Outbound Call Sprints — the therapist campaign operating manual

How the therapist shortlist actually gets worked: time-boxed sprints, a warm-path-first rule, WhatsApp-before-phone, and explicit stop rules. Built for ~4–6 h/wk of sales time.

Parents: Client acquisition (ICP, ladder, qualification — not restated here) · Pricing 🔒 · Delivery. Inputs: who to approach · who can be reached warm. Everything is logged in PROSPECTS.md.

0. Standing constraints this must live inside

Non-negotiable, inherited from 01-client-acquisition.md and PRICING.md 🔒:

⚠️ Verify before the first build lands: HPCSA advertising rules may restrict patient testimonials and practitioner-quoted case studies — see the network map §4. This changes what can be built and what can be published as proof. Ask Alexa and Nadia.

1. The domain fact everything is designed around

Therapists sit in 50-minute sessions all day and cannot answer a phone. An unknown missed call reads to them as a patient in distress — intrusive, and an ethical loose end.

So the medium order is inverted from normal B2B outbound:

WhatsApp first → phone only once they've replied → email last.

At group practices, call the practice manager, not the clinician. They're non-clinical, at a desk, and paid to take exactly this call. (Therapy in Action has one; solos don't.)

Never leave clinical-sounding voicemail. Identify as a website supplier in the first sentence so nobody wonders whether it's an emergency.

2. The two-track rule

Cold outreach is a dropped channel — with one honest exception. A site that is measurably broken supplies its own introduction; there's no need to borrow trust.

The rule: never spend a cold touch on a prospect whose only problem is that their site looks dated. Dated-but-working is a taste argument, and taste arguments lose cold.

3. Path classes — assign one before any contact

Record the class in each PROSPECTS.md row. ⏸ HOLD is a real state — the discipline is in what doesn't get called.

4. Sprint 0 — the network harvest (do this first, ~3 h, zero cold contact)

The cheapest cold→warm conversion available. Two conversations with existing clients.

The mechanics that make a referral ask work

1. Named lists beat open questions. "Do you know Rianie, Liizl, or Annalize?" vastly outperforms "know anyone who needs a website?" — recognition is easy, recall is work.
2. Cap the list at ~5 names, chosen for her network (Alexa gets Pretoria + Winelands
   • southern-suburbs; Nadia gets ed-psychs).
3. Ask for permission to mention, not for an introduction. "Would you mind if I said you suggested I get in touch?" is a far smaller favour than "will you introduce me", and yields the same opening sentence.
4. Give before asking. Alexa's ask is already paired with the case-study metric request; Nadia's with her round-1 changes just shipped. Never make the referral ask the only reason for the call.
5. Colleagues only. ⚠️ Never ask a psychologist about patients, or for an intro to anyone she has a clinical relationship with. This is the one way to destroy both relationships at once.

Sprint 0 checklist

Exit condition: every shortlist row carries a path class, and at least 3 rows are 🤝 INTRO. If Sprint 0 yields zero intros, that's a finding, not a failure — it means the warm well is dry for this vertical and the 01-client-acquisition.md M1 pivot should be considered (pull Upwork/INTL forward) rather than grinding cold therapists.

5. Anatomy of a call sprint

One sprint = 2 weeks ≈ 10 h. One cohort of 8–10. One hypothesis. One stop rule.

Weekly shape (~5 h)

Cadence per prospect — 3 touches, then stop

Three touches is the whole cadence. A fourth touch on a health professional reads as pestering and costs more (in a referral-dense profession) than the marginal reply is worth.

6. The sprint schedule

Sprints run sequentially — one cohort hot at a time. Order is deliberate: cash-fastest and cold-legitimate first, highest-leverage second, template-reuse third.

Sprint 1 — 🔧 The Defect sprint (cohort: 3)

• Cohort: Lisa Dales (TLS dead), Kelly Jacobs (503, 9+ days), Edith Buchanan (no viewport, <font> tags, © 2026 → she cares).
• Hypothesis: a verified break converts cold, because the message is a service.
• Why first: needs no network, carries real urgency, and Buchanan gives the best before/after asset in the pipeline.
• Stop rule: 3 touches each, then park. Don't rewrite copy at n=3.

Sprint 2 — 🏢 The Hub sprint (cohort: 1 practice + 2 individuals)

• Cohort: Therapy in Action (~20 practitioners, 2 towns, has a practice manager), Natalie Sadie (the doorway — she's the best-connected individual found), Somerset West play-therapy cluster via whichever of the two lands.
• Hypothesis: one institutional relationship outperforms ten solo pitches.
• Requires: Sprint 0 item 3. Pairs with physioscpt — if that lands first, lead with it (allied-health → allied-health is transferable proof, not a vertical jump).
• Note: this deliberately overrides July's "deprioritise group practices" rule. That rule targets diffuse 2-partner practices; 20 practitioners with a dedicated practice manager is a concentrated buyer.

