Menu Close
Close
Get in Touch

kōdōkalabs How We Work

HomeAboutHow We Work

Our Engagement Principle

kōdōkalabs uses a six-phase transformation system — Diagnose, Architect, Build, Enable, Measure, Scale — and works toward governed client ownership, not permanent operational dependency. Every engagement is designed as one connected system spanning strategy, implementation, governance, enablement, measurement, and ownership transfer, rather than as a set of separate deliverables handed off in sequence. This distinction matters because it’s the most common failure mode in marketing transformation work more broadly: a vendor delivers a strategy document, a workflow redesign, or a technology implementation as a standalone artifact, and the client is left to figure out governance, training, and measurement on their own — or worse, remains dependent on the vendor to operate what was built. kōdōkalabs treats installation, enablement, and ownership transfer as one designed system from the outset, not as an afterthought once the “real” delivery work is finished.

The Six Phases of the Transformation System

The kōdōkalabs Transformation System moves through six phases:

  1. Diagnose (assess current maturity, workflows, and governance)
  2. Architect (design the target operating model and workflow structure)
  3. Build (implement the workflows, knowledge infrastructure, and governance controls)
  4. Enable (train and certify the client’s own team on the new system)
  5. Measure (establish baselines and track outcomes against them)
  6. Scale (extend proven workflows across the organization while sustaining governance and capability)

Each phase produces a specific set of deliverables and a specific handoff to the next — see the Framework page for the full detail of what each phase covers.

How an Engagement Begins

Most engagements begin with the Executive AI Marketing Assessment, which qualifies fit and produces an initial view of an organization’s current maturity across strategy, workflow, knowledge, technology, governance, capability, and measurement. From there, an engagement typically proceeds through executive sponsor alignment, confirmation of source and data access, scope definition, establishing a measurement baseline, and agreement on success criteria — before any workflow design or implementation work begins.

Establishing a measurement baseline before any implementation work starts is deliberate, not procedural box-checking. Without a documented “before” state, later claims about improvement are impossible to substantiate credibly — a pattern kōdōkalabs specifically avoids repeating both in its own client work and in what it recommends clients do internally, per the AI Marketing Maturity Model guide.

Roles and Responsibilities

Delivery involves a defined set of roles, though the specific individuals filling them vary by engagement size and scope:

Role
Responsible For

Executive sponsor (client)

Overall accountability for the engagement’s success within the client organization

Business owner (client)

Defines the business outcome the engagement is meant to achieve

Workflow owner (client)

Owns the specific workflow being redesigned, day to day

Subject-matter expert (client)

Provides domain expertise and validates factual accuracy

kōdōkalabs lead

Owns the transformation methodology, workflow design, and delivery quality

Technical or implementation resources

Build and configure the workflow, tooling, and knowledge infrastructure

Legal, privacy, or security reviewer
(client)

Reviews data handling, governance, and compliance-relevant decisions

Final approver (client)

Holds final sign-off authority at each of the three human gates below

This is not a universal contract — the specific division of responsibility is confirmed and documented for each engagement individually.

How We Use AI

AI is used throughout the Transformation System to accelerate research, drafting, analysis, and workflow design — always within the governance structure described in the AI Governance for Marketing guide and the AI Ethics & Safety Statement. AI-assisted output moves through the same risk-based human review that kōdōkalabs recommends to clients: higher-risk, higher-exposure work receives more review, not less, and final accountability for delivered work always sits with a named human, not an AI system.

Human Review and Decision Gates

Three specific points in every engagement require explicit human approval before work proceeds:

  1. Source and evidence approval — before research, data, or client-provided information is used as the basis for strategy or workflow design, its accuracy and appropriateness are confirmed.
  2. Strategic and editorial approval — before a proposed operating model, workflow design, or piece of published content moves forward, it’s reviewed against the engagement’s stated business outcome and editorial standards.
  3. Publication, deployment, or scale approval — before anything goes live, is deployed into production, or is extended beyond its initial pilot scope, a final approver signs off.

No deliverable skips these three gates, regardless of how much AI assistance was used to produce it.

