Trueness brand platform

A written proposal for Trueness, using the structure of the reference brand-page template and brand-pyramid copy specification: purpose, essence, promise, attributes, benefits, reasons to believe, positioning, audience, products, identity direction and voice. The known name is Trueness and the domain is trueness.ai. The rest is proposed. No logo, palette, typography, imagery, screen or motion design has been created.

25 October 2026. Joint Claude and Codex planning edition. The underlying rule inherited from the reference brand template is useful here: describe the service specifically and make only the claims its evidence can support. Trueness is a proposed research business with AI-assisted delivery, paid human expertise and research suppliers; it is not an accredited laboratory or standards body.

The name

Trueness has a useful connection to measurement. The International Vocabulary of Metrology distinguishes agreement with a reference value from the repeatability of measurements. This gives the brand a meaningful idea: a system can repeat itself consistently while missing what it is supposed to represent. See the VIM definition of trueness and precision.

Use that connection as an explanation, not a claim of metrological accreditation or access to an objective truth about people. Human surveys also contain sampling and measurement error. VIM itself does not define a universal numerical “trueness score,” which reinforces the decision to report several relevant metrics.

Proposed naming convention: Trueness in prose, trueness.ai for the domain. Keep the familiar word intact. Trademark clearance and domain ownership are launch checks; the planning exercise has not established them.

Purpose

Help people use synthetic audiences with a clear understanding of what the human evidence supports.

Essence

Faithful to the people it stands for.

Promise

We check your synthetic audience against real people and show you where it can help.

Attributes

Exact. Candid. Human-minded. Practical.

Exact means a claim carries its scope and evidence. Candid means an inconclusive result is allowed to remain inconclusive. Human-minded means variation and disagreement matter. Practical means the customer can see what decision to make next.

Illustrative voice, not a real research finding:

The audience missed a concern that appeared in the human interviews. Use it to explore wording. Test the final choice with people.

Emotional benefits

• A defensible next step. The buyer can explain why the tool was useful and where they stopped relying on it.
• Less guesswork. A visible comparison replaces an unexamined impression that the answer sounded plausible.
• Respect for the audience. Disagreement, indifference and minority concerns remain part of the picture.
• Permission to be uncertain. An unresolved question becomes a research action rather than a polished answer that hides doubt.

Avoid promising that a report removes accountability, guarantees regulatory acceptance or leaves the customer with “nothing to defend.” It helps a responsible decision; it does not make the decision consequence-free.

Functional benefits

• A panel you convene in a minute on any piece of work (an ad, an email, a page, a deck): people who behave like your audience react in their own voices, score it, and disagree with each other.
• A verdict you can act on, a second reader who argues with it, a person you can interview, and a research bench that answers with citations or says the documents do not cover it.
• A one-page export the room will actually read.
• A scoped human comparison of an existing synthetic audience, or a bespoke build followed by blinded validation.
• A portable Calibration Record containing measured errors, uncertainty, relevant bases, supported uses, limits, reviewer and expiry.
• Four bounded studies a month under the proposed maintained service, with human review and scheduled drift checks.
• A visible route to more human research when the evidence is inadequate.
• Versioned exports and, after the core workflow works, controlled access through APIs and assistant integrations.

Audit targets three weeks after readiness. Builds target approximately eight weeks. Recruitment, access and agreed scope determine the actual schedule. A quarterly pulse is a drift screen, not automatic renewal of every validation claim.

Reasons to believe

These are proof requirements for the pilot and later claims. They are not achievements already established by this plan.

A properly held-out portion of seed data can test generalization within its design. Fresh respondents and stimuli test additional transfer conditions. Neither design is automatically sufficient for every decision, and a split should not be dismissed as a memory test merely because it uses an existing dataset.

Positioning

Proposed position: bespoke synthetic audiences, checked and maintained against real human research.

The entry offer makes that position concrete: an Audit of a system the customer already uses. It establishes a useful role before asking the buyer to replace a supplier or fund a new build. The maintained offer then keeps a bounded, useful audience available between human studies.

