gooseworks-ai/goose-skills / skills/marketing/improve-search-visibility
improve-search-visibility
VerifiedSEO and AI-answer visibility. Use for a site SEO or technical audit with ranked fixes, competitor keyword and topic gaps, a topic cluster, SEO metrics for domains, a programmatic SEO page-set plan, or checking where ChatGPT and Google name the brand. Delivers the requested audit, comparison, map or blueprint with dated evidence and clear data limits. Also covers app store listings and local search. Drafts changes; never edits live. Full pages go to make-content; paid search goes to run-paid-campaigns.
references/question-list.md
View on GitHub# Building the question set
Load at step 1 (atom question-set). No approval is needed before the free pass runs.
## Where questions come from (most trusted first)
1. Search Console queries (if shared).
2. The words customers use: research-audience's `research/audience/voc-bank.md` and research-market's `research/market/demand.md` (how buyers search), reviews, support tickets, sales calls; `knowledge_search` with `source_types: ["document", "evidence"]` (query "Voice-of-customer bank …") for "customer questions, objections, words customers use".
3. The brand brain's audience, pains and use cases; the product catalogue.
4. Google autocomplete and "People also ask" seen during the check.
5. Reddit and forum thread titles in the category.
6. Competitors' "vs", "alternatives" and pricing pages.
7. App-store search suggestions (apps).
## Two forms per question
- **AI form:** conversational, with the buyer's real constraints ("best moisturizer for sensitive skin that won't pill under makeup, under $40"). Answer engines split one prompt into several searches; the constraints decide which brands survive.
- **Google form:** the short query ("moisturizer for sensitive skin").
Never put the brand's name in a non-brand question.
## Mix (first run 12–15; grow to 25–40)
| Stage | Share | Pattern |
|---|---|---|
| Category discovery | ~30% | "best X for Y", "top X for [segment]" |
| Problem-led | ~20% | "how do I stop [pain]", "why does [symptom]" |
| Comparison / alternatives | ~15–20% | "A vs B", "alternatives to [leader]" |
| Price / purchase | ~10% | "how much does X cost", "is X worth it" |
| Brand / trust (accuracy) | ~15–20%, always 3–6 | "what is [brand]", "[brand] pricing", "is [brand] legit", "[brand] reviews", "does [brand] ship to [country]" |
| Local | when physical | "[service] near me", "[service] in [city] open Sunday" |
**Brand names that are common words** ("Clearday", "Notion", "Halo"): every brand question carries the category ("Clearday skincare"), and the namesake is confirmed with the human, so brand rows never read as false misses.
## Value and winnability (non-brand questions: value 1–5 × winnability 1–3)
Scored provisionally before the check (some signals only the check reveals) and re-scored after it; tell the human which battles changed.
**Value = buying intent, 1–5:** 5 — maps straight to what the brand sells ("best AI ad tool for DTC brands"); 4 — the buyer has the problem but no product in mind ("how to make more ad creative without a designer"); 3 — an adjacent pain ("Meta ad rejection reasons"); 2 — educational, tangential ("what is a UGC ad"); 1 — traffic that rarely buys. Keep questions scoring 3 or more; a 1–2 question earns a place only as the anchor of a hub page that the buying questions link up to.
**Winnability, 1–3.** Winnable when: the brand has or can make a page that answers it directly; today's answers name brands of a similar size, not only giants; the cited sources are ones the brand can influence (own site, review platforms it is on, comparison content); the brand's real difference is the deciding factor; it is not a health or money question owned by institutions. Keep the top by value × winnability; pick the 3–5 key questions yourself; the human can change them after the first message. Brand questions are about accuracy, not ranking — always kept. They include the spellings a buyer would type after hearing the name (`entity-and-local.md`, "How the name is said and misheard") when the name is easy to mishear.
## Examples (fictional)
**Halden — DTC skincare, fragrance-free barrier cream, $34:** best moisturizer for sensitive skin that doesn't sting *(discovery, key question)* · best ceramide cream under $40 *(discovery/price)* · moisturizer that won't pill under sunscreen *(problem)* · how to repair a damaged skin barrier *(problem; winnable with an expert-reviewed guide)* · why does my moisturizer burn *(problem; owned by dermatology sites — low winnability)* · Halden Barrier Cream vs [drugstore ceramide leader] *(comparison)* · cheaper alternatives to [premium barrier cream] *(alternatives)* · is Halden legit · Halden reviews · does Halden ship to Canada *(brand)*.
**Shiftwise — restaurant scheduling SaaS:** best scheduling software for restaurants *(key question)* · scheduling app for a 20-person restaurant · how to stop staff no-shows *(problem)* · 7shifts alternatives *(key question)* · Shiftwise vs 7shifts · scheduling software that integrates with Toast *(highly winnable)* · restaurant scheduling software cost per location *(price)* · free restaurant scheduling software *(low value)* · what is Shiftwise · Shiftwise pricing *(brand)*.
## The saved set (`search/questions.md`)
```
version: 1 · agreed: [date] by [human] · country: US · language: en · city: —
who ships: [you / developer / agency] · pages a month: [n] · may name competitors: [yes/no, names never to touch] · locations: [list or —]
engines: chatgpt-search, google-organic, google-aio, google-ai-mode[, perplexity]
runs·route: chatgpt 1 (key questions 3) browser | 2 pasted · aio 1 | 1 pasted · ai-mode 1 | 1 pasted · organic 1 lookup (paid top-up only)
| id | AI form | Google form | stage | value | winnable | key question | why |
```
Changing the set bumps the version; diffs compare only questions present in both versions with the same id (a reworded question gets a new id; the old one stays as `retired (reworded → Q<n>)`). Edits the human makes after the first message apply from the next questions asked, and the run record names the version actually asked.
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: 194k | Updated | Claude CodeCodex | Verified | Install |