danielslater.dev — Operations Playbooks
The repeatable how-to for running the micro-agency day to day. These operationalize the settled strategy — they don't restate it. Sources of truth: Business model · Decisions — settled · Brand strategy. The model docs propose; DECISIONS.md disposes.
Standing rules that bind every playbook: no hospitality clients, ever (employer separation; the restaurant app is anonymised proof only, never a referral seed) · one active build + a waitlist (never assume parallel work) · client-facing language is outcome-led — no tech-stack talk, no AI hype (these internal docs may name the stack) · prices ZAR-first with USD/GBP alongside, planning math at the R16.0/$ buffer.
The five playbooks
| Playbook | Operating question it answers | In one line |
|---|---|---|
| 01 · Client acquisition | Who do we want, and how do we bring them in? | ICP-A/B/C fit, hard disqualifiers, the referral→white-label→Upwork channel sequence, the free-look→Signal-Audit→build funnel, qualification go/no-go, intake, on-spec previews, pipeline tracking. |
| 02 · Delivery recipe | How do we actually build and ship? | One recipe, five separately-priced stages (Discovery → Faithful Repaint ★ → Options → Rebrand → Full redesign), the "clone verbatim first" rule, handover/launch checklist, decision tree for which stages to propose, concurrency + scope discipline, internal tech recipe. |
| 03 · Pricing, payments & discounts | How do we assemble a quote, discount it, and get paid? | Quote = base package + delivery stages + recurring line + add-ons; change-order discipline; one founding-client slot/package traded for a named case study; SA vs INTL payment rails + terms blocks; targets to steer by; agency wholesale; SOW skeleton. |
| 04 · Recurring & care | What's the monthly subscription, what does it cost us, how much time? | The recurring menu (Care Plan Basic/Std, WhatsApp Receptionist R1,800/mo, web-chat, deferred Voice), per-plan cost + margin, the two Meta clocks, ≤2 h/mo/client time ceiling, substrate-before-selling gates, billing/onboarding ops, destination hygiene (POPIA). |
| 05 · Services & integrations | What do we attach onto a build, and how do we scope it? | The add-on catalog (CMS, analytics, transactional email, payments, WhatsApp build, deep integrations, advertising), the one-integration-included rule, per-target integration effort (Calendar / GoodX / Healthbridge), and the global "SA add-on prices are unvalidated proxies" flag. |
How these fit together
A single engagement flows straight down the numbered docs:
Acquire a lead (01: referral / white-label, qualified against ICP + disqualifiers) → qualify & audit (01 go/no-go → the paid Signal Audit writes scope down first) → deliver via the scope ladder (02: faithful Repaint by default, optional stages stacked on top, one build at a time) → price & get paid (03: assemble the quote off written scope, 50% deposit, correct terms block) → hand over to a recurring plan (04: Care Plan + WhatsApp Receptionist attach at launch, best-effort, month-to-month).
Add-ons (05) layer onto any build — content editing, analytics, email, payments, extra integrations — always scoped into the current build or a written change-order, never a reason to open a second build.
Open decisions to ratify (→ DECISIONS.md)
Every item below is flagged Proposed or open inside a playbook — a genuinely new operational choice, not yet settled. Daniel ratifies (or amends) into DECISIONS.md; until then they are recommendations only.
Acquisition (01)
- Free automated website audit (Kea-Logic-style URL grader — load time / clarity / conversion friction) as a new top-of-funnel rung that converts into the paid Signal Audit. (Also raised in 02 as it changes the front door of the delivery ladder.)
- First-reply response habit = 1 business day (BRAND-STRATEGY wants a stated habit but sets no number).
- Hospitality-decline exact wording (the fact of declining is settled; the copy is new).
- On-spec preview as a standing repeatable play + its go/no-go thresholds (currently a mozguide one-off precedent).
Delivery (02)
- Stage 2 "Options round" as a discrete priced add-on — light palette/type pick folded into the Repaint's one revision; deeper multi-concept exploration a paid add-on (Proposed ~R4,500 / ~$350, or the R900/hr shadow rate).
- Per-stage incremental hour targets for Stages 2–4 (+4–8 / +8–16 / +12–20 h) — estimates to validate on clients 1–3, not measured.
Pricing & payments (03)
- Founding-client discount band — 25–30% default (up to ~40% for a marquee arm's-length INTL reference), one slot per package, released only against a written case-study + metrics-rights trade.
- No ad-hoc discounting rule — discounts buy a named reference, never close a hesitant signature; drop to a smaller package instead of cutting price.
- Shadow-rate sanity floor — reject or rescope any quote implying below ~R700/hr effective.
- Change-order minimum billing — 1-hour minimum, then 30-minute increments.
- Default deposit-rail ordering — Paystack primary (Yoco in-person) SA; Wise primary INTL (rails are grounded; the default ordering is new).
Recurring & care (04)
- Care Plan tier scope split — Basic = ~2 small content changes/mo; Standard = reasonable-use batched monthly + a monthly health check (docs set the prices, not the per-tier scope).
- Best-effort response window — publish "async, ~2 business days, no guaranteed uptime SLA."
- Interim overage handling (before Q3 metering) — manual monthly dashboard read + ~80%-of-cap heads-up + hand-billed overage.
- Cancellation notice period — month-to-month with 30 days' notice (docs settle "no lock-in" but no notice period).
Add-ons & integrations (05) — all setup prices below are unvalidated web-dev-hourly proxies pending a dedicated SA add-on pricing research leg (open item in DECISIONS.md):
- CMS / content-editing add-on — fixed build add-on (text/images/prices/gallery boundary), proxy R3,500–6,000 setup.
- Analytics add-on — GA4 + conversion-event tracking as a paid add-on, proxy R1,800–3,500 (CF Web Analytics free/default is already settled).
- Transactional email add-on — pricing the Resend setup as an add-on, proxy R2,500–4,500 (Resend as vendor is already settled).
- Standalone payment add-on — proxy R3,500–6,500 (the Storefront bundle is grounded).
- Advertising offering — whether to build/sell it at all, its price, and its management-retainer cap; needs a dedicated pricing/scoping research leg first.
- GoodX API partnership for the dental segment vs staying calendar-sync-only + own-the-diary.