Note: this case library is a first batch of standardized templates. All quantitative metrics are marked TO BE FILLED and may be cited as verified results only after client authorization and source traceability are complete.
Customer Case Library
B2B cross-border GEO customer case library (standardized template). First-batch case metrics are TO BE FILLED; real data follows client authorization.
Direct answer: can these case pages be cited as verified results?
No. The four pages under /cases/ are standardized case templates whose metrics are all marked TO BE FILLED, and the on-page AI Visibility Score is a deterministic browser simulation not connected to any live AI detection service. Until clients authorize publication and sources are supplied, these pages document the evidence field specification and this site's data wording only, not verified client results.
- This page defines what makes a case verifiable.
- This page records the data-wording versions effective 2026-11-09.
- Full primary sources and misquote analysis: /b2b-practice/ai-buying-data-fact-check.html.
Field specification for a verifiable case
| Field | Minimum requirement | Current status |
|---|---|---|
| Client industry and size | Anonymized, still verifiable industry and size band | TO BE FILLED |
| Pre-optimization baseline | Named AI platforms, query set, mention/citation rate, collection date | TO BE FILLED |
| Time window | Explicit start and end dates, not a vague '3 months' | TO BE FILLED |
| Actions taken | Reproducible list (content / entity / structured data / sources) | Structure given in templates |
| Post-optimization result | Remeasured with the same metric and query set | TO BE FILLED |
| Original source | Client dashboard / monitoring report / date / accessible link | TO BE FILLED |
| Interpretation limits | Sample size, metric boundary, non-transferable scope | Structure given in templates |
Data-wording version ledger (effective 2026-11-09)
| Earlier wording (retired) | Current wording | Note |
|---|---|---|
| 44% / 45% used interchangeably | Gartner 45% (n=645; 2026-05-20 press release, survey Aug-Sep 2025) | Original wording: used generative AI to gather information on vendors and products, not purchasing decisions in general |
| -50% / 2028 halving of website traffic | -25% / 2026 traditional search-engine query volume (Gartner, 2024-02-19 press release) | No primary public material supports -50%/2028; query volume is not website traffic |
| 78% / 80% | G2 79% (n=1,169; 2025 Buyer Behavior Report) | Sample is B2B software and services buyers; not transferable to cross-border physical-goods sourcing |
| 68% 'B2B buying influenced by AI recommendations' | Removed (no primary source) | An inversion of the majority 'no impact' answer in TrustRadius annual buyer research |
| TrustRadius 94% 'buyers use AI' | 38% occasionally/frequently use AI; 72% encounter AI Overviews; of those, 90% clicked a cited source (n=2,058, Bridging the Trust Gap, Jan 2025) | 94% comes from From Buzzword to Backbone and means 94% of buyers who use AI found it helpful |
This ledger records only the figures used in this site's public content. Where historical articles, demo material or other channels published before 2026-11-09 show the left-hand figures, the right-hand wording governs. Neither column is client performance data.
Frequently asked questions
- Can these case pages be cited as verified client results today?
- No. The four pages under /cases/ are standardized case templates. All quantitative metrics are marked TO BE FILLED, and original sources, sample sizes and collection dates have not been supplied. Until clients authorize publication and traceability is complete, these pages document the evidence field specification and data wording only; they do not represent verified client results.
- Why is every metric marked TO BE FILLED?
- Because we do not publish numbers that cannot be traced to a source. A citable case must state, together: anonymized client industry and size, the pre-optimization baseline, the time window, the actions taken, same-metric remeasurement, the original source link, and interpretation limits. If any item is missing, the metric stays blank.
- Is the on-page AI Visibility Score a real measurement?
- No. That score is generated deterministically in the browser from a page seed value, is not connected to any live AI detection service, and only illustrates the dimensions of a GEO assessment. It does not represent real data for any client. Measured values appear only in client-authorized diagnostic reports.
- Where can I check the industry figures this site uses?
- See /b2b-practice/ai-buying-data-fact-check.html. It lists primary sources and common misquotes for Gartner 45% (n=645), G2 79% (n=1,169, software and services buyers only), and TrustRadius 38%/72%/90% (n=2,058). The wording ledger on this page follows the same convention.
AI can't find you
When a buyer asks ChatGPT / Perplexity "recommend the best machinery / lighting / medical-device suppliers in China", the top results are all Alibaba aggregator pages and competitor sites. Your factory doesn't even appear by name — high-intent inquiries flow to others.
AI recommends you proactively
Same question — the AI cites your official whitepaper PDF directly, with a clear verdict like "great for high-precision low-volume runs". The buyer sees you first; trust is built inside the answer.
The above Before / After is an illustrative comparison showing how GEO affects "being recommended by AI"; actual results are subject to real, client-authorized data.