The growth flywheel

How product-led sharing, creator sharing, rev-share, and the impressions model compound into one loop. The premise is the central thesis: the more truth-seeking a company is to its customers' real experience, the faster it grows, so the whole flywheel runs on one context warehouse that every user, creator, and recipe writes back into.

The simple picture (the one to explain to marketing)

Amazon drew its growth on a napkin as one wheel that spins faster every turn. Ours is the same shape:

                 ╭─────────── the wheel spins ───────────╮
                 │                                        │
                 ▼                                        │
          ┌──────────────┐                     ┌──────────────┐
          │  FREE VALUE  │                      │  MORE DATA,  │
          │  in 5 min    │                      │  SHARPER     │
          └──────────────┘                      │  PRODUCT     │
                 │                               └──────────────┘
                 ▼                                        ▲
          ┌──────────────┐                     ┌──────────────┐
          │ USERS SHARE  │  ────────────────▶  │ NEW SIGNUPS  │
          │ "powered by  │                      │ land on free │
          │  Intempt"    │                      └──────────────┘
          └──────────────┘

   FUEL:  some free users pay  ──▶  revenue pays creators + referrers
          ──▶  creators bring more signups  ──▶  back to the top of the wheel

Read it out loud: a team gets real value free in five minutes, they share it (every shared page says "powered by Intempt"), the people they shared with sign up free, their use adds data that makes the product sharper, which makes the free value better, which makes them share more. Each turn is cheaper than the last. The money is the fuel, not the wheel: the paying customers fund the creators and referrers who spin it faster.

The four inputs, and how each feeds the loop

• PLG sharing (the base loop). A new user gets value free in under five minutes, then shares: invites teammates into the same workspace, sends their ?via link to peers, or shows the product to someone. Every user is a potential referrer; the Collective Share track pays a customer who refers peers the same 30% a creator earns. Free value in, sharing out, more free users back in.
• Creator sharing (the amplifier). Vetted creators put Intempt in front of audiences the product loop never reaches. They earn two ways at once: a flat fee for the placement (impressions) and 30% rev-share on the customers it brings. A per-creator code discounts their audience and credits them, so the placement both converts and attributes.
• Rev-share (the fuel pump). 30% of seats and AI Pass for 12 months turns every referred customer into recurring income for the referrer, which funds more creator activity and more sharing. It is the edge that makes the loop spin faster the longer it runs, not a one-time cost.
• Impressions (the reach the loop cannot buy organically). Flat and CPM sponsorships put the product in front of new audiences up front, before any rev-share has accrued. They prime the pump; rev-share keeps it running.

The two reinforcing engines

• The recipe engine. Users and creators publish recipes (named plays Blu runs). A published recipe is both organic discovery that pulls new users in and an in-product play existing users run, and it earns its author 20% of the credits it burns in other orgs. More users publish, more recipes exist, more acquisition and more royalty, so more creators publish. No competitor has this leg.
• The monetization engine. Free users convert to paid (seats, AI Pass, usage). That revenue is what the payouts are drawn from, so the rev-share and royalty that fuel the creator and sharing loops are self-funding, not a subsidy. The loop pays for itself once conversion and retention hold.

Why it compounds

Each turn makes the next turn cheaper. A paying customer funds the rev-share that funds a creator whose placement brings the next paying customer. A free user publishes a recipe that acquires the next free user. The context warehouse is the shared substrate: every user, every recipe, every creator-sourced signup writes behavior back into it, which makes the product work better for everyone and gives the next cohort a sharper reason to stay. Tiering and the ambassador bundle (see influencers) are the levers that make the highest-performing creators spin the loop harder over time.

Distribution is free and branded; creation is metered

The PLG-sharing loop only spins if the shared artifact carries Intempt to the recipient. The rule, the way Calendly, Loom, and Fathom grow with no referral incentive: gate creation, never distribution.

