Decision
The staged TAM expansion (startup decisions -> work decisions -> life and work) becomes per-channel and audience-led, dropping the strict global sequence. TikTok runs all three altitudes from day one via the three-pillar structure ratified in the Aug 10 Michael x Tatem meeting; the feed’s job is growth, the funnel’s job is founder-sorting. A non-founder majority on TikTok is the expected success state, never a drift alarm. Converting surfaces (site, newsletter, LinkedIn, X) stay founder-weighted until yield data argues otherwise; the brand promise migrating toward “high-stakes calls in work and life” is the expected path, pending that data, with site copy frozen until then.
Context
The Aug 10 meeting ratified Tatem’s three simultaneous content pillars (broad personal story / mid audience-directed / niche frameworks) as the answer to decision content running thin at ~90 videos/month, with the mission “normalize making hard decisions and pivoting, then give people a framework for when and how.” After the meeting Michael pressure-tested the positioning with the Granola AI, brought that conversation in for objective evaluation, and explicitly instructed that outside advice be judged on merits rather than canon-match. The re-evaluation flipped the frame: on a growth-first mandate the algorithm picks the de facto niche, because the broad pillar will out-distribute the framework pillar and build a stage-3 audience regardless of what any positioning doc declares. Declaring “founder decisions” while the feed builds ambitious-people-in-transition is the actual incoherence risk.
Supporting reads that survived scrutiny: the funnel infrastructure (lead magnet -> scored survey router -> Pro / install call) is built precisely to sort a mixed audience, so broad top-of-funnel adds volume to lower bands while the router extracts founders (founder yield, not founder majority, is the requirement); the product’s differentiators per the research synthesis (identity as load, entangled ledgers, five currencies spanning life) make the broad frame more on-thesis than the narrow one; the James Clear comp was method-niche not audience-niche; and Michael already did the earn-authority-in-a-niche phase on other channels (60k list, 45k LinkedIn, $4M+ receipts), so TikTok is distribution for authority already owned. Two independent outside signals (Tatem’s pillar design, the context-free Granola AI) converged on the same shape.
Options considered
- Strict staged sequence (the 2026-08-09 doctrine as written): rejected for TikTok because it fights the channel’s physics; the broad pillar was already ratified in the meeting, so the niche label was the least true part of the plan.
- Full brand repositioning now (“Hard Calls helps ambitious people make high-stakes decisions in work and life” everywhere, the Granola AI’s recommendation): rejected as premature; it rewrites converting surfaces that monetize founders today for zero additional TikTok growth. Adopt after yield data, not before.
- New consumer-priced brain product: rejected; Pro at $197 already is the consumer rung, the Engine/journal is the consumer brain. No new SKU.
Consequences
- Old tripwire (“watch for non-founder over-index in save data as the stage-2 signal”) is retired: that outcome is now guaranteed and desired. New gate: founder yield through the funnel by week 6-8 of real distribution (per-pillar UTM signups, survey band distribution, install calls booked). Broad growth with zero founder yield after the window = re-weight toward the niche pillar; reversal is cheap.
- Broad-pillar guardrail (register, not audience): every story ends in a priced decision (what it cost, the tripwire, what the memo would say); educational never motivational; no woo woo. This is the structural version of Tatem’s tone rule.
- TikTok bio and pinned videos get written broad-with-receipts (ambitious people, founder credentials as the authority layer, “I want to be him” over name-dropping).
- Per-pillar UTM instrumentation from week 1, owner: Michael/Hermes when the bio link ships.
- Docs updated same day: tiktok-strategy-brief v1.3, content-dna.md (pillars, audience split, founder-yield gate, production register, 60k figure, Megaphone flag state), hard-calls-stack skill decision 19.
- Meeting ratifications recorded alongside: three pillars, low-production register, use the warmed-up account with the 250-view test gate, credibility rule, Tatem owns mid-pillar research.