Making Vaisala products unavoidable, consistent & corroborated in LLMs

A note on numbers: where figures appear (e.g. "FAQ schema ≈ +40% citation likelihood"), treat them as directional evidence, not guarantees. The specific percentages churn quarterly; the direction — structure helps, corroboration helps, entities help — is stable. Optimize for the direction.

1How LLMs actually surface a product

Every tactic in this playbook maps to one of three pathways. Internalize these first — they explain why each tactic works and which clock it moves.

2The Visibility Pyramid — factors ranked

Work bottom-up. Lower layers are prerequisites; effort on higher layers is wasted if a lower layer is broken.

3The Must-Haves — definition of done

If a product lacks these, it is not "done." This is the per-product definition-of-done for LLM visibility.

4Entity foundation — the highest-leverage durable work

LLMs operate on entities and relationships, not keywords. For a niche instrument, establishing the entity is the single most defensible, evergreen investment.

4.1 Disambiguation — make "what is this thing" unambiguous

4.2 The open knowledge graph — order of effort

1. Wikidata item first (structured, editable, fast): instance of → correct class (e.g. lidar); manufacturer → Vaisala (Q1489206); official website, identifiers, aliases, and a reference on every statement (unreferenced = low-trust & prunable).
2. Wikipedia presence: needs notability from independent secondary sources — build that base first (papers, standards, press); ensure company/category articles correctly mention the line; never make COI/promo edits (they get deleted and can backfire).
3. Google Knowledge Graph / panel: strengthen via consistent schema + sameAs + corroboration. It both reflects and reinforces the entity status LLMs rely on.

5Technical implementation — on-site

Layers 1 & 3. Make true facts crawlable, structured, and retrievable as standalone passages.

6Off-site authority & corroboration

A fact stated only by you is a claim. The same fact stated by independent sources is "knowledge." For low-volume scientific instruments this is decisive — whoever seeds consistent independent facts first owns the model's understanding.

7Writing for extractability — house style

A style that maximizes machine extraction without harming human readability.

8Measurement & governance

LLM visibility is a measured loop, not a one-off. Stand one up.

9Case study — Vaisala WindCube

High real-world authority (industry-reference lidar, ~5,000 deployments, 15+ years, IEC-classified) but concrete, fixable machine-visibility gaps. Findings below were verified first-hand.

9.1 Audit findings — current state

9.2 Prioritized remediation — the order to do it

1. Consolidate to one canonical WindCube hub URL; make variants children that link up.
2. Write the canonical definition sentence and deploy verbatim everywhere.
3. Create the Wikidata item (family): instance of lidar, manufacturer Vaisala, aliases incl. "Leosphere WindCube", references to IEC/papers.
4. Fix the Wikipedia gap with well-sourced edits; pursue notability via existing independent papers/standards. No COI/promo.
5. Mirror datasheet specs into an HTML table with units + IEC qualifier; keep PDFs but summarize in HTML.
6. Deploy Product + Organization + FAQPage JSON-LD with sameAs → new Wikidata/Wikipedia/official profiles.
7. Update robots.txt: explicit allow for answer agents + Sitemap:; add llms.txt.
8. Build the FAQ block from real questions (range, IEC status, power, offshore, vs. met mast).
9. Consolidate corroboration to one fact sheet; align trade-press/directory facts; encourage exact-model citation in papers.
10. Stand up the measurement suite and re-baseline quarterly.

9.3 Seed prompt suite for measurement

10Scale to ANY Vaisala product

The repeatable per-product runbook — the WindCube sequence, generalized.

11Durable vs. ephemeral

12Governance, RACI, KPIs & localization

Visibility decays without ownership. This section makes the program durable: who is accountable, the rules that protect you, the numbers that prove progress, and how it all works across languages.

