agentsclimarketplace

Productized service business

Skill S3YED/appie-kit/skills/client/productized-service-business

Validate, launch, and deliver productized service businesses, service-first SaaS wedges, managed marketplaces, and audit/retainer offers. Use when the user asks which business idea is worth doing, wants competitor or Dutch/NL market validation, wants a monetization ladder, wants to validate a service before building software, or wants delivery SOPs/templates for a first paid client offer such as Review-to-Revenue / Customer Voice Conversion Audits.From its SKILL.md

Install
npx -y skills add S3YED/appie-kit --skill productized-service-business

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 6 stars6 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

7.9 KB, ~1.6k tokens by cl100k_base, as published. Nobody here has run it

Productized Service Business

Use this class-level skill to move from raw idea → market verdict → first paid offer → repeatable delivery. Prefer service-first validation over building SaaS until paid demand is proven.

Default workflow

  1. Define the paid proposition before researching

    • Who pays?
    • What painful job gets removed or improved?
    • What exact deliverable is produced?
    • What outcome is promised?
    • What is the narrow first wedge?
  2. Research competitors by category, not just exact names

    • Direct exact competitors.
    • Adjacent service providers.
    • Software/tools that solve part of the problem.
    • Agencies/consultants with broader overlap.
    • Blogs/content that describe the method but do not sell it.
    • Enterprise versions that are too heavy for the target customer.
  3. Compare overlap with a fixed grid For each found party, check whether it uses the same primary input, targets the same buyer, delivers the same artifact, promises the same outcome, has similar pricing/productization, and can expand into the wedge quickly.

  4. Score attackability 1-10 Judge pain intensity, buyer ability to pay, ease of first sale, competitor density, clarity of ROI, delivery simplicity, ability to productize into retainer/MRR, and copy risk from tools/agencies.

  5. Finish with the operator answer Always include: hard verdict, sharp positioning sentence, first paid offer, free entry wedge, upsell/retainer path, biggest risk, and next 3 actions.

Research lane

Load exa-plus for deep search when useful. If Exa fails because EXA_API_KEY is missing, continue with web_search + web_extract and mention the fallback. For Dutch checks, search both Dutch and English:

  • Nederland [problem] automatisering [buyer]
  • site:.nl "[exact phrase]" "[industry]"
  • "[method]" "webshop" "Nederland"
  • "[problem]" "AI" "conversie" "Nederland"
  • [global competitor category] Netherlands [industry]

Absence of exact keywords is not proof of uniqueness; adjacent categories count.

Service-first / managed-marketplace pattern

When the user asks for Airbnb/Uber-like ideas or simple people-to-people/service marketplaces, use a concierge MVP rather than recommending broad marketplaces.

  • Avoid generic “Werkspot but with AI,” “Temper but better,” or broad leadgen/offerte sites.
  • Look for narrow wedges where AI/automation improves intake, matching, trust, scheduling, pricing, payments, evidence photos, backup coverage, and follow-up.
  • Start with landing page + Tally/Typeform + WhatsApp Business + Google Sheets/Airtable + Mollie/Stripe/Tikkie.
  • Validate paid transactions manually before building software.
  • Prefer recurring/high-frequency demand: short-stay turnovers, VvE maintenance, landlord repairs, office services, event staffing, practical care relief, subscriptions/memberships.
  • Use references/service-marketplace-ai-ideas-nl.md for Dutch marketplace gaps, crowded categories, MVP stack, and current best bet.

Productized audit/retainer pattern

When a validated wedge becomes client delivery, switch from research to SOP execution:

  1. Intake only what is needed: URL, data source links/exports, priority pages or assets, business priority.
  2. Start with public data when safe; do not block on perfect access if a free quick-win or sample audit can be produced.
  3. Turn evidence into exact artifacts: page copy, FAQ answers, trust blocks, priority fixes, scorecards, checklists, or implementation-ready recommendations.
  4. Reject vague advice. Replace “improve trust” / “optimize copy” / “make clearer” with exact text, exact placement, evidence, and expected impact.
  5. Sell a ladder: free quick win → paid audit → done-for-you fix → managed retainer.

Review-to-Revenue / Customer Voice Conversion Audits

Use this subsection when the business is Harry's Review-to-Revenue wedge or any client task that turns reviews/customer feedback into conversion fixes.

Positioning

Do not position it as generic AI, CRO, review management, or a chatbot. Position as:

We turn customer reviews into concrete websitecopy, FAQ answers, trust blocks, and page fixes that reduce buyer doubt and improve conversion.

Best first niche: Dutch webshops with 100+ reviews, products above roughly €30, visible product/category pages, and repeated doubt around sizing, quality, delivery, installation, compatibility, returns, or trust.

Delivery ladder

  1. Free entry: 3 Review Revenue Leaks — public review scan, 3 concrete leaks, each with review evidence → page problem → exact fix text.
  2. Paid entry: Review-to-Revenue Audit — analyze 100-300 reviews and key pages; deliver top buying reasons, objections, complaints, return risks, 10 website fixes, 5 copy blocks, 5 FAQ answers, 3 trust snippets, and a priority plan.
  3. Upsell: Audit + Copy Fix — implementation-ready copy for homepage, PDP, category, FAQ, checkout, return/shipping trust copy.
  4. Retainer: Managed Customer Voice Agent — monthly review monitoring, new complaint/competitor patterns, and copy/FAQ update recommendations.

Review source order

  1. Client CSV/export from review platform.
  2. Public review pages, copied or semi-automated.
  3. Official API/feed if client provides access.
  4. Browser/scraping only if allowed and necessary.

Platform notes: Trustpilot public reviews/export; WebwinkelKeur API ratings/product-review endpoints with client credentials; Kiyoh XML/API/feed with client access; Shopify/Judge.me/Loox exports through client admin/app; Google Business official access is cleanest, public samples are enough for quick wins.

Labeling and page comparison

Tag meaningful reviews as buying reason, doubt/objection, complaint/friction, return risk, trust signal, missing product info, delivery/service, sizing/fit/compatibility, emotional trigger, competitor advantage, and FAQ candidate.

Compare at minimum homepage hero/USP, main category page, 3-5 product pages, FAQ/help, shipping/returns, and cart/checkout trust messaging if visible.

Bundled resources

  • references/review-to-revenue-netherlands.md — Dutch competitor-map and positioning notes.
  • references/review-to-revenue-delivery-kit.md — SOP notes, platform blockers, delivery standards.
  • templates/quick-wins-template.md — copyable 3 Review Revenue Leaks template.
  • templates/paid-audit-template.md — copyable paid audit report structure.
  • templates/client-intake-message.md — message to request client inputs.
  • scripts/review_csv_helper.py — local CSV preprocessing helper for review text.

Harry output style

For Harry, keep it caveman concise and direct: no academic report tone, no 20 vague options, use clear scores, say when a market is crowded, say when an idea only works with a narrower wedge, and prefer founder-led service validation over building SaaS first.

Common pitfalls

  • Do not confuse “nobody uses our name” with “no competition.” Category competitors matter.
  • Do not recommend broad automation bots as a first wedge unless the buyer pain is already narrow and paid.
  • Do not sell AI as the headline when the buyer cares about money, time, risk, or conversion.
  • Do not let a service become “we optimize everything for everyone.” Narrow the beachhead.
  • Do not overbuild. First sell an audit, pilot, or fixed-scope service; automate after proof.
  • Do not wait for perfect data exports. Public data is enough for quick wins and proof-of-work.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,736. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.