Sprint 3 — 🎓 The Ed-psych cluster sprint (cohort: 8–10)

• Cohort: the 5-province cluster — De Bruin, Miller, Green, Wessels, van Niekerk, Mojapelo, Steyn, Styles, Sadie (if not closed in Sprint 2).
• Hypothesis: one template + one before/after (nadiapreview) × ~10 near-identical practices is the cheapest repeatable pitch available.
• Requires: Nadia's intros (Sprint 0 item 2) — this sprint is materially better warm.
• ⚠️ Wessels' "no HTTPS" opener is dead (valid cert as of 2026-07-30). Open on his IIS/ASP.NET age. De Bruin's no-mobile-viewport is now the strongest visual before/after.
• Proof framing (standing): nadiapreview is "a working demonstration I built" — never a client site. berleinpsych.com is a real client site, name cleared.

Sprint 4 — 🌱 The Greenfield sprint (cohort: 8–10)

• Cohort: the no-website practitioners — Somerset West play-therapy five, Stellenbosch four, Bloemfontein five, Siphiwe (Mbombela, Facebook-only).
• Hypothesis: "own what you're renting" beats "replace what you have" — no incumbent site means no taste argument, which is what killed most WEAK rows.
• The pitch: they already pay ~R199/mo to Psychology Today for a profile they don't own, can't brand, and which ranks the directory above them.
• First: verify each actually has no site (directory-inferred, not proven).

7. Message templates

Register per BRAND-STRATEGY.md: quiet craftsman, short declaratives, outcomes over adjectives, zero stack talk, zero AI hype. One CTA. No prices in a first touch.

🔧 DEFECT — WhatsApp first touch

Hi [Name] — Daniel Slater here, I build websites for health practices in SA. I noticed something you'd probably want to know: your site currently [shows a security warning to visitors / has been offline for over a week]. I checked it twice to be sure.

Happy to tell you straight what's worth keeping and what it'd take to fix — free, about two minutes, no obligation. Want me to?

Rules: the defect is stated as fact, verified, and dated. No adjectives about their site. The CTA is the free look.

🤝 INTRO — WhatsApp first touch

Hi [Name] — Daniel Slater here. [Mutual] suggested I get in touch. I build websites for psychology and allied-health practices — [Alexa Berlein's practice / a working demonstration for an educational psychologist] is the kind of thing I do.

No pitch: if you send me your site link I'll tell you straight what's worth keeping. Free, two minutes.

🏢 HUB — practice manager, phone

Hi, is that [Name]? I'm Daniel Slater — I build websites for health practices, I'm not a patient. Do you have two minutes, or should I call back at a better time?

I work with psychology and allied-health practices in the Cape. With [N] practitioners across two sites, the thing that usually costs practices like yours is enquiries arriving by email and WhatsApp with no single place they're captured. Is that roughly your situation, or have you already solved it?

A question, not a claim. Lets them correct you — which is the conversation.

🌱 GREENFIELD — WhatsApp first touch

Hi [Name] — Daniel Slater, I build websites for SA therapy practices. I found you on Psychology Today rather than a site of your own, so you may already be paying a monthly fee for that listing.

The difference: that profile ranks the directory, and you can't shape it. Your own site ranks you. If it's useful I'll show you what that looks like for a practice your size — free, two minutes.

Voicemail (≤15 seconds)

Hi [Name], Daniel Slater — I build websites for health practices, this isn't a clinical call. I'll send you a WhatsApp so you can read it between sessions. No need to call back.

Referral ask (to Alexa / Nadia — never a first touch)

One more thing, and say no freely — I'm doing more work with psychology practices. Do you happen to know [Name 1], [Name 2] or [Name 3]? If you do, would you mind if I mentioned you suggested I get in touch? Only if it's comfortable — no obligation either way.

8. Metrics — and an honest warning about them

Log per sprint: touches → replies → free looks → Signal Audits → builds.

⚠️ The targets below are unvalidated guesses, not benchmarks. The only measured number in these docs is cold email's 0.215%. Treat sprint 1–2 as calibration, and replace these with real figures once ~20 touches exist.

The real target is one qualified build per sprint — the one-active-build cap means a second simultaneous close creates a waiting client, not more revenue.

Stop rules (the point of the whole document)

9. Weekly cadence

Single source of truth is PROSPECTS.md — update the row on every touch, per 01-client-acquisition.md §7. Cross-check active work against BOARD.md.

10. Open decisions (→ DECISIONS.md)

• Does the therapist campaign open at all before the site + honest contact channel are live? 01-client-acquisition.md §2 gates all outbound on it. Sprint 0 is gate-free (existing clients); Sprints 1–4 are not. Needs Daniel's call.
• ⭐ PsySSA 30th Congress, 13–15 Oct 2026 (JHB) — theme "Psychology in Transition: Humanity, Technology, and the Future of Care" with an AI-and-digital-practice sub-stream. Unusually well-aimed, and a talk there is credibility in front of the whole profession. But BNI was dropped at 75–100 h/yr — a congress must clear the same bar. ⏳ Volunteer applications close 31 July 2026; abstract deadline unconfirmed. Decide or consciously skip — the date is tomorrow.
• Group-practice rule amended? July said deprioritise; Sprint 2 argues a 20-practitioner practice with a practice manager is a concentrated buyer. Ratify the exception.
• Is samedicalwebsitedesign.co.za / SA Medical Specialists a competitor or a white-label partner (ICP-C)? Its client base is ICP-A. Resolve before pitching against it.
• HPCSA advertising limits on testimonials + published client metrics — affects delivery and proof assets. Currently 🔴 unverified.
• Sprint conversion targets (§8) are guesses; replace with measured figures after Sprint 2.