Delivery Plans — how J&M builds each fix

One playbook per common fix the audit surfaces. Each is written to be run inside the AI Concierge model — done with the client, on a screen-share, following AOA (Audit → Optimize → Automate): fix the process before you automate it. Never automate a broken process (AOA Concierge / AOA Framework).

The rule that makes the retainer renew: every session ships something and logs it in the Notion hub with the quantified result — recapped as “what WE built today,” not “what I delivered.” The renewal mechanism is the client saying the result out loud on the call (“that saves me ~7 hours a week”) and seeing it logged in their own words in Notion (AOA Notion Documentation Hub).

The delivery arc (every engagement)

1. Onboard — intake form (hard requirement, done before call 1), Cowork/Claude setup if new (AAIC 4.1; AOA Strategy Call Engine).
2. Per workflow, run AOA — Audit the current process → Optimize (cut steps, question each one) → Automate the clean version.
3. Ship + log — build at least one thing per call; record it, the takeaways, action items (two-sided), and the quantified impact in Notion — and end every session by booking the next one on-screen and naming what we’ll ship in it. Never leave a session without the next session: the retainer is a chain of next steps.
4. Weekly async — Voxer/Slack, 12-business-hour SLA.
5. Prove + renew + refer — quantified results → testimonial → next deal (AAIC 4.6). The referral ask is a designed moment: right after a quantified win is logged, ask “who else do you know running the same leak?” Testimonials must name trade + town + number (“plumber, Pocatello, 7 hrs/wk back”) — similar-others proof, not generic praise.

Fix playbooks

Front office (capture + convert) — fix these first: - missed-call-text-back.md — the highest-ROI local fix. - speed-to-lead-followup.md — instant personalized response to every inbound. - ai-receptionist-voice-agent.md — Retell.ai voice agent taking calls/booking. - quote-proposal-skill.md — Claude skill: notes → priced quote in 2 min. - review-reputation-engine.md — automated review requests → local SEO. - knowledge-base-second-brain.md — company knowledge system / Custom GPT. - crm-and-pipeline.md — lightweight CRM + pipeline automation. - scheduling-dispatch.md — booking + team dispatch automation.

Marketing & growth (generate demand) — turn on after capture works: - answer-engine-visibility.md — get cited when buyers ask AI (AEO/GEO). Organic/earned. - paid-social-ads-and-creative.md — Meta ads: research, creative, campaign, reporting. Paid.

Each playbook: what it fixes · the stack (with alternatives) · AOA build steps · time to ship · compliance · what to log.

The two marketing plays, compared

Jake flagged two tool launches as strong fits (sources: ../sources/reference/marketing-tool-launches.md). They are not competitors — they sit at opposite ends of the funnel, and a client can run both.

The through-line: both replace an expensive agency function with agent/skill-driven work — which is J&M’s entire thesis (see the gooseworks “$15k/mo agency → <$100/mo skills” framing; it’s your offer ladder’s logic, shipped publicly). More tools like these launching is a tailwind, not a threat (../00-thesis.md): the toolbox is not the moat — anyone can buy the tools. The moat is being the expert who’s already done the hunting, learning, and wiring, wrapping it in the audit, ROI framing, setup/tuning, and accountability, and running it all behind the scenes so the client never has to become a tool operator. Sequencing: fix front-office capture first, then visibility, then paid — never pour demand into a leaky funnel.

Tiers & pricing when a client wants it built for them (done-for-you upsell): Process Redesign $3–5k · Automation Build $1–3k · Knowledge Systems $3k+ · Custom Workflows $3–5k · Full Implementation $5–10k+, plus a $497–1,497/mo care retainer (AOA Upsell Menu). Don’t offer Full Implementation until you’ve shipped 5–10 lower-tier builds.