Based on your answers, this assessment suggests improvement priorities; it does not independently verify your business or predict AI recommendations, sales, or business quality.
One offer. Seven areas.
Answer for one main offer throughout. The bank contains 27 questions; the machine-use question appears only when you report an interactive tool. A useful static guide is a valid form of decision support.
Unknown is different from absent.
“Not sure” and “Skip for now” remain distinct unknowns. Only the specific credential exception and the conditional machine-use question permit an exclusion. Conflicting answers are clarified; unresolved disputes remain unknown, with any reported risk still visible.
Four descriptive statuses
NEEDS CHECKING
We don’t have enough information yet, or something important still needs verification.
NEEDS ATTENTION
A known foundational condition is missing.
PARTLY IN PLACE
Some of the condition is present, but not enough to call it complete.
IN PLACE BY YOUR ACCOUNT
You report that the applicable condition is fully in place.
How we determine this
Each dimension requires at least two applicable items and at least 75% known, rounded upward, before a substantive status. Anchors are U1, E1, F2, A1, T1, C1, and M1. Known conditions have values 0 (absent), 1 (partial), or 2 (fully in place); unknowns have no value.
- If there are fewer than two applicable items, or known items are fewer than ceil(0.75 × applicable items), the status is Needs checking.
- Otherwise, if the anchor is 0, or the sum of known values is less than the number of known items, the status is Needs attention. This also covers known conditions that are mostly missing.
- Otherwise, if any known value is below 2, the status is Partly in place.
- Otherwise, if every applicable item is known and every value is 2, the status is In place by your account.
- Otherwise, the status is Needs checking: unknowns remain alongside otherwise fully positive answers.
E2 may be excluded only by its approved credential exception. A4 is applicable for an interactive resource, excluded for a reported static resource or no interactive tool, and unresolved when F2 is unknown or disputed. Excluded items do not count toward coverage. Disputed answers remain unknown unless clarified as referring to different resources; scope qualifications remain visible. Explicit 0H risk reports are retained even when a related answer is disputed.
A full seven-dimension orientation requires at least 14 known applicable answers and one known answer in every area. Otherwise the result is a Preliminary picture. A reported absent anchor remains a visible foundation gap even if coverage is insufficient for a substantive status.
Exactly three next actions
We don’t simply pick your three lowest areas.
The assessment looks first for important risks, then missing foundations, prerequisites, ordinary gaps and things that still need checking. Advanced machine-access work comes later because there is little value in making a business easier for machines to use before the business itself is clear, useful and trustworthy.
View the exact prioritization rules
A fixed catalog selects the actions. No AI model substitutes its judgment at runtime. An action becomes a candidate when a mapped, applicable condition is not fully in place. Related triggers are merged into that action.
Priority tiers, in order
- AC00: one bounded review action for all explicit 0H risks reported at E1, T3, T4 or C3. It absorbs the corresponding AC04, AC14, AC15 and AC17 repairs.
- Actions mapped to an absent anchor: U1, E1, F2, A1, T1, C1 or M1.
- Actions that establish an unmet prerequisite for another candidate (excluding AC11 from dependency reach), or check an unknown anchor or F1.
- Other specific reported gaps.
- Unknown-only checks.
- AC11: optional machine-access work, subject to its stricter gate.
Dependencies and tie-breaks
Normal prerequisites must already have value 1 or 2. Selecting a prerequisite repair does not make dependent work eligible. AC11 additionally requires U1, U2, E1, F1, F4, A1, A2, A3, T1, T4, C1 and C3 at 2; F2 at 2I and not disputed; E2 at 2 or properly excluded; and no active 0H risk.
Sort by tier ascending, severity descending, dependency reach descending, catalog decision value descending, evidence basis descending, effort ascending, then numeric action ID ascending. Severity is 3 for AC00 or a trigger at 0, 2 for a trigger at 1, otherwise 1. Dependency reach counts other non-AC11 candidates whose unmet prerequisites intersect the action’s mapped conditions. Evidence basis is 0 for unknown-only checks and 1 otherwise. Choose the first three eligible candidates.
| Action | Mapped items | Prerequisites | Value / effort |
|---|---|---|---|
| AC00: Review the specific practices you flagged | None | 3 / 2 | |
| AC01: Make the offer and fit clear | U1, U2 | None | 3 / 1 |
| AC02: Align the public facts | U3 | U1 | 2 / 1 |
| AC03: Explain cost and commitment | U4 | None | 3 / 1 |
| AC04: Put support beside important claims | E1 | None | 3 / 2 |
| AC05: Make credential claims checkable | E2 | None | 2 / 1 |
| AC06: Add context and a review habit to evidence | E3, E4 | None | 2 / 2 |
| AC07: Choose one customer decision to help with | F1 | None | 3 / 1 |
| AC08: Build or improve one useful decision aid | F2, F3, F4 | U1, F1, E1, T1 | 3 / 2 |
| AC09: Make essential information publicly readable | A1, A3 | U1 | 3 / 2 |
| AC10: Test the customer path | A2, C2 | A1 | 3 / 2 |
| AC11: Verify whether the interactive tool can be used through AI | A4 | None + stricter gate above | 1 / 3 |
| AC12: State where the guidance stops | T1 | None | 3 / 1 |
| AC13: Include a legitimate alternative to buying | T2 | None | 3 / 1 |
| AC14: Make commercial interests clear | T3 | None | 3 / 1 |
| AC15: Reduce and explain information requests | T4 | None | 3 / 2 |
| AC16: Give the customer a clear next step | C1 | U1 | 3 / 1 |
| AC17: Clarify commitments and authority | C3 | None | 3 / 2 |
| AC18: Set realistic response expectations | C4 | None | 2 / 1 |
| AC19: Define and record a useful completed step | M1 | C1 | 2 / 2 |
| AC20: Learn whether the help clarified the choice | M2 | F2, C1, M1 | 3 / 2 |
| AC21: Keep a repeatable record of AI responses | M3 | U1, A1, E1 | 2 / 2 |
Checks and completion to three actions
Unknown-only candidates ask the owner to establish what is present, missing or uncertain before undertaking a repair. With no known answers, the fixed plan is AC01 (check U1), AC07 (check F1), then AC16 (check C1).
If fewer than three repairs or checks are justified, append eligible validation/maintenance actions in this order. They are optional activities, not diagnosed weaknesses:
- V01, validate one important claim: excluded by AC04, AC05 or AC06.
- V02, recheck one customer journey: excluded by AC09, AC10, AC16, AC17 or AC18.
- V03, observe whether the help clarifies a choice: excluded by AC07, AC08 or AC20, and requires F2 to be known and above 0.
- V04, recheck the public explanation: excluded by AC01, AC02 or AC03.
- V05, set a review trigger: no exclusion.
For exclusions, AC00 also counts as the repairs it absorbed. Stop when the plan contains exactly three unique actions. No overall score, grade, ranking, probability or prediction is produced.
Evidence layers
V1 uses only BY YOUR ACCOUNT. A future public observation, technical test, or behavioral observation would need its own source, date, scope, permissions, and limitations. Those future layers are not implied by completing this assessment.
What still needs evaluation?
The timing estimate, coverage thresholds, anchors, effort/value judgments and action ordering are pilot conventions. They are not calibrated maturity measures. A consented user pilot should test comprehension, timing, burden, and whether the first action is useful. Completion and conversion alone do not establish a Better Choice.
Ruleset: de-readiness-1.0.0. Read the exact question bank and input contract. View the change record.