Xinhui Chenpi Brand: A 90-Day GEO Repair Path Starting from Conflicting Brand Facts
The engagement started with public search diagnosis and conflicting brand facts, then organized a 90-day GEO repair path covering entity facts, evidence-led content, technical crawlability and fixed AI probes.
AI-generated category illustration, not a photograph of the client's products or project.
This case comes from a real project and the published content has been reviewed. The client name and identifying details are withheld within confidentiality and public-authorization boundaries.
Project evidence and disclosure scope
- Evidence as of
- 2026-08-28
- Evidence status
- Public search facts verified; AI outcome measurement pending
- Basis
- CASE-GEO-01 records the public search diagnosis, fact conflicts and 90-day action plan.
- Public boundary
- No public real-world AI before-and-after measurement loop exists yet. The historical 30.5% figure was a virtual estimate and must not be presented as a verified outcome.
Project Background
The client is a Xinhui chenpi and premium tea brand. Public information was fragmented across pages and platforms, and key facts lacked a consistent, verifiable source that search engines and AI systems could cite.
Core Problems
- Company, brand, product and origin facts were not expressed consistently.
- Public pages did not answer high-intent user questions directly.
- Evidence sources and disclosure boundaries were not clearly recorded.
- No fixed AI probe set or repeatable visibility baseline had been established.
Diagnosis Findings
The immediate issue was not content volume alone. The brand first needed one consistent entity definition, verifiable evidence and clear boundaries before expanding content or measuring AI visibility over time.
Action Direction
- Audit public search results and record conflicting brand facts.
- Build a fact ledger with sources, dates, owners and public authorization status.
- Organize content with the Claim, Evidence, Boundary and Action framework.
- Plan a fixed AI probe set and retest schedule for T0, T+14, T+30, T+60 and T+90.
- Create a 90-day repair path across the official website, structured data and selected external sources.
Why start with this work?
When company, people and origin information conflict, adding articles can spread inconsistent versions. The diagnosis first separates sourced facts from unresolved items, then organises technical work, content and fixed questions for later measurement.
DELIVERABLE OUTLINE
What would the working material look like?
An outline based on the published working approach, not the client's original file. It contains no client data or measured results.
| Deliverable | What it records | What to check |
|---|---|---|
| Brand fact-check sheet | Record company and brand relationships, people, origin, product facts, sources and conflicts. | Separate verified facts from pending checks and preserve source links. |
| Fixed-question baseline | Keep question IDs, platforms, modes, times, answers and citations. | Leave missing tests blank and measure mentions, accuracy and citations separately. |
| 90-day action list | Organise fact alignment, technical repair, content work and later measurement. | Assign owners and deliverables and distinguish targets from completed results. |
What your team can take from it
Give each core fact a source, owner and disclosure scope before agreeing a repeatable question set. The brand confirms product and company materials; consulting organises the diagnosis and priorities. This case completed a public-search diagnosis and 90-day roadmap, not a demonstrated AI recommendation uplift.
Before the next phase
Before implementation, the brand needs to confirm one fact version and disclosure scope and collect actual AI baseline answers. Later outcomes require comparable recorded tests.
Stage Result
The public search diagnosis and 90-day method have been documented. As of August 28, 2026, there is no public real-world AI before-and-after measurement loop, so no uplift percentage is published.
Project Insight
GEO credibility begins with consistent facts and repeatable measurement, not with an unsupported visibility claim.
If your brand faces a similar problem, start with a basic diagnosis.
If your brand is described inconsistently or rarely cited in AI answers, request a basic public-information GEO report.
