Menu Close
Close
Get in Touch

kōdōkalabs
AI Marketing Diagnosis and Readiness Assessment

HomeAI Marketing Framework - The kōdōkalabs Transformation SystemAI Marketing Diagnosis and Readiness Assessment

Diagnose: Establish the Evidence for AI Marketing Transformation

Diagnose is the first phase of The kōdōkalabs Transformation System. It establishes an evidence-based baseline of the organization’s strategy, workflows, knowledge, technology, governance, capability, and measurement before a target operating model is designed.

A diagnosis is not a generic audit, a tool inventory, a vendor proposal dressed up as analysis, or a maturity score produced without supporting evidence. Those all produce an opinion about where an organization stands. A diagnosis produces evidence — sourced, labeled by strength, and organized so leadership can see exactly what it’s based on before committing to a redesign.

Executive Summary

Diagnose exists because AI marketing transformation decisions made without a documented baseline tend to be driven by enthusiasm, anecdote, or whichever tool a competitor recently adopted – not by evidence about what’s actually constraining performance. The phase evaluates seven areas (strategic alignment, workflow reality, knowledge and data readiness, technology and integration, governance and risk, people and capability, and measurement and value), labels every finding by evidence strength, and produces a decision package that tells leadership whether to stop, pilot, move to Architect, or investigate further. It’s the evidence phase, not the fix phase — Architect is where the redesign happens.

Key Takeaways

  • Diagnose produces evidence, not opinion — every finding is labeled by evidence strength.
  • Seven evidence areas are assessed together, not as isolated audits.
  • The phase produces a specific decision package, not a general impression of “how AI-ready” the organization is.
  • A diagnosis is distinguishable from a sales-oriented audit by what it’s willing to conclude, including “not now” or “not this.”
  • Diagnose feeds directly into Architect — nothing gets designed before it’s diagnosed.

What Is the Diagnose Phase?

Diagnose is the evidence-building phase of the Transformation System. Before any workflow gets redesigned, any tool gets selected, or any budget gets allocated, Diagnose establishes what’s actually true about the organization’s current state — not what leadership assumes is true, not what a vendor proposal implies is true, and not what a generic AI-readiness quiz estimates from a handful of self-reported answers.

The distinction matters because the three most common substitutes for a real diagnosis all skip the evidence step. A tool inventory catalogs what software is installed, which says little about whether it’s used consistently or well. A vendor proposal typically diagnoses exactly the problem the vendor’s product solves. A maturity score without supporting evidence produces a number that feels authoritative but can’t be interrogated — nobody can ask “based on what?”

Why AI Marketing Transformation Requires a Baseline

Without a documented baseline, two things happen. First, later claims of improvement become impossible to substantiate credibly — if nobody recorded the “before” state, there’s no defensible way to demonstrate the “after” state is actually better, only that it’s different. Second, and more consequentially, transformation priorities end up driven by enthusiasm and visibility rather than evidence: the loudest problem, the most recently discussed technology, or the initiative a competitor announced gets resourced, while the actual highest-value or highest-risk gap goes unaddressed because nobody measured it.

A baseline solves both problems by giving Measure (Phase 5) something concrete to compare against later, and by giving Architect (Phase 2) a documented, defensible starting point for prioritization rather than a set of assumptions.

What Diagnose Evaluates

kōdōkalabs - Framework - AI Marketing Diagnose Phase - Seven-layer diagnose evidence stack
kōdōkalabs - Framework - Measure Phase - Measure to Scale Feedback Loop
kōdōkalabs’ Diagnose Evidence Stack organizes the diagnosis into seven evidence areas. This is kōdōkalabs’ own methodology for structuring a diagnosis, not an externally validated industry standard.

Strategic alignment

Objectives, ideal customer profile, positioning, growth priorities, and executive decision rights. A transformation effort disconnected from actual business priorities — however technically sound — tends to lose executive attention the moment competing priorities emerge.

Workflow reality

How work actually moves through the organization, including the exceptions, handoffs, delays, rework, and approvals that a documented process map often omits. The gap between how a workflow is supposed to work and how it actually works is frequently where the real constraint lives.

Knowledge and data readiness

Source quality, access, ownership, structure, sensitivity, and retrieval. AI-assisted work is only as reliable as the knowledge it draws on — this evidence area determines whether that knowledge is actually usable, not just theoretically present somewhere in the organization.

Technology and integration

Tools, licences, data flows, access controls, duplication, and technical constraints. This goes beyond a simple inventory to examine how tools connect (or don’t), and where redundant or conflicting systems create friction.

Governance and risk

Policies, risk classes, human review requirements, privacy, security, intellectual property, and accountability. An organization can have sophisticated technology and still carry significant undiagnosed risk if governance hasn’t kept pace.

People and capability

Role clarity, practical skill, adoption barriers, incentives, and ownership readiness. Technology and workflow redesign fail when the people expected to operate the result haven’t been assessed for whether they’re actually positioned to do so.