Against an integrated audience builder, emphasize a separately purchasable assessment and portable record. Against an enterprise research platform, emphasize a focused use case and clear delivery scope. Against a cheap discovery tool, explain what the added human comparison buys. Against traditional research, show how the service uses human evidence to support repeated exploration between studies.

These are positioning choices to test, not claims that competitors lack human grounding, disclosure or independent evaluation. The market assessment identifies close overlaps. Do not claim to be the first neutral provider or suggest that neutrality follows automatically from the name.

Audience

The buyer is the executive who owns a nine-figure marketing budget and a seven-figure line of agency strategy and research, in a consumer business of about $1bn or more: travel, financial services, automotive, healthcare, education, premium durables. They have already seen a synthetic audience flatter itself. They want to test more, faster, and they have to defend the result to a board.

The buyer's question: "How far is this synthetic audience from the people we are paying to reach, and can I show that?"

Never their question: "Which synthetic tool should the research team buy?"

Three audiences, in priority order:

1. The chief marketing or customer officer and their brand and insights leads. They buy a programme the way they buy a tracker or a strategy workstream, and they expect a named person in the room with the agency.
2. The holding-company agency or consultancy serving them. They need an independent record they can put in front of the client, and a safer way to be right.
3. The synthetic-audience vendor. They want their product measured by a neutral party and the result published.

Markets: English-language first (US, UK, Australia, Canada), because human holdouts are recruited and interviewed in the audience's language. Price in USD.

Products

Prices are hypotheses until three clients have paid them. Say so. The single-audience audit and build of earlier drafts exist as units of work inside the programme, not as products.

Look and feel direction

The future identity should make evidence easy to read and people easy to remember. A useful reference is the clarity of a well-edited research report, with enough warmth to avoid making a boutique service feel like an impersonal institution. This is direction for a later design brief, not a selected visual system.

The later design should make source, generated interpretation and measured result distinguishable; give distributions and uncertainty readable space; work in print and on screen; and avoid imagery that could imply a generated persona is a photographed research participant. Numbers, notes and qualifiers must remain legible. Status should never depend on color alone.

Explore a restrained identity whose distinctiveness comes from consistent editorial judgment and useful evidence. Consider a recurring human-versus-synthetic comparison as a content pattern, but do not fabricate charts for decoration. Assess any future palette, type, photography, illustration and motion choices against clarity, accessibility and truthful representation. No specific colors, fonts, marks, layouts or assets are chosen here.

Tone of voice

Measured, direct and useful. Explain the result before the method, then make the method available. Use ordinary words in customer actions and precise research terms where the reader needs them.

We say: checked against people; human evidence; synthetic response; Calibration Record; scope; comparison; uncertainty; inconclusive; next human check; what changed; what to do next.

We explain when needed: holdout, effective sample, distribution, calibration and drift. Do not force specialist terminology into every headline or button.

We avoid: guaranteed truth, a universal accuracy percentage, certification without accreditation, human-level as an undefined claim, replacement for people, an unsupported “first” or “only,” and a model-written confidence number presented as measured certainty.

A quantitative research claim includes its metric, base, scope and uncertainty where appropriate. This does not mean a price, date or count needs a statistical interval. Label every illustrative example and every generated quotation. Distinguish historical validation from a new study's actual human base.

Illustrative customer copy:

• Before a study: “This audience was checked for message appeal in this market. Your question asks about price sensitivity. That needs a different study.”
• At a result: “No new human responses were collected for this run. The historical comparison and its limits are below.”
• On drift: “The model changed. This audience is paused for the affected task while we check it again.”
• On an inconclusive Audit: “The evidence is not precise enough to support this use. Here is the smallest next study that would help.”

Competitive voice references

Decisions deferred

Logo, wordmark treatment, palette, typography, imagery, motion, screen layouts and final tagline remain open. Public founder visibility is a preference separate from the legal identity and accountability customers require. A name/domain and written platform are sufficient for the next buyer conversations; identity production should follow a clearer offer and evidence from the pilot.