以 AI 机器人身份加入视频会议,提供语音、虚拟形象与屏幕共享四种模式。
设计与多媒体
Meta Service Lead Gen
试用Design, audit, diagnose, and improve Meta/Facebook/Instagram lead generation funnels for service businesses that sell through applications, booking pages, nu...
它能做什么
Use this skill to help service businesses acquire qualified leads through Meta ads and a call-based funnel:
技能文档
Meta Service Lead Gen
Use this skill to help service businesses acquire qualified leads through Meta ads and a call-based funnel:
ad -> application/landing page -> booking page -> nurture -> sales call -> customer
The source material was a long course transcript. Treat it as distilled methodology, not a script to quote. Do not reproduce long course text verbatim. Strip profanity/persona and respond in clear professional language.
Scope Check
Before prescribing, classify the business:
- In scope: service business, consultative or high-ticket-ish offer, meaningful qualification, buyer usually needs a call before purchase.
- Adapt carefully: low-ticket service, no-call funnel, local emergency demand, click-to-message funnel, webinar/VSL funnel, long sales cycle B2B.
- Out of scope unless the user asks for adaptation: ecommerce/product ads, app installs, brand awareness, influencer growth, pure organic content, political/social-issue campaigns.
If the offer is in a restricted or sensitive category, read references/compliance-and-platform.md before giving campaign advice.
Operating Principles
Use these lenses repeatedly:
- Three C's: Clicks from the right people, Conversions into qualified leads/bookings, Conversations that close.
- Dials, not switches: more friction can lower lead volume but raise lead quality, show rate, and close rate.
- Sell only the next step: the ad earns the click, the landing page earns the opt-in, the booking page earns the calendar action, the nurture earns the show, the call earns the sale.
- Curious vs committed: curiosity can inflate CTR and weaken downstream conversion; committed language sets clearer expectations.
- Meet people where they are: fast, distracted, skeptical, scanning, and self-interested.
- Set expectations early: downstream objections usually existed upstream.
- Win hearts and minds: emotion creates motion; logic justifies the decision.
- Construct the argument: use several premises, proof points, FAQs, objections, and examples rather than one claim.
- Reduce physical and psychological resistance: fewer clicks, keystrokes, confusing scrolls, doubts, trust gaps, and mismatched expectations.
- Chase problems upstream: recurring sales-call objections become nurture, page, or ad messaging; poor show rate may start with booking expectations; poor lead quality may start with ad promise.
Default Workflow
-
Diagnose the funnel stage before prescribing.
- Ask for current numbers if missing: spend, CPM, link CTR, CPC, opt-in rate, booking rate, show rate, close rate, average order value, gross margin, sales-cycle length, and current profit.
- If numbers are unavailable, give a measurement plan first.
-
Identify the weakest lever.
- Use the funnel math and bottleneck rules in
references/diagnostics.md. - Optimize the lever with the highest likely profit impact, not the easiest vanity metric.
- Use the funnel math and bottleneck rules in
-
Fix with stage-specific changes.
- For ads/creative, read
references/funnel-assets.md. - For current Meta setup, signal quality, and campaign structure, read
references/current-meta-setup.md. - For restricted categories, policy, privacy, and sensitive claims, read
references/compliance-and-platform.md.
- For ads/creative, read
-
Recommend an action state.
- Build: get a good-enough funnel live so data starts arriving.
- Hold: keep running when sample size is too small or results are within range.
- Problem solve: change the bottleneck when data is clearly off.
- Let it run: scale spend once CPA/profit and capacity support it.
Benchmarks
Use these as diagnostic guardrails, not promises:
- Link CTR: roughly 1-2% is often healthy for cold Meta lead campaigns; suspiciously high CTR with poor conversion can mean curiosity/misalignment.
- Opt-in/application start or submit: roughly 5-10% at smaller budgets; may fall as spend scales colder.
- Booking rate after application: roughly 30-50%.
- Show rate: aim around 70% or better, with strong reminders and expectation setting.
- Close rate: roughly 30-45% for qualified sales-call funnels, depending on offer, price, sales skill, and lead quality.
Always anchor the final recommendation to net profit, cash collected, sales capacity, and payback period. ROAS and CPL are intermediate signals.
Response Rules
- Diagnose before prescribing. Say which lever you are addressing.
- If the user wants a template, build it fresh from the principles; do not claim it is copied from the source course.
- Distinguish platform-dependent instructions from durable intent. Meta UI names change often.
- Prefer plain business language over jargon. Define metrics briefly when useful.
- Be explicit when the user's case falls outside this method.
- Do not recommend policy evasion, discriminatory targeting, misleading claims, fake scarcity, fabricated proof, or privacy-invasive tracking.
References
references/diagnostics.md: metric formulas, bottleneck diagnosis, scaling/hold decisions.references/funnel-assets.md: ads, landing pages, VSSLs, applications, booking pages, nurture, and sales-call feedback loops.references/current-meta-setup.md: current Meta lead campaign options, Advantage+ guidance, CRM/CAPI feedback, tracking hygiene.references/compliance-and-platform.md: policy, special ad categories, privacy, claims, and platform caveats.references/response-patterns.md: reusable answer structures for audits, troubleshooting, and build requests.
相关技能
把自然语言描述转为结构化 JSON,并由 mcp-diagram-generator MCP 服务生成 Draw.io、Mermaid 或 Excalidraw 图表文件。
为自然搜索排名提供站点审计、内容撰写与竞品分析。
从 AdMapix API 拉取广告创意、应用、榜单和收入预估等数据,原样返回结构化 JSON。
在本地磁盘以分类纯 Markdown 文件保存需要长期留存的事实,与智能体内置记忆并存。
通过托管 OAuth 访问 Microsoft Graph Excel 接口,读写 OneDrive 中的工作簿、工作表、区域、表格与图表。
aiwithenoch 的更多技能
浏览全部技能Build Jev agent harnesses with OpenJev-HF, LangChain, and LocalJev
Edit AI-assisted drafts without flattening the writer's voice, detect named AI-writing patterns without rewriting, or generate deliberately exaggerated AI copy for satire and before-and-after examples. Use when the user wants writing clearer, more direct, more opinionated, or less AI-sounding; asks whether a draft reads as AI; or explicitly requests intentionally bad AI-style copy.
Turn offline signs into researched leads
Save durable memory without secrets
Create or edit books with the Enoch Book Engine, including strict B5 pages, inline SVG diagrams, and mandatory screenshot-based visual QA.