A COMMUNICATIONS CANON,
OPERATIONALISED.
Our AI-augmentation layer is trained against a communications canon, then constrained by the live brief, current evidence and human accountability.
Training gives the system a serious starting point. Live evidence and explicit constraints decide what is supportable. A human owner decides what survives.
Start with the mind, not the media plan.
Al Ries and Jack Trout remain foundational for positioning, category thinking and the reality that communication competes for limited mental space.
We use those principles as a discipline: identify the position, the alternative, the proof and the sacrifice before producing more material.
Clarity, proposition and disciplined persuasion.
The house doctrine draws from enduring advertising thinkers including Claude Hopkins, David Ogilvy, Rosser Reeves and Bill Bernbach: make the proposition clear, respect the audience, test what can be tested and do not let execution bury the idea.
We do not treat any author as scripture. We keep the principles that survive the brief and the evidence.
Understand behaviour, reputation and relationships.
Persuasion is more than copy. Our framework also draws on established work in influence, public relations and stakeholder communication, including Robert Cialdini and James Grunig, alongside modern evidence and channel realities.
The purpose is not to imitate historic campaigns. It is to give the system better questions and better constraints.
Approved sources beat confident invention.
For a live brief, project facts, approved client material and trusted sources outrank model memory. When the evidence is uncertain, the system should surface the uncertainty rather than smooth it over.
The system can widen research and challenge assumptions. It cannot manufacture proof.
Not AI first drafts. Senior-review-ready work.
The system is expected to return structured research, positioning routes, message architecture, stakeholder analysis, media angles, campaign territories, executive communications, content, variants, risk papers and red-team material complete enough to review as work.
Senior-review ready does not mean automatically publishable. It means the human owner spends time judging and improving the answer rather than rescuing raw output.
Pay for judgement, not the pyramid.
AI augmentation reduces the paid production time behind research, synthesis, first drafts, versioning, adaptation and repetitive production.
We price the brief rather than a staffing pyramid. The saving is created by changing the operating model, not lowering the standard.
The machine can produce. A human has to answer for it.
Every consequential recommendation stays under human review and has a named owner. The machine can argue, draft and pressure-test. It cannot own the reputational consequence.
AI increases speed and breadth. It does not replace accountable professional judgement.
The system does not get to skip the hard questions.
What position are we trying to own? What are we willing to sacrifice? What is the specific promise? Where is the proof? Who has influence? What behaviour must change? What could destroy trust? What would a hostile journalist attack? What result are we trying to create?
The canon is useful only when it improves the questions applied to the live evidence.
The doctrine ends at the outcome.
We do not score success by how many assets the system produced. We agree what should change, choose evidence that can tell us something useful, and judge the work against that.
AI makes us faster. Doctrine keeps us good. Humans keep us accountable. Results decide whether it worked.
Give us the second look.
Tell us what you are trying to change. We will tell you whether we think we can help.
Start a project ↗