These gates exist specifically because AI-assisted speed can otherwise outpace review discipline — a workflow that can produce a first draft in minutes doesn’t automatically deserve minutes of review. The gates scale in rigor with what’s at stake: internal working drafts move through them quickly, while anything reaching a client’s audience, brand, or regulatory exposure receives the full weight of review each gate is designed to provide.

Evidence and Source Management

Strategy and workflow design decisions are traced back to identifiable sources — client-provided data, documented research, or verified subject-matter input — rather than asserted without a traceable basis. Where evidence is genuinely unavailable, that gap is disclosed rather than filled with an assumption presented as fact. This applies as much to kōdōkalabs’ own internal process as to the content and workflows built for clients.

Scope, Change, and Risk Management

Engagement scope, once agreed, is managed through a defined change process rather than informal drift: a change request is documented, its impact assessed, and it’s approved before scope actually changes. Risks identified during the engagement are logged and escalated to the appropriate decision-maker rather than absorbed silently into the existing plan.

Documentation and Intellectual Property

Workflows, decision logic, and knowledge infrastructure built during an engagement are documented as part of the deliverable, not left as undocumented tacit knowledge — this is what makes the Enable and Scale phases possible.

Team Enablement

Client teams are trained through real, live workflow work rather than generic training sessions disconnected from the engagement — the Enable phase specifically pairs client team members with the actual workflows being built, with documentation, role-based practice, and structured feedback along the way. This mirrors the Capability Transfer Framework used across kōdōkalabs’ work, most directly productized in the AI Capability Academy.

Measurement and Improvement

Every engagement establishes a measurement baseline early — during the Diagnose phase — so that later claims about improvement have something concrete to compare against. Measurement covers operational metrics (workflow efficiency, consistency, cycle time), quality (accuracy, review pass rates), and business contribution (the outcomes the engagement was designed to move), with findings feeding back into ongoing refinement rather than being treated as a one-time report.

Handover and Ongoing Support

The goal of every engagement is governed client ownership: the client team can run, maintain, and extend the workflows and governance built during the engagement without depending on kōdōkalabs indefinitely. Handover is assessed against documented acceptance criteria — agreed in advance — rather than an informal sense that things are “probably fine.” Advisory or specialist support may continue after handover where a client wants it, but this is a choice, not a structural dependency the model requires.

What We Need From the Client

Successful engagements require specific client inputs: executive sponsor access and attention at key decision points; subject-matter expertise to validate accuracy; timely access to source data and systems; timely review at each of the three human gates; reasonable technical cooperation from IT or engineering where systems integration is involved; and clear decision-making ownership so that approvals don’t stall waiting for an unclear internal process. Engagements move at the pace of the slowest required input, most often client-side review turnaround rather than kōdōkalabs’ own delivery pace.

Frequently Asked Questions (FAQ)

This varies by engagement scope and phase — sponsor attention is most needed early (Diagnose and Architect) and during the three approval gates, while workflow owners and subject-matter experts are most needed during Build and Enable. Specific time expectations are set during scoping, not assumed in advance.
AI accelerates research, drafting, analysis, and workflow design; humans retain strategy, final approval, accountability, and any judgment-heavy or sensitive decision — see [How Our Solutions Fit Together] and the AI Governance for Marketing guide for the fuller breakdown of what AI assists versus executes versus never touches.
Ownership and licensing terms are set out in each engagement's signed agreement.
Through the same risk-based data classification and access controls described in the AI Governance for Marketing guide — sensitivity determines the required review level and approved handling environment, not a one-size-fits-all rule.
The gap is disclosed and flagged rather than filled with an unverified assumption presented as fact — this applies to kōdōkalabs' own internal process as much as to client-facing deliverables.
Engagements are designed to integrate with a client's existing technology and agency relationships rather than requiring their replacement, with roles and responsibilities clarified during scoping to avoid duplicated or conflicting work.
The client team operates the transformed workflows independently once handover acceptance criteria are met; ongoing advisory or specialist support remains available if the client wants it, but the model is designed to end structural dependency, not create it.

Request an Executive AI Marketing Assessment

View Our Solutions