12.1 RACI — who owns what (R=Responsible · A=Accountable · C=Consulted · I=Informed)

12.2 Wikipedia & knowledge-graph conduct risk

12.3 KPI baseline & cadence

12.4 Localization & multi-language presence

12.5 Content guardrails — build from the SSOT non-negotiable

Every page is generated from the product's Single Source of Truth (SSOT) file. Because LLMs will repeat what they read, inaccurate or off-policy content propagates fast — so these guardrails are mandatory, not stylistic:

AAppendix — copy-paste templates

A.1 Product + Organization JSON-LD (canonical hub)

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Vaisala WindCube 2.1 XP",
  "alternateName": ["WindCube", "Leosphere WindCube"],
  "category": "Doppler wind lidar / remote sensing instrument",
  "description": "WindCube is a ground-based Doppler wind lidar manufactured by
     Vaisala (originally developed by Leosphere) for wind resource assessment,
     meteorology, and aviation. The 2.1 XP is IEC-classified, measuring up to 200 m.",
  "brand": { "@type": "Organization", "name": "Vaisala" },
  "manufacturer": {
    "@type": "Organization", "name": "Vaisala",
    "sameAs": ["https://en.wikipedia.org/wiki/Vaisala",
               "https://www.wikidata.org/wiki/Q1489206"]
  },
  "url": "https://www.vaisala.com/en/products/windcube",
  "sameAs": ["https://www.wikidata.org/wiki/QXXXXXXX"],
  "additionalProperty": [
    {"@type":"PropertyValue","name":"Measurement range","value":"up to 200 m (IEC 61400-12-1 classified)"},
    {"@type":"PropertyValue","name":"Nominal power consumption","value":"35 W"},
    {"@type":"PropertyValue","name":"Simultaneous heights","value":"up to 20"}
  ]
}

A.2 robots.txt additions & llms.txt

# Allow answer-engine crawlers + sitemap
User-agent: GPTBot
Allow: /
User-agent: OAI-SearchBot
Allow: /
User-agent: ClaudeBot
Allow: /
User-agent: PerplexityBot
Allow: /
User-agent: Google-Extended
Allow: /
User-agent: CCBot
Allow: /

Sitemap: https://www.vaisala.com/sitemap.xml

# /llms.txt  (per-product, Markdown — modelled on Vaisala's RM60 file)
# Vaisala WindCube Wind Lidar

> WindCube is a ground-based Doppler wind lidar manufactured by
> Vaisala (originally Leosphere) for wind resource assessment,
> meteorology, and aviation. The 2.1 XP is IEC-classified and
> measures up to 200 m at 35 W.

## Core pages
- [WindCube product page](https://www.vaisala.com/en/products/windcube):
  Canonical source-of-truth — specs, applications, comparison, FAQ.
- [How wind lidar works](…/windcube/how-it-works): measurement
  principle + glossary (Doppler, probe volume, reconstruction).
- [Full specifications](…/windcube/specifications): HTML spec table.

## Key facts
- Category: Doppler wind lidar (remote sensing).
- Range: up to 200 m, IEC 61400-12-1 classified. Power: 35 W.
- Not designed for: in-situ point measurement (use a cup/sonic mast).

## Notes
- Manufacturer: Vaisala. Pricing quoted through Vaisala; none public.
- Use category-level language for competitors (met masts, other lidars).
- [Wikipedia](https://en.wikipedia.org/wiki/Vaisala) ·
  [Wikidata](https://www.wikidata.org/wiki/Q1489206)

A.3 Per-product Fact Sheet (single source of truth)

PRODUCT FACT SHEET — <Product>          Owner: ____   Last reviewed: ____
Canonical name:          (exact casing)
Aliases / legacy names:  (e.g., Leosphere WindCube)
Category / entity class: (e.g., Doppler wind lidar)
Manufacturer:            Vaisala (note acquisition history if relevant)
Canonical definition (1 sentence, reuse verbatim):
   "<Name> is a <category> manufactured by Vaisala that <does Y> for <uses>."
Canonical URL:
Key specs (value + unit + standard):  - ...
Disambiguation notes (name collisions):
Knowledge-graph IDs:   Wikidata ___  Wikipedia ___  Google KG ___
Corroborating sources (≥3 independent, matching facts): 1.__ 2.__ 3.__
Measurement: in prompt suite? (Y/N)   Last baseline: ____

