JasonColapietro/suede-creator-skills / skills/suede-ai-seo
suede-ai-seo
VerifiedSuede Answer Engine Optimization Skills. Use when asked to "improve our AI citations" or optimize for ChatGPT/Perplexity: crawler policy, entity consistency, citation-ready pages, off-site evidence and prompt-panel measurement. NOT FOR: nine-lane technical SEO audits (suede-seo-audit), quick promotion grades (suede-visibility-grader), or funnel implementation (suede-site-alchemy).
| File | Kind | Size |
|---|---|---|
| CARD.md | markdown | 4.4 KB |
| SKILL.md | skill md | 15.9 KB |
| agents/openai.yaml | config | 574 B |
| evals/evals.json | config | 12.2 KB |
| references/ai-crawlers.md | markdown | 3.4 KB |
| references/citations-vs-recommendations.md | markdown | 4.1 KB |
| references/content-patterns.md | markdown | 5.5 KB |
| references/entity-graph.md | markdown | 4.6 KB |
| references/measurement.md | markdown | 6.6 KB |
| references/okf.md | markdown | 2.5 KB |
| references/platform-notes.md | markdown | 6.8 KB |
| references/sources.md | markdown | 2.4 KB |
| references/worked-contract.md | markdown | 2.1 KB |
references/citations-vs-recommendations.md
View on GitHub# Citations vs. Recommendations: The Suede Citation Ladder
Being cited and being recommended are different outcomes with different
causes. A citation means a page was useful to consult. A recommendation means
the engine put the brand on the buyer's shortlist. Conflating them leads
smaller brands into content plans that mostly help competitors.
## The ladder
| Rung | What it means | What governs it | How to see it |
|---|---|---|---|
| 1. Retrieved | The engine read your page while answering | Crawl access, indexation, query relevance | Bot hits in server logs |
| 2. Cited | Your URL appears as a source | Usefulness of the passage: direct answer, specific facts, clear structure, freshness | Prompt panel, Bing AI Performance, GSC |
| 3. Mentioned | Your brand is named in the answer | Entity clarity and how often the web associates you with the topic | Prompt panel |
| 4. Recommended | You are on the shortlist the buyer acts on | Consensus across sites you do not own: reviews, communities, analysts, press, video | Prompt panel framing |
There is a shadow rung: **recommended against**. On requirement-heavy prompts,
engines name products to avoid and cite why. Track the framing around every
mention (recommended, neutral, hedged, against), not just the count.
## The self-ranked listicle problem
The common play: publish "best [category] tools", rank yourself first, expect
AI answers to repeat the ranking. Practitioner studies of AI Overview
citations in 2026 observed that self-promotional listicles often earned the
citation while the answer recommended other brands. The mechanism is
straightforward: the engine treats the guide as a source about the category,
extracts the competitor names and criteria you compiled, then recommends from
broader consensus, where established players dominate.
Run this as a hypothesis on your own prompt panel before investing at scale:
for the prompts where your guide is cited, is your brand recommended?
By stage:
- **Category leaders** usually get both outcomes, because third-party sources
already corroborate them. A definitive buyer's guide shapes how the whole
category is described.
- **Emerging brands** may win the citation and influence the category's
evaluation criteria, which is real positioning work, but should not expect the
shortlist from it. Weight effort toward the off-site signals below.
Useful buyer's guides still belong in the plan at every stage. What changes is
the expectation and the budget split.
## What earns recommendations
| Channel | Why it moves recommendations | Related skill |
|---|---|---|
| Review platforms and app stores | Independent validation engines treat as evidence | `suede-customer-research`, `suede-aso` |
| Analyst and industry reports | High-authority category framing that answers echo | `suede-public-relations` |
| Communities and forums | Unprompted practitioner discussion, widely retrieved and hard to fake | `suede-community-marketing` |
| Earned media | Independent restatement of your positioning | `suede-public-relations` |
| Directories and marketplaces | Category placement with consistent entity facts | `suede-directory-submissions` |
| Video and podcasts | Transcripts tie the brand to the category | `suede-social` |
The test: if an engine ignored everything on your domain, would the rest of
the web still put you on the shortlist? If not, that gap is the priority.
## What a recommendation is worth, and why attribution undercounts it
Behavioral studies published in 2026 found that an assistant recommendation is
followed by more branded searches and site visits than a passing mention, and
that most of that follow-on traffic arrives as branded search or direct rather
than as an assistant referral. Do not quote their percentages; measure your own
with the four-source method in `measurement.md`.
## Applying this
- Report the ladder, not one "AI visibility" number.
- A rising citation share with a flat recommendation share is a specific gap:
the web does not yet corroborate your content.
- For requirement-heavy prompts, check whether engines recommend against you,
and trace the sources they cite when they do.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |