## P.R.E.X.™ ANALYSIS — UX AUDIT REPORT --- ### ── FINDING 01 ──────────────────────────────── **TYPE:** Pattern Signal **SIGNAL:** Mixed format — prose lead + bullet observations + imperative recommendation **EVIDENCE:** Opening paragraph in declarative prose, mid-section shifts to a 3-item bullet list, closes with a "Recommendation:" block in imperative voice **INFERENCE:** The original prompt instructed a structured audit format with distinct sections for diagnosis, evidence, and action — likely via explicit section labels or a template scaffold **CONFIDENCE:** 87% --- ### ── FINDING 02 ──────────────────────────────── **TYPE:** Role Signal **SIGNAL:** Practitioner-level UX vocabulary with measured authority **EVIDENCE:** Terms like "friction," "abandonment risk," "CTA language," "post-activation," "profile enrichment" — all domain-precise without being academic **INFERENCE:** Role was defined as a UX/product specialist, not a general consultant. The vocabulary suggests a prompt that named the domain explicitly (UX audit, onboarding review, or product analysis) **CONFIDENCE:** 85% --- ### ── FINDING 03 ──────────────────────────────── **TYPE:** Pattern Signal **SIGNAL:** Flat assertion style — no hedging, no qualifications **EVIDENCE:** "introduces unnecessary friction," "increases abandonment risk," "Reduce verification steps" — zero softeners ("may," "might consider," "could") **INFERENCE:** The prompt either specified a direct, confident tone or defined a role with high-authority stance (senior reviewer, lead auditor). No instruction to hedge was present. **CONFIDENCE:** 78% --- ### ── FINDING 04 ──────────────────────────────── **TYPE:** Constraint Signal **SIGNAL:** Deliberately shallow depth — observations are named but not quantified or explained **EVIDENCE:** "CTA language shifts between screens" is stated without specifying which screens or what the shift is; depth score of 6/10 confirms this ceiling **INFERENCE:** The prompt likely imposed a brevity constraint (word limit, "concise," or "summary format") — or depth was limited by the absence of input data (no actual screenshots or metrics were provided) **CONFIDENCE:** 72% --- ### ── FINDING 05 ──────────────────────────────── **TYPE:** Structure Signal **SIGNAL:** Numbered-action recommendation following prioritized reduction logic **EVIDENCE:** "Reduce to one primary action and delay secondary profile enrichment" — implies prioritization framework (primary vs. secondary actions), not just a list of fixes **INFERENCE:** Prompt may have included an instruction to prioritize recommendations by impact or to distinguish between critical-path and post-activation actions **CONFIDENCE:** 68% --- ### ── FINDING 06 ──────────────────────────────── **TYPE:** Persona Signal **SIGNAL:** Detached third-person analytical stance — no "I" or "we" **EVIDENCE:** No first-person language anywhere in the output; the subject is always the product or the user, never the reviewer **INFERENCE:** Role framing was analytical/advisory rather than collaborative. The prompt likely defined the output as a report artifact, not a conversational response. **CONFIDENCE:** 80% --- ## ── RECONSTRUCTION CANDIDATES ───────────────── ### CANDIDATE A — Structured UX Audit Template **Confidence: 84% Signals: 6** ``` You are a senior UX auditor specializing in SaaS onboarding flows. When given a product area or feature description, produce a concise audit report using this structure: 1. Opening diagnosis: one to two sentences identifying the primary friction point and its risk implication. 2. Key observations: three to five bullet points identifying specific, named UX issues. Be precise — name the element, name the problem. 3. Recommendation: one to two sentences of direct, actionable guidance. Prioritize the critical path action first; defer secondary improvements to post-activation stages. Tone: direct and confident. No hedging. No first-person. Format: report artifact, not conversation. ``` --- ### CANDIDATE B — Role-First with Implicit Format **Confidence: 71% Signals: 4** ``` Act as a UX product reviewer conducting a SaaS onboarding audit. Analyze the described flow for friction points, drop-off risks, and inconsistent UI patterns. Lead with the core problem, support with specific observations, and close with a prioritized recommendation. Keep it brief and professional. Use bullet points for observations. ``` --- ### CANDIDATE C — Minimal Prompt, High-Signal Role **Confidence: 58% Signals: 3** ``` You are a SaaS UX auditor. Review the following onboarding flow and identify friction points, UX inconsistencies, and one concrete improvement recommendation. Be direct and specific. ``` *Note: Below 60% confidence threshold on candidate-level basis — included as a baseline test variant only. Expect lower structural consistency across runs.* --- ## ── IMPLEMENTATION NOTE ─────────────────────── Test **CANDIDATE A** first — it replicates both the format scaffold and the role precision that account for the high Structure (8) and Specificity (8) scores. If depth remains at 6/10 and you need richer analysis, add this line to Candidate A: *"For each observation, briefly explain the UX principle being violated."* To replicate at scale for SaaS onboarding reviews, parameterize the flow description as a variable input block and keep the role and format instructions fixed. ---