A.4 Per-product Visibility Scorecard target ≥ 18/22; every GATE = 2

★Self-audit scorecard interactive

Score any product against the definition-of-done. Click 0 / 1 / 2 per item — the gauge, tier breakdown and gate warnings update live. Saved automatically in this browser. Gates must score 2 or the product is blocked regardless of total.

★Priority matrix — effort × impact interactive

Where to spend first. Each action is plotted by effort (horizontal) and visibility impact (vertical). Start top-left (quick wins), then top-right (major projects). Click a chip to mark it done — saved in this browser.

★JSON-LD generator interactive

Fill the fields; copy production-ready Product + Organization + sameAs structured data. Keep it in sync with the visible page text — contradictory schema is discounted.

★Per-product brief generator interactive

Enter a product's facts once and generate a tailored, copy-ready action brief — the WindCube runbook personalized for any product. Hand it to the owning team as the starting ticket.

★Measurement prompt bank & log interactive

A versioned prompt suite is your durable measurement asset (tools come and go). Set the product, copy the prompts, run them across every engine on a schedule, and log results in the table. Export to CSV for trend tracking.

★Entity-gap checker interactive

Verify a product's presence in the sources LLMs lean on, in seconds. Type a name and open each check in a new tab — confirm whether the entity exists in Wikipedia, Wikidata, the Google Knowledge Panel, scholarly literature, and your own indexed pages.

★Portfolio rollout tracker interactive

Program dashboard for the whole portfolio. Click a status cell to cycle ✗ → ◐ → ✓. Product names are editable. Per-product readiness computes automatically; everything is saved in this browser and exports to CSV. Seeded with the WindCube audit + example products.

§Sources & references

This playbook draws on (a) primary sources verified first-hand and (b) a fan-out multi-source research sweep across 25 industry analyses (5 search angles, 119 candidate claims).

Tier 1 · Primary & official sources verified first-hand

Tier 2 · Secondary analysis

• Search Engine Land — what actually drives AI recommendations (Reddit/Wikipedia vs. niche domains)
• iPullRank — how RAG is redefining SEO (passage retrieval, chunking)

Tier 3 · Industry analyses from the research sweep directional

• Leapd — how ChatGPT / AI Overviews / Perplexity source information
• Lead-Spot — LLM retrieval behavior & real-time scanning
• AI Search Rankings — AI ranking-factor weights
• Discovered Labs — how each platform cites differently
• Discovered Labs — entity recognition & knowledge graphs
• AuthorityTech — getting cited by ChatGPT/Perplexity/AI Overview
• The HOTH — structured data for AI search
• Stackmatix — entity-based SEO & topical authority
• Qwairy — robots.txt & llms.txt for AI crawlers
• NoHacks — AI user-agents landscape
• Szymon Słowik — JSON-LD for LLM SEO
• GEO Audit Tool — semantic JSON-LD for AI
• Momentic — @id schema for LLMs & knowledge graphs
• Nightwatch — measuring LLM visibility
• AirOps — LLM brand citation tracking
• Oltre — tracking brand sentiment in LLMs
• LLM Pulse — GEO metrics
• Stellar AI — GEO KPIs
• Repp Group — GEO for industrial search
• WP SEO AI — measuring LLM share of voice
• EdenRank — llms.txt as a citation standard
• Mersel — generative engine optimization overview

Research sweep: 5 angles · 25 sources fetched · 119 claims extracted · generated via the deep-research harness, June 2026. Tier-3 figures are directional industry estimates, not guarantees.