flo.health
Citation Rate
64%
9 of 14 prompts · baseline
Target: 85%
First-Named Rate
21%
3 of 14 · period, TTC, pregnancy
Target: 45%
Avg Sentiment
2.9
out of 5.0 · cited prompts only
Target: ≥ 3.5
Absent Prompts
5
PCOS · endo · peri · private · free
Target: ≤ 2
Branded Links ↗
—
post May 7 · run sweep to measure
New metric · ChatGPT only
Baseline: ChatGPT · 13 May 2026 · 14 prompts × 3 platforms. Click "Run Sweep" to update.

Flo Health — AI Search Visibility Audit

GEO baseline · ChatGPT · 13 May 2026 · 14 seed prompts × 5 archetypes · Prepared by Valentina D'Orefice

MetricBaseline (May 13)LiveTarget (30d)Progress
Citation Rate % 64% (9/14) 64% 85%
First-Named Rate % 21% (3/14) 21% 45%
Avg Sentiment (1–5) 2.9 / 5 2.9 ≥ 3.5
Negative Framing % 33% (3/9 cited) 33% < 10%
Absence Count 5 prompts 5 ≤ 2
Branded Links ↗ New Not measured pre-May 7 Run sweep Baseline M1

Three AI narratives about Flo today

● Category Leader
Period tracking, TTC, pregnancy. 3/14 prompts — cited first, consistently. Category dominance clear.
◑ Privacy in Remediation
Anonymous Mode + FTC response page already cited. ChatGPT credits "later changed policies" — vague. Chronology gap is closeable.
○ Specialist Gap
PCOS, endometriosis, perimenopause — 5 prompts absent. URL architecture problem, not content. /pcos, /endometriosis, /perimenopause all 404.
Drift Detection — 13 May 2026
"Flo today is much safer than it was in 2020–2021. They added Anonymous Mode and tightened policies after FTC action. But the trust damage is real — lawsuits and news stories connected to historical data-sharing practices still surface."
— ChatGPT (P05 'Flo vs Clue') · 13 May 2026 · cites FTC.gov · The Verge · Mozilla Foundation · flo.health
↑ +0.4 sentiment drift on P03, P05, P13 — narrative already softening. Chronology page accelerates this.

Specialist gap — who owns the absent prompts today

Balance
Dr. Louise Newson · Apple Editor's Choice
Owns: P06 perimenopause
P06 absent
AskPCOS
Monash University · Highest MARS score
Peer-reviewed · Owns: P11 PCOS
P11 absent
Phendo
Columbia University · Peer-reviewed
Owns: P12 endometriosis
P12 absent
Spot On
Planned Parenthood · 180+ birth control brands
Owns: P13 teens + P14 IUD
P13 NEG · P14 NEU
Drip / Euki
Open-source, on-device storage
Owns: P04 most private
P04 absent
Flo's edge
Lifecycle continuity: cycle → TTC → pregnancy → perimenopause in one app. Up to 15 years of personal data. Structurally non-copyable. No competitor has this.

14-Prompt Heatmap

Green = cited positively · Yellow = neutral/friction · Red = negative framing · Grey = absent · Baseline: ChatGPT 13 May 2026

DRIFT DETECTION · 13 May 2026 · P03, P05, P13 all show narrative softening
"Flo today is much safer than it was in 2020–2021. They added Anonymous Mode and tightened policies after FTC action."
ChatGPT P05 · cites FTC.gov, The Verge, Mozilla Foundation, flo.health

Drivers of Representation

Sources shaping how AI describes Flo — and which are already in Flo's citation set

Dominant — defines the frame
FTC.gov
The Verge
The Guardian
Wikipedia (Flo entry)
FTC complaint + Aug 2025 Meta verdict anchor the privacy narrative. .gov + tier-1 media. These define the frame ChatGPT uses for all privacy prompts.
High — specialty + privacy media
Privacy International
Teen Vogue
Financial Times
TechCrunch
Mozilla Foundation
Cited alongside tier-1 sources on privacy prompts. Shape qualifications — "privacy concerns remain" framing.
Emerging — Flo-owned (already cited)
flo.health/anonymous-mode
flo.health/newsroom/response-to-FTC (Oct 2024)
flo.health/flo-privacy-faqs ← gap
flo.health/clinical-evidence ← missing
Mechanism proven: when Flo publishes structured dated content on the right URL, ChatGPT surfaces it. The work is not "get cited" — Flo is already cited. The work is "get cited on the things that matter."

Day-1 fix: publish the privacy chronology

✓ What ChatGPT already credits Flo for
Anonymous Mode — cited by name
Response to FTC page (Oct 2024) — cited
"Later changed policies" — directionally positive
+0.4 sentiment drift on 3 prompts (13 May)
✗ What's missing from the citation set
Dated chronology 2019→2025 — no source material
Specific dates for settlement + OHTTP launch
Post-quantum layer (2023) — not surfaced
ISO 27001 + ISO 27701 dual certification — absent
14 Seed Prompts · 5 Archetypes Baseline: ChatGPT 13 May 2026 · Live data updates on sweep
IDArchetypePromptBaselineLive StatusPositionSentimentSpecificityBranded Link ↗
Select a prompt to run
Choose a prompt from the list on the left, select a platform, and click Run.
ChatGPT
Perplexity
Gemini

GEO Strategy — 30-Day Plan

No budget. No engineering. Content, schema, URL architecture only.

01 — Win the privacy conflict check · Days 1–14
Affects 5+ prompts: P03, P04, P05, P13 + all privacy-adjacent queries.
Fix: Update /flo-privacy-faqs/ — add 2019→2025 settlement chronology, reframe Q1 as 'Is Flo safe to use?', add FAQPage JSON-LD.
Why: flo.health's FTC response page (Oct 2024) is already cited by ChatGPT. The model wants to tell the full story — it doesn't have the source material. One CMS edit enters the citation set on Day 1.
Note: Google deprecated FAQ rich results in 2023. Schema now serves LLM citation selection, not SERPs.
02 — Unlock the specialty gap · Days 15–30
/perimenopause · /endometriosis · /pcos · /iud · /trying-to-conceive — ALL 404 today.
Fix: Publish 5 top-level hub pages + 2 comparison cornerstones (/flo-vs-clue/, /flo-vs-apple-health/). 301-redirect /menstrual-cycle/.../ clusters.
Why: AI engines prefer comprehensive content that answers follow-up questions in one retrieval pass. Fragmented sub-paths force chained retrievals; top-level hub pages don't.
03 — Surface clinical assets · Month 2
Mayo Clinic perimenopause study · >50% endometriosis diagnosis study · ACOG partnerships · 120+ reviewing doctors — none appear in AI answers.
Fix: Move research from /newsroom/ → /clinical-evidence/. Pitch Mayo Clinic study to Healthline + SELF. Build citation chain: peer-reviewed paper → health media → AI.
Why: Closes the Clue gap. ChatGPT frames Flo as "popular app (scale only)" vs Clue as "science-based". Clinical credibility is the structural shift that changes that framing.
Execution scaffold — 7-agent content pipeline
Audit → Answers → Topics → Angles → Writer → Publish → Monitor. Live-tested: 40 questions answered · mean citation potential 7.90/10 · total spend $5.56.
Open-source: github.com/ValentinaDorefice/geo-content-pipeline