# Sample Audit — "Company A": Flat-Rate All-In-One Client-Portal SaaS

**Hundred Euro Labs · AI Search Visibility SNAPSHOT · 2026-08-25**
Tier: **Snapshot** (Method v1.0 §5.1: reduced battery, one assistant surface + citation layer; *directional, not multi-engine*). Anonymized public sample — identity withheld; all measurements are real and reproducible from the sources described.

> **How this was measured (honesty box).** Citation layer: live web-search runs executed 2026-08-25 on three battery queries representative of intent classes C1/C4/C6 (category head, alternatives, brand trust), top results logged per run. Assistant layer: unaided recall check across query paraphrases performed by the audit agent itself — a frontier LLM — single pass per paraphrase. Disclosed limitation: one assistant surface, one run set, training-data bound. No ChatGPT/Gemini/Perplexity API or UI runs are claimed; the full multi-engine protocol is the paid standard tier.

---

## 1. Subject profile (anonymized)

Bootstrapped all-in-one **client-portal / agency-management SaaS**: CRM, client portals, projects, invoicing, white-labeling; flat-rate pricing ($19/$49/$99/mo, unlimited users). Competes against per-seat rivals (Copilot/Assembly, HoneyBook, Dubsado, Plutio, Moxo…). Buyer: SMB service businesses and agencies. Mature product (15+ years), large review-platform base, thick affiliate/comparison ecosystem.

## 2. Battery slice run

| Query (as run) | Class | What it tests |
|---|---|---|
| "best client portal software for agencies 2026" | C1 head | Do roundups include the brand at shortlisting time? |
| "[Brand] alternatives comparison" | C4 switching | Who controls the exit narrative? |
| "[Brand] review pricing client portal" | C6 trust | What do buyers read when validating? |

Assistant-recall checks mirrored these plus a C3 pair ("[Brand] vs [top competitor]").

## 3. Findings

### 3.1 Presence — solid mid-funnel, leaking at the head

**Citation layer (C1 query): 7 distinct category roundups sampled → brand present in 4 (≈57%).**

| Roundup (type) | Included? | Framing when included |
|---|---|---|
| Independent 9-tool buyer's guide | ✅ | "**Best value for most small agencies**" — its #1 verdict |
| 24-tool mega-comparison | ✅ | #2 of 24; praised flat-rate white-label |
| Workspace-vs-dedicated guide | ✅ | Featured dedicated-portal pick |
| Competitor's own roundup | ✅ | Listed behind the publisher |
| Newer rival's 8-tool guide | ❌ | Absent entirely |
| Agency-workflow blog guide | ❌ | Absent |
| No-code vendor top-5 | ❌ | Absent |

Every omission is a buyer who never learns the brand exists at the exact moment of shortlisting. The misses cluster around newer, aggressively SEO'd competitors whose comparison pages outrank neutral guides.

**Assistant layer:** in unaided "best client portal" recalls, the brand appears inconsistently — reliably when the stated constraint is *flat pricing / unlimited seats / all-in-one*, often absent from unconstrained lists that default to Copilot, HoneyBook, Moxo, Dubsado, Basecamp. Directional estimate: present in roughly one third to one half of C1-type answers, rarely listed first.

### 3.2 Narrative — value-praised, UX-caricatured

Across the trust and alternatives queries, third-party content repeats one story arc:

- **Positive (recurring):** flat $19/mo unlimited users; white-label even on entry tier; breadth ("replaces 5–8 subscriptions"); lifetime-deal availability.
- **Negative (recurring, near-verbatim across 5+ independent pages):** *"dated interface"*, *"steep learning curve"*, *"setup takes weeks"* — one competitor page quantifies "20–40 hours over the first two weeks."

**Alternatives-SERP check (C4):** the "alternatives" results page is dominated by **competitor-owned** comparison pages (project-management and portal rivals), which open by amplifying exactly those UX criticisms before introducing their own product. The brand's own alternatives page exists but competes against half a dozen adversary-controlled narratives. In assistant recalls, the same tropes resurface almost verbatim — evidence that training data mirrors the SERP's affiliate layer. **This is the audit's core insight: the company's AI-search narrative is currently written by its competitors.**

### 3.3 Accuracy — clean in this sample

Pricing ($19/$49/$99), plan structure, and headline features were reported correctly everywhere we checked. **AEC = 0 in this snapshot.** (In full audits this is re-verified across surfaces; stale pricing is the most common error class for SaaS.)

### 3.4 Snapshot scorecard (directional)

| Metric | Value | Read |
|---|---|---|
| Presence Rate (sampled slice) | ~55–65% | Healthy but not safe |
| Share of Voice vs 3 named rivals | mid pack | Wins budget contexts, loses default lists |
| Sentiment Index | ≈ −0.1 … +0.1 | Value praise offset by UX tropes |
| Citation Inclusion | moderate-high | Review platforms + guides cite it readily *when included* |
| Accuracy Errors | 0 | Nothing to repair — keep it that way |

## 4. Prioritized fix plan (evidence-linked)

| Prio | Action | Workstream | Evidence |
|---|---|---|---|
| P0 | Publish honest head-to-head pages for every rival that owns an "[Brand] alternatives" page — lead with the setup-time objection, answer it with real onboarding data | Comparison content | §3.2 alternatives-SERP |
| P0 | Machine-readable pricing/specs + schema markup; publish an `llms.txt` and a "who is this for / not for" page so assistants can describe the product correctly | Structured clarity | §3.2 trope mirroring |
| P0 | Systematic review-platform hygiene: respond to learning-curve reviews with concrete onboarding resources; surface recent UI updates in review responses | Review posture | §3.2 recurring negatives |
| P1 | Earn placement in the three absent category roundups via legitimate routes (product data, testing offers, news hooks) — no fake reviews, ever | Cited-source placement | §3.1 omission table |
| P1 | Case-study program with specifics assistants can cite (setup hours achieved, team sizes, migration stories) to counter the "weeks of setup" trope with sourced counter-evidence | Content placement | §3.2 |
| P2 | Wikidata/Wikipedia eligibility assessment; industry-directory presence audit | Knowledge-base entities | Method §7.5 |

**What success looks like in 14 days (verification re-run):** presence in ≥1 previously-missing roundup process started; own comparison pages indexed for C4 queries; zero new accuracy errors. Not promised: ranking or recommendation changes — those lag and are never guaranteed.

---

*Snapshot limitations apply (Method §10): single assistant surface, single run set, citation layer from web search only, one date. Full audits run 40–60 queries × multiple engines × n=3 runs. Company identity withheld; source list available on request.*