Ask Avi - Growth & Go-To-Market Strategy

_Committed on 2026-09-11. This is the live growth strategy. If a decision changes it, edit this file - the internal operations spec (docs/internal-spec.md) states the binding requirements ("new-entrant supply wedge" + "consented, time-boxed claim"), and the code implements them (shared/claim-lead.ts, agents/marketing/index.ts, the portal claim banner). Behaviour must keep falling out of this document; the doc must not drift from behaviour.

1. The core insight (why this is the wedge)

Ask Avi is a two-sided marketplace. The classic trap is trying to solve both sides at once. The research this strategy banks on says the supply side has a cheap, overlooked entry point: new entrants.

• Provisional psychologists and new graduates have **empty capacity and

no way to fill it**: unpaid or piece-rate placements don't bring their own clients, and Medicare makes provisionally-registered clinicians nearly impossible to bill under (their placement is the income, not private clients). At the graduate level, the supply overhang is severe

• e.g. 158 applications for 48 South Australian _provisional

placements_ (PsychVault / AAPi 2026 reporting).

• Newly registered solo therapists take **6-18 months to fill a

caseload** and consistently list marketing/lead-gen as their #1 unsolved problem (Growtherapy & solo-practice survey coverage).

• In the same period they're desperate for clients, they are the segment

that LANDS on Ask Avi: they already use groups, forums, and TikTok to chase referrals.

The wedge: recruit the empty-capacity new-entrant supply (students first, new grads second, established practices curated last). When a seeker can't be matched in-network, the demand-backed "claim this lead" offer dips into that supply pool with ONE real client on a warm, time-boxed, free claim - and converts a stranger into a registered, proven provider. The two-sided cold start becomes one-sided: supply is cheap, comes to us, and pays us once a client lands.

2. The claim offer - the demand-backed onboarding loop

When match-provider finds no in-network fit for a non-crisis lead (open_demand row), marketing reverse-matches that lead to a not-yet-registered prospect and emails them a claim link instead of a generic "join our network someday" pitch.

Rules (binding; see internal-spec "consented, time-boxed claim"):

• No seeker PII in the email. The copy carries only type/region/

urgency/keywords + a private claim link; name & contact unlock on acceptance, and only to the exact inbox the offer went to.

• Time-boxed and free. 24h by default; the first claim never charges

(charge columns stay null). New-entrant proof of value.

• Never double-offered. A lead is claimed at most once, whether sold

(sell-demand-lead) or claimed free - same claimed_by_provider_account_id=is.null compare-and-swap, so the two paths can never race to two owners.

• One seeker is never emailed twice for the same lead: the first claim

minted in a batch exhausts that lead for the rest of the pass.

• Accepting drops the client straight into the provider's existing

"Accepted clients" queue (a real referrals row, already unlocked) and notifies both parties exactly like a matched referral.

3. Self-serve onboarding is table stakes

Every competitor with any traction in adjacent space leads with "no sales call, get set up in minutes" (Caresensy, ClinicSync, Phensys, Apex Bookings et al). Ask Avi's bar is already email-only + one click (the register-provider magic-link flow). This strategy holds that bar: nothing in the next milestone may put a paywall, a form, or a call between a new-entrant provider and their first registered profile or first claim. The dashboard carries that pitch and the portal login is the same route as registration.

4. Pricing posture (anchors, not commitments)

Anchored on what the market actually charges for demand in this niche:

• Sold leads (existing members, via sell-demand-lead): Ask Avi

defaults to $15/lead (DEFAULT_LEAD_FEE_CENTS=1500). The defensible band for a qualified, matched, consented lead is $35-60. Current default undercuts that - it's an early-launch subsidy, not the ceiling.

• Comparator anchors (2026, AU primary):
  • Google LSA (local services): $20-75 per lead in mental health.
  • Non-branded mental-health pay-per-click: $110-155 cost per lead.
  • PeptideLeads-style qualified B2B/B2C lead products: ~$150 / lead.
• New-entrant wedge: first claim free, every subsequent claim/cost

path follows registered-provider pricing. Volume and tier pricing are documented elsewhere in the spec.

The guiding shape: free first lead; cheap matched leads; **discount the acquisition, never the compliance moat.**

5. Compliance is the moat - the "no cold blasts" rule

The investor brief pitches Ask Avi as the referral service that can be trusted with client data because of how it's built (Privacy Act posture, consent-first flows). A lead-offer email preserves that moat; an anonymous lead dump to every practice in a suburb burns it. Therefore:

• Every outreach/claim send goes through send-compliant-email (the one

send path, suppression + consent-basis + crisis gates).

• Claim offers are only ever about a real seeker who actually asked for

help (service_request consent basis), carry sender identity and a working one-click unsubscribe (Spam Act 2003).

• Copy never fabricates demand evidence: no invented counts, no fake

urgency. What the email says is exactly what open_demand recorded.

6. 90-day plan

1. Ship the claim flow (this change set):

shared/claim-lead.ts issue/redeem/lookup; open_demand claim columns; claim-offer email template; runOutreachBatch routes non-members through claims; portal ?claim= cookie parking + accept endpoint + "client on hold" banner; tests green.

1. Keep register-provider self-serve. No sales call, one-field

signup, first claim card-free. (Bar is already met - defend it in reviews, don't regress it.)

1. Wedge content: publish student/grad supply content on the TikTok

channel (scripts in campaign/new-entrant-supply.md) aimed at provisional/graduate clinicians: "your clients aren't coming - here's how one shows up this week." Drive to the portal; demand-backed claim emails do the rest.

1. Measure and tune the two numbers that matter:
   • Claim-click → redemption rate (the offer is the funnel).
   • Redemption → booked-appointment rate (the claim is the promise).

Tune expiry (24h), copy, and which domains/types get claims first.

7. KPIs for this strategy

• Supply conversion: % of claim offers redeemed.
• Supply velocity: days from first contact to first booked client for

new-entrant providers.

• Demand non-loss: % of open demand that resolves to a claim or sale

within 24h.

• Compliance intact: 0 sends outside send-compliant-email (invariant,

tested).

8. Sources

• Provisional-psychologist placement scarcity: PsychVault & AAPi 2026

coverage of 158 applications / 48 SA provisional placements.

• Solo/new-grad caseload ramp: Growtherapy 2026 industry reporting -

new solo providers and small practices take 6-18 months to fill a caseload and name client acquisition as the top unsolved problem.

• Self-serve onboarding bar: Apex Bookings, Caresensy, ClinicSync,

Phensys product pages (2026) - "no sales call / set up in minutes" is the shared competitive baseline.

• Pacing anchors: Google Local Services lead-cost bands for mental

health ($20-75); non-branded mental-health CPC CPL ($110-155); qualified-lead products such as PeptideLeads (≈$150/lead).

• Medicare/provisional-funding constraint: current Medicare Better Access

arrangements do not rebate sessions with provisionally-registered psychologists.

9. What this doesn't change

• sell-demand-lead (paid path, at-most-once) - unchanged.
• bill-on-booking (bill on booked appointment) - unchanged.
• Crisis demand is never captured, offered, or sold - unchanged.
• Seeker-facing spec (openspec/specs/agentic-system/spec.md) - not

touched; this is internal ops + marketing mandate only.