stage as a place to start a conversation, not as a finding.

Some statements may draw agreement from almost everyone. A stage that holds one of them will be named more often for reasons that have nothing to do with your teams. We will check for this as results come in.

If you gave the same answer to 21 or more of the 25 statements, the page adds a note saying so. Your result is still shown, and Randhir sees the flag.

Do not compare or average results across roles or company sizes without separating them. Four statements (ig3, sa1, sa4, la5) describe leadership behavior, which a CEO and a functional leader are likely to rate from different positions. That is an inference to test, not a finding.

What this can and cannot tell you

These are practitioner heuristics, not a psychometrically validated scale. The statements, the zone cut-offs, the 0.6 rule and the 21-of-25 flag are judgment. Reliability, factor structure and cut-off accuracy have not been tested. We make no claim of research validation, benchmarks or precision.

We have not yet tested the new statements, or the wording of the results page, in interviews with leaders. Wording may change, and each change is recorded in QUESTIONS.md with a new revision number.

Every score is one leader’s view of the teams. Check the result with two or three people who do the work.

Version 2 does not measure project-level decision ownership, capacity and resourcing, technical debt, the quality of your strategy, go-to-market readiness, business outcomes, formal versus informal reviews beyond recording lessons, or industry-specific release limits.

The books that shaped version 2 are listed on its welcome screen. They are not evidence that the questions work.

About 7 to 10 minutes for the 25 statements, plus a minute or two for each optional reflection you write, so 11 to 18 minutes if you write all four. These are estimates from reading and typing speeds, not measurements. We will replace them with the measured median once enough people have taken version 2.

Results are immediate. We save your answers, including any reflections you write, under a report ID so the link we email you works. Randhir sees each submission. Written reflections are not shown when you reopen a report from that link.

What you get

• Your five stage scores and a radar chart of the shape.
• A plain statement of where to look first, or of why your answers do not point to one place, with the zone of each stage named.
• One action for every statement you agreed with, and a list of the ones you came closest to agreeing with.
• A PDF you can share with your team or use in a planning session.

Version 2 is a free beta, and anyone can take it. Results are immediate.

Why friction, not maturity

Most organizational diagnostics are maturity models. They ask: “How capable are you?” Then they score you against a benchmark and tell you where you sit on a spectrum from Level 1 to Level 5.

The risk is social desirability bias. When people answer a capability question, they may answer it aspirationally, describing how the team operates on a good day or how they wish it operated. That can inflate scores and make the picture look better than reality.

Friction framing tries to avoid that. Instead of asking “how good are you?”, it asks “where does it hurt?” Friction statements describe specific, recognizable pain points, for example: Unclear priorities slow us down when new opportunities arise. We often discover gaps between what stakeholders expect and what we actually deliver.

Leaders can identify with those statements honestly. Agreement is not an admission of failure. It is a practical description of a real problem. We chose this framing to make honest answers easier. We have not tested whether it does.

Version 1: reports made before November 3, 2026

This section explains the earlier version of the diagnostic, with four dimensions and 20 scored items. It applies to every report made before November 3, 2026, and to any later report that shows “Your Primary Constraint”. A version 2 report is marked “Version 2 (beta)” and shows five stage scores. Version 1 reports still open from their original links.

The four dimensions in version 1

Version 1 maps your team across four dimensions that it treated as a pipeline. It assumed that friction upstream compounds downstream: that a team that struggles to define the problem will misalign on strategy, ship the wrong things, and have little worth learning from.

How scoring works in version 1

Each of the 20 scored items uses a 6-point scale with no neutral midpoint, which removes the easy escape of “neither agree nor disagree”. (Four short open-ended reflections bring the full assessment to 24 questions.)

All 20 scored items are friction statements. Agreement with a statement means the friction is present. Responses are reverse-scored so that higher scores represent lower friction. A dimension score of 6.0 means the friction is essentially absent. A score of 1.0 means it is pervasive.

A version 1 report names the lowest-scoring dimension as “Your Primary Constraint”, with a short paragraph about it. If two dimensions tie, it names the earlier one in this order: Insight Generation, Strategic Alignment, Execution Excellence, Learning & Adaptation. It names a dimension even when that dimension is in the Strength zone. Version 1 treated the lowest-scoring dimension as the best place to start. That is a design choice we have not tested.

Sources behind version 1

Version 1 drew on published work across product development, organizational behavior, and execution science. Each construct drew on at least one published body of work.

These are practitioner heuristics, not a psychometrically validated scale. The zone cut-offs and the rule that picks your result paragraph are judgment calls. No scored item measures whether teams worked out their approach, planned how to launch and learn, or reached detailed agreement on what to build.

• Reinertsen (2009) — The Principles of Product Development Flow
  Foundational framing for queue management, batch size, and flow efficiency as the primary levers of product development throughput. The pipeline metaphor and bottleneck logic derive directly from this work.
• Forsgren, Humble & Kim (2018) — Accelerate (DORA Research)
  Four-year research program identifying the metrics and practices that distinguish high-performing software delivery organizations. Execution Excellence items draw on the delivery practices DORA's research associates with high performance.
• Torres (2021) — Continuous Discovery Habits
  Practitioner framework for structured, ongoing customer discovery. The Insight Generation dimension draws on Torres' distinction between assumption-testing and continuous opportunity exploration.
• Cagan (2018, 2020) — Inspired / Empowered
  Widely adopted frameworks for product team structure, outcomes-vs-output thinking, and the conditions that enable teams to do their best work. Strategic Alignment items reflect Cagan's emphasis on shared context over directives.
• Edmondson (1999, 2019) — Psychological Safety / The Fearless Organization
  Peer-reviewed research establishing psychological safety as the most consistent predictor of team learning and performance. Learning & Adaptation items include constructs from Edmondson's team learning framework.
• Kaplan & Norton (2001) — The Strategy-Focused Organization
  Research on how organizations translate strategy into execution. The Strategic Alignment dimension incorporates Kaplan & Norton's finding that most strategy failures are execution failures, not strategy failures.
• Pfeffer & Sutton (2000) — The Knowing-Doing Gap
  Analysis of why organizations fail to translate knowledge into action. The diagnostic's friction framing addresses the knowing-doing gap directly — by surfacing the specific points where knowledge stops converting into output.

What a version 1 report shows

We save your answers so your emailed report link works, and Randhir sees each submission. A version 1 report shows the following items.

• A “Your Primary Constraint” card: the lowest-scoring dimension, a short paragraph about it, and its zone and score.
• A radar chart and a table of the four dimension scores, each with its zone.
• A list of strengths: the statements you answered Disagree or Strongly disagree.
• A list of friction areas, in dimension order: every statement you answered Somewhat disagree or higher, each with a short note and one “Try this” action.
• A line telling you to address friction in order, because upstream constraints compound downstream. Version 2 does not make that claim.
• Every statement with a score from 1 to 6, where higher means less friction, and your written reflections.
• A PDF you can share with your team or use in a planning session.