Measurement and value

Present baselines, costs, quality signals, commercial contribution, and reporting gaps. This evidence area asks whether the organization can currently demonstrate value from its marketing operations at all — a surprising number cannot, independent of AI. If the organization can’t establish a credible measurement baseline today, that gap itself becomes a priority finding, since it means any AI-related improvement claim made later will face the same substantiation problem.
These seven areas are assessed together, deliberately, because they interact. A workflow can look ready for AI-assisted redesign on its own terms while sitting on top of knowledge that isn’t structured well enough to support it, or governance that hasn’t caught up to the risk the redesign would introduce. Evaluating the areas in isolation — the mistake a narrow “AI readiness” audit often makes — tends to miss exactly these interaction effects, which are frequently where the highest-priority findings live.

Evidence Required for an AI Marketing Diagnosis

Every finding produced during Diagnose is labeled using kōdōkalabs’ Evidence Strength Levels – four explicitly defined levels, not an external standard:

Level
Definition

Asserted

Someone stated it; no supporting documentation or observation yet exists

Observed

A member of the diagnosis team directly witnessed the behavior or condition

Documented

Written evidence exists (a policy, a workflow document, a data record)

Measured

Quantified with a defined method, ideally against a comparison point
A high-priority finding should present both the evidence behind it and the consequence of being wrong — if a finding is only “asserted,” that’s stated plainly, not disguised as something more certain. This matters because decisions based on asserted-level evidence carry meaningfully more risk than decisions based on measured-level evidence, and leadership should be able to see that distinction before acting on a recommendation.

How the AI Marketing Maturity Model Is Used

Diagnose uses the AI Marketing Maturity Model – kōdōkalabs’ seven-stage, seven-dimension framework — as one input to the diagnosis, not the entire diagnosis itself. The maturity model estimates where the organization currently sits across the same seven dimensions the Diagnose Evidence Stack examines, providing a comparative structure. But the maturity model alone doesn’t replace the workflow-level, evidence-labeled detail a full diagnosis produces – it’s a framing tool inside a larger evidence-gathering process, not a substitute for it. The interactive AI Marketing Maturity Assessment offers a lighter-weight, self-assessed starting point for organizations not yet ready for a full Diagnose engagement.

How Workflows and Use Cases Are Assessed

Within the workflow-reality evidence area, individual workflows and candidate AI use cases are evaluated against six criteria:

Business relevance

Does this workflow connect to a genuine business priority, or is it a convenient candidate because it’s visible or easy to discuss?

Frequency and effort

How often does this workflow run, and how much effort does it currently consume? High-frequency, high-effort workflows tend to offer the clearest evidence of value if redesigned — low-frequency workflows rarely justify the investment regardless of theoretical improvement potential.

Knowledge and data requirements

What does this workflow need to know to run well, and is that knowledge currently accessible in a usable form?

Risk and reversibility

What happens if this workflow produces a wrong or harmful output, and how easily can that be corrected? Low-reversibility, high-consequence workflows require materially more governance than low-risk, easily-corrected ones.

Human judgment requirements

Which parts of this workflow require genuine judgment calls versus following a defined process? This determines what proportion of the workflow can realistically be AI-assisted versus what must remain human-led.

Measurement feasibility

Can this workflow’s performance actually be measured with available data, or would measuring it require building new instrumentation first?

Who Participates in Diagnose?

A credible diagnosis requires more than a single stakeholder’s perspective. Participants typically include an executive sponsor (accountable for the transformation decision), a business owner (accountable for the outcome the transformation is meant to achieve), a workflow owner (day-to-day responsibility for the process being examined), a subject-matter expert (domain knowledge the diagnosis team may lack), a technical representative (visibility into systems, data, and integration constraints), governance stakeholders (privacy, security, legal, or compliance perspective where relevant), and a final decision-maker (who will actually act on the diagnosis’s recommendation).

Skipping any of these roles tends to produce a diagnosis with a specific kind of blind spot – for example, omitting the workflow owner in favor of only executive interviews commonly produces a diagnosis that describes the process as documented rather than the process as actually run. Omitting governance stakeholders tends to produce a diagnosis that looks complete on workflow and technology but discovers a privacy or compliance gap only after Architect has already begun designing around the omission — a far more expensive place to discover it.

Participation doesn’t require every role to be present for every conversation. What it requires is that each perspective is captured somewhere in the evidence register before the diagnosis is considered complete, with the source of each contribution documented – a workflow owner’s observation carries different evidentiary weight than an executive’s assertion, and the diagnosis should make clear which is which.

kōdōkalabs - Framework - AI Marketing Diagnose Phase - Current state evidence map
kōdōkalabs - Framework - AI Marketing Diagnose Phase - Current state evidence map

Outputs of the Diagnose Phase

Diagnose produces a defined Decision Package, not a general summary of findings:

  • Executive problem statement
  • Current-state map
  • Maturity profile by dimension
  • Workflow and dependency inventory
  • Governance and risk gaps
  • Opportunity backlog
  • Evidence register
  • Baseline metrics
  • Priority recommendations
  • An explicit decision: stop, pilot, move to Architect, or investigate further

That last item matters as much as any other output. A diagnosis that only ever recommends “proceed to the next phase” isn’t functioning as a genuine evidence-gathering exercise — sometimes the evidence supports stopping, or investigating a specific area further before committing to anything.

