Professional Practice & Everyday Jargon

Say-do gap

The discrepancy between what an organisation publicly says it will do or values and what its decisions, resources and observable behaviour actually demonstrate.

Emerging · Version master-draft-2026-08-10

Expert review openNo editor-accepted expert review yet

Definition

The discrepancy between what an organisation publicly says it will do or values and what its decisions, resources and observable behaviour actually demonstrate.

Overview

Development status

For expert editorial review.

“The say-do gap is not measured in words; it is measured in the decisions the words fail to change. ”

The say-do gap is fundamentally a credibility problem: stakeholders observe inconsistency between narrative and behaviour, even where individual statements may be technically defensible. A company may publicly champion supplier living wages while procurement rewards lowest price and short contracts. The gap is not solved by adding another policy because the purchasing behaviour contradicting the claim remains intact.

This is why say-do gap should be treated as a decision concept rather than a decorative label. A definition earns its place in practice only when it helps someone distinguish a stronger course of action from a weaker one.

The term is practitioner shorthand used across responsible business, strategy and behavioural research. It differs from an implementation gap. Implementation can fail despite sincere intent and internal effort; a say-do gap emphasises inconsistency between public representation and actual behaviour, especially where the organisation continues to communicate confidence.

That distinction is important because sustainability language often migrates between regulation, management, investment and communications, where the same word can imply different duties. Responsible use begins by naming the purpose and boundary rather than assuming a shared meaning.

Collective sustainability language often sounds cooperative even when incentives remain fragmented. A useful test is to follow resources, decision rights and accountability: who can commit whom, who pays, who carries risk, and what happens when interests diverge. Collaboration becomes substantive only when those questions have operational answers.

Commitment registers, decision audits and claim substantiation can make the gap visible. Governance should compare major public positions with capital allocation, procurement, remuneration, lobbying and operational performance. This shifts attention from the visible artefact - a title, workshop, pledge, platform, score, report or process - to the governance and evidence beneath it.

A practical way to interrogate the concept is to ask what would be observable if it were working well. Trust erodes faster from visible inconsistency than from transparent difficulty. Naming the gap creates permission to align communications with evidence and to redirect attention toward the organisational causes of non-delivery.

Useful indicators should therefore include not only completion or participation, but the decisions, behaviours, outcomes or reductions in uncertainty that the practice is expected to produce.

Organisations often manage the gap as communications risk - refining language, adding caveats or producing more disclosure - instead of investigating the incentives, decision rights and resource choices producing the contradiction. This is rarely solved by adding another layer of terminology.

The corrective is usually more concrete: clearer ownership, better evidence, fewer contradictory incentives, stronger stakeholder participation, or a more honest statement of what the organisation can currently support.

Evidence should be proportionate to the claim. Where the concept describes a formal process, practitioners should retain criteria, decisions, source information and changes over time.

Where it is practitioner jargon, the need for discipline is greater rather than smaller: the organisation should explain what it means, avoid implying a universal definition and choose language that a reasonable reader can test against observable facts.

Context also matters. A multinational, a small supplier, a public authority and a civil-society organisation may face the same sustainability issue with radically different power, resources and obligations. Good practice does not use context to excuse severe impacts, but it does use context to design proportionate implementation, support and evidence.

This is particularly important where requirements travel down supply chains from actors with more influence to those with less.

The concept becomes most useful when it changes a question. Instead of asking whether the organisation can say it has say-do gap, ask what the term requires us to see, decide or do differently. That shift from label to consequence is the recurring discipline of this book: clearer definitions should create better decisions, not simply more sophisticated language.

Practical Application

Choose a sample of major sustainability claims and trace each backwards to the decisions, budgets, metrics and behaviours that would have to be true for the claim to remain credible. Where contradiction exists, decide whether action must change or language must narrow. Do not preserve an aspirational claim by repeatedly explaining why implementation is someone else’s responsibility.

Build the result into normal management rather than leaving it as an annual sustainability exercise. Assign an owner, a review point and a small number of evidence tests that would reveal whether the practice is improving. When conditions change, update the decision openly rather than preserving an obsolete classification or claim for the sake of consistency.

Why It Matters

Trust erodes faster from visible inconsistency than from transparent difficulty. Naming the gap creates permission to align communications with evidence and to redirect attention toward the organisational causes of non-delivery. The broader value is organisational clarity: people can see what the concept is for, what evidence belongs to it and where responsibility sits.

That makes it easier to challenge weak practice without turning every disagreement into a debate over vocabulary.

Common Misconception

Every missed target is not automatically a say-do gap. The term is most useful when representation and behaviour diverge, not simply when difficult implementation underperforms. A more useful test is substantive rather than semantic: what would have to be true in the real world for the term to be justified, and what evidence would make us withdraw or narrow the claim?

Connections

Implementation Gap explores execution failure more broadly. Pledge and Commitment describe the statements against which behaviour is judged. Impact Washing and Virtue Signalling examine more deliberate or perceived forms of symbolic communication.

These connections matter because no sustainability term operates alone; each creates boundaries that determine which evidence and responsibilities are carried forward into the next decision.

A Question Worth Asking

Which public sustainability statement would be hardest to defend if a reviewer looked only at budgets, incentives and operating decisions?

Selected References

• OECD. 2026. OECD Responsible Business Outlook 2026.

• United Nations High-Level Expert Group. 2022. Integrity Matters: Net-Zero Commitments by Businesses, Financial Institutions, Cities and Regions.

• Science Based Targets initiative. 2026. Corporate Net-Zero Standard, Version 1. 3. 1 and published Version 2. 0 transition materials.

• ISO. 2010. ISO 26000: Guidance on Social Responsibility.

Core chapter length: 973 words.

Review

Public comments appear only after editor acceptance. Draft comments stay in the review queue.

0
How people contribute

Reviewers choose the definition or an overview paragraph, leave a comment or replacement, and attach evidence or a source link.

How comments are used

Editors compare reviewer cards side by side. AI may help find agreement, conflicts, unsupported claims and possible source issues.

What gets published

Only an editor-accepted synthesis changes the public page. Reviewer identities are shown only with consent and verification.

No verified experts yet

Submitted reviews stay private until accepted.

Loading verified endorsements… Endorsements are not votes and never determine publication.

Endorse this definition

Endorse the exact version shown here. This is not a vote, and publication remains an editorial decision.

vmaster-draft-2026-08-10

Sign-in supplies your email for verification and necessary follow-up; it is not displayed publicly. We do not ask you to enter it again.

Sign in with a passwordless email link before submitting.

Review board

Comment on a specific line. Each reviewer stays separate until an editor accepts a merged draft.

1Separate reviewer cards

Each person comments on the definition or overview in their own draft card, with role, evidence and suggested wording kept together.

2AI comparison

AI can compare comments against the current text, flag conflicting claims, surface missing evidence and identify where reviewers agree.

3Editor synthesis

An editor merges compatible suggestions into a draft change, checks sources, records disagreements and decides what can be published.

Text to reviewChoose the exact definition or overview paragraph.
Reviewer commentDraft only. Not public until editor accepted.
Definition
Reviewer identityYour signed-in account identifies the submission. We use its email only for verification and necessary follow-up, and never display it publicly.
Before you submit

This proposal follows the editorial and AI-assistance rules. The live definition will not change until an editor accepts it.

  • Add the proposed wording or note.
  • Explain why the change is needed.
  • Ready
  • Ready

Sign in with a passwordless email link before submitting.

You can still save a draft, but completing these items makes editorial review faster. Multiple reviewers can suggest changes on the same text. Editors compare, merge, accept or decline them before any public change.