10 questions across 5 dimensions. Output isn't a score — it's a decision memo with three priorities for the quarter and the solution that fixes each. Takes about 8 minutes.
Question 1 of 100%
Data
Product Fit
Pricing
Compliance
Team
Dimension 1 · Data Readiness
Last time a buyer asked "can you show us your model's performance on our patient population?" — what happened?
Enterprise buyers increasingly require hold-out performance evidence on their own data. "It works in our lab" stalls procurement.
A
We couldn't answer
No held-out evaluation framework exists
B
We showed internal benchmarks
But they don't reflect real-world distribution
C
We ran a live pilot on their data
But documentation wasn't audit-ready
D
We showed documented, third-party validated results
Auditable and publishable
Dimension 1 · Data Readiness
How much of your training and evaluation data is labelled, governed, and traceable to a source?
Data governance is a procurement gating requirement in regulated sectors — not a post-launch concern.
A
Most is unlabelled or ad-hoc
Significant manual work required
B
Partially labelled, inconsistent provenance
Coverage gaps remain
C
Mostly labelled with minor gaps
Edge cases and some missing audit trails
D
Fully governed, labelled, and auditable
Ready for regulated-industry deployment
Dimension 2 · Product-Market Fit
When you describe what your AI does to a CFO, do you lead with the technology or the dollar outcome?
Buyers buy outcomes, not models. The biggest AI GTM failure: describing the technology instead of what changes in the business.
A
Technology-led ("our LLM is fine-tuned on…")
We lead with architecture and capabilities
B
Feature-led ("it writes X in seconds")
Good features, but ROI story is unclear
C
Outcome-led with 1–2 examples
We can quote specific impact from pilots
D
Outcome-led, buyer-validated, repeatable
Multiple buyers confirm the value and we have a consistent sales motion
Dimension 2 · Product-Market Fit
How would you characterise AI usage among your active customers today?
AI products often suffer "demo-ware" syndrome — impressive in demos but unused in production. Real PMF requires habitual usage.
A
Still in pilot / POC
No customers in production yet
B
Occasional usage
Sporadic — limited to power users
C
Regular, embedded in core workflows
Users notice when it's unavailable
D
Mission-critical
High switching cost; operational dependency
Dimension 3 · Pricing Model
When procurement asks for a single annual subscription price that covers your AI feature, can you give one without discounting on the spot?
If the answer requires back-calculating against seat count and compute exposure, the pricing architecture doesn't exist yet.
A
No — we bundle AI into existing SaaS pricing
No separate commercial model for AI
B
We have a number but always discount it
No floor — deals close wherever the buyer pushes
C
We have tiered pricing with some flexibility
But compute cost variance makes margins unpredictable
D
Yes — defended with a margin model
CFO-ready, stress-tested against compute cost scenarios
Dimension 3 · Pricing Model
Do you know your gross margin on the AI feature specifically — separate from your overall SaaS margin?
AI gross margins can be dramatically different from SaaS margins. PE boards increasingly ask for AI-specific unit economics.
A
Not tracked separately
AI infrastructure cost is buried in total opex
B
We know cloud AI spend but not revenue attribution
Can't produce a per-feature margin
C
Modelled but not yet board-reportable
Internal estimates — not tied to pricing decisions
D
Tracked, reported, and tied to pricing
Board-ready metrics with three compute-cost scenarios
Dimension 4 · Regulatory & Compliance
Has your AI product cleared an enterprise InfoSec review? If so, how long did it take?
In regulated enterprises, AI governance reviews run 3–9 months. "We can clear it" and "we've cleared it" are months apart.
A
Never attempted
We haven't mapped our product to InfoSec requirements
SOC 2 or equivalent; reusable across the portfolio
Dimension 4 · Regulatory & Compliance
Does your AI product leave an audit trail that a compliance officer can follow without calling your engineering team?
Enterprise buyers require audit trails as a procurement prerequisite — especially in healthcare, legal, and financial services.
A
No audit trail exists
Decisions are made but not logged
B
Partial logging — requires engineering to interpret
Not usable by compliance without help
C
Full logging, internally documented
But hasn't been tested against enterprise procurement
D
Compliance-ready audit trail
Passed enterprise security reviews; self-service for compliance officers
Dimension 5 · Team & Operating Model
If your best AI product manager left tomorrow, would the roadmap survive?
AI product management is a distinct discipline from SaaS PM. The skills gap is one of the most common blockers to AI commercialisation at scale.
A
No — the roadmap lives in one person's head
Engineering is leading product by default
B
We'd recover but lose 2–3 months
Context exists but isn't systematically documented
C
The roadmap is documented; the judgment isn't
We'd lose product quality for a quarter
D
The roadmap and the reasoning are both in the OS
New PM would be productive in days, not months
Dimension 5 · Team & Operating Model
Is there a single owner of AI commercial strategy who sits at the C-suite table?
Companies with a dedicated AI product executive commercialise AI 3× faster than those where strategy is distributed across engineering and product teams.
A
No dedicated owner
AI strategy is distributed — no single accountable leader
B
Informal ownership
Someone cares but it's not a formal role with authority
C
VP or Director owns it
Below the C-suite, limited board visibility
D
CPO or Chief AI Officer
Board-level accountability, integrated with commercial strategy