Diagnose Exit Criteria

The phase is complete only when leadership can state, specifically: what problem is being solved; what evidence supports the diagnosis; which constraints matter most; which opportunities warrant design; what should not proceed; what baseline will be used later; and who owns the Architect decision. If leadership can’t answer these questions in specific terms — not general impressions — the diagnosis isn’t finished, regardless of how much time has been spent on it.

A useful test: if the answer to any of these questions is a vague restatement of the original business goal (“we want to use AI to grow marketing”) rather than a specific, evidence-backed statement (“workflow X consumes Y hours per week with Z percent rework, and the knowledge base supporting it is undocumented”), the exit criteria haven’t actually been met yet, even if a report has been produced and a meeting has been held.

Common Diagnose Failure Modes

  • Tool-first assessments – starting from “what tools should we adopt” rather than “what does the evidence show is actually constraining us.” This is the single most common failure mode, largely because it’s the easiest question to answer quickly and the one most vendors are incentivized to help answer.
  • Self-reported maturity without evidence – accepting leadership’s own assessment of organizational maturity without corroborating it against observed or documented evidence. Leadership’s view of their own organization’s maturity tends to run more optimistic than what frontline observation supports, particularly on governance and knowledge readiness.
  • Missing frontline workflow owners – diagnosing based only on executive interviews, missing the reality of how work actually happens day to day. Executives frequently describe the documented process; workflow owners describe the actual one, exceptions included.
  • Ignored governance – treating governance and risk as an afterthought rather than one of the seven core evidence areas. A diagnosis that skips governance can produce a technically sound redesign that turns out to be materially riskier than anyone realized once implementation begins.
  • Invented benchmarks – comparing the organization against unverified or fabricated industry figures rather than its own documented baseline. Comparisons to the organization’s own prior state are almost always more defensible than comparisons to an unverifiable external number.
  • Recommendations that precede findings – arriving at a preferred solution first, then gathering evidence selectively to support it. This is the pattern that most closely resembles a sales audit rather than a genuine diagnosis, and it’s often the hardest one for an organization to self-detect without an outside perspective.
kōdōkalabs - Framework - AI Marketing Diagnose Phase - Failure Modes

From Diagnose to Architect

Diagnose hands Architect a specific package: the evidence register, the maturity profile, the prioritized opportunity backlog, and the explicit decision about what warrants design. Architect doesn’t re-diagnose — it takes Diagnose’s findings as the input for designing a target operating model. If Architect discovers during design work that the diagnosis was incomplete in a specific area, that’s a signal to return to Diagnose for that area specifically, not to proceed on an assumption.

Frequently Asked Questions - Diagnose Phase - (FAQ)

Because tool selection driven by assumption rather than evidence tends to solve the wrong problem — or a real problem that isn't actually the organization's highest-priority constraint.
The same seven the Diagnose Evidence Stack covers: strategic alignment, workflow reality, knowledge and data readiness, technology and integration, governance and risk, people and capability, and measurement and value.
A mix of asserted, observed, documented, and measured evidence across all seven areas — with the diagnosis being explicit about which level supports each finding, rather than presenting everything with equal confidence.
At minimum an executive sponsor, business owner, workflow owner, subject-matter expert, technical representative, relevant governance stakeholders, and the final decision-maker — see "Who Participates in Diagnose?" above.
Against six criteria: business relevance, frequency and effort, knowledge and data requirements, risk and reversibility, human judgment requirements, and measurement feasibility.
Higher-risk, lower-reversibility workflows require more governance design before they're prioritized for redesign, regardless of how attractive their potential value looks in isolation.
The full Decision Package: problem statement, current-state map, maturity profile, workflow inventory, governance gaps, opportunity backlog, evidence register, baseline metrics, priority recommendations, and an explicit next-step decision.
Diagnose's Decision Package becomes Architect's primary input — Architect designs the target operating model based on what Diagnose found, rather than starting from a blank page or an assumption.
Yes, in principle — the Diagnose Evidence Stack and Evidence Strength Levels are a methodology an internal team can apply. In practice, internal diagnoses often struggle with the same blind spots an outside perspective helps surface, particularly around self-reported maturity and governance gaps nobody inside the organization is incentivized to flag.
A sales audit is structured to arrive at a predetermined recommendation — usually the auditor's own product or service. A genuine diagnosis is willing to conclude "stop" or "not this," and shows its evidence at every level of confidence, including where the evidence is weak.

Ready to
Diagnose Your Organization's AI Marketing Readiness?

If you’re tired of fragmented, keyword-centric strategies, it’s time to build a foundation that withstands algorithm shifts. Let’s start with the blueprint that guarantees high-ROI execution.
Get In Touch With Us