Every sharing surface should do three things: let the recipient view, book, or receive with no login; carry a "powered by Intempt" badge plus a signup CTA; and let the badge be removed only on a paid plan. The free shared page is the distribution.

Where the product stands today:

So the booking page already runs the Calendly loop; the two highest-frequency surfaces (a call recording sent to a prospect, a dashboard sent to a stakeholder) spread the content but not Intempt. Adding the badge and a signup CTA there is the cheapest acquisition win in this whole model.

The goal: shared booking pages carry a powered-by badge today; shared dashboards and recordings should too, each with a signup CTA (optionally carrying a ?via). The hide-badge option stays paid-only on all three, and recipients can view without logging in. Do not make the create side free: AI and token cost and monetization live there. Gate creation, never distribution.

The acquisition inputs, and how they land in the free plan

Three inputs pour new people into the free plan. Each carries Intempt to someone who has not seen it, and each lands on the same free plan where AI credits do the activating.

• Free tools (the marketing lead-magnet tools on /free-tools and /tools). A visitor runs a free tool, gets a real result, and the result page carries the same powered-by badge and signup CTA as any shared surface. The tool is the demo; the CTA is the door into the free plan. Same rule as distribution: the output is free and branded.
• Creator impressions (sponsored placements and per-creator codes). A creator puts Intempt in front of an audience the product loop never reaches. The impression lands as a free signup, often carrying an offer credit through the creator's code.
• Product sharing (the powered-by loop above). Every shared booking page, recording, and dashboard is an impression that lands the viewer in the free plan.

AI credits are the activation fuel, not the acquisition. The free plan gives a bounded daily AI grant (the free plan's AI allowance). That grant is what turns a signup into an activated user: enough free Blu runs, images, and analysis to reach real value in the first session, bounded so the token cost of free never runs away (and the heaviest AI is paid-only). Acquisition gets them to the free plan; the AI credits get them to value; value is what makes them share and convert. That is the hinge of the whole engine.

The CMO strategy

The strategy is to make the product its own largest acquisition channel, then pay creators to amplify it, and let owned content and recipes compound underneath.

• Positioning anchor. Every surface ladders to the thesis in who-we-build-for: a context warehouse plus five tools plus Blu agents is a self-driving GTM engine, and the more truth-seeking a team is to its customers, the faster it grows. We sell the engine, to the SMB growth team that cannot stitch and babysit eight tools.
• Channel order, cheapest-compounding first.
  1. PLG distribution loop (free tools, product sharing, powered-by badges). Lowest cost per signup, compounds, and most of it is already built. First priority is closing the badge gaps, not new spend.
  2. Creator and affiliate (rev-share plus the impressions lane, influencers). The amplifier that reaches past the product loop. Self-funding once rev-share is live.
  3. Recipes (the Build track royalty). A moat no competitor has: every published recipe is both organic discovery and an in-product play, and it pays its author.
  4. Owned content (the PostHog model: deep, first-party, no AI slop, no paid-for-links). Slowest to start, most durable, lowest long-run cost. A standing bet, not a quick win.
• Monetization path. Free plan, to self-serve PAYG (monthly, card or ACH), to a Savings Plan commitment for the teams that scale (order-form, invoiced). Price is never gated by feature; it is gated by usage and seats.
• The one number. Activated free workspaces per week, and self-serve new revenue behind it. Everything above is a lever on that number; report it with a source and a date, never a target pulled from the air.

Execution: phases, owners, cadence

• Phase 1, close the loops that are already half-built (product owns). Add the powered-by badge plus a signup CTA to the shared-meeting and shared-dashboard views, paid-gate the hide-badge toggle, confirm no-login viewing, and apply the same badge plus CTA to free-tool output. This is the cheapest acquisition work on the board and needs no new budget.
• Phase 2, stand up the creator program (marketing owns, payments supports). Pick the rails (the recommendation is Dub for links, codes, and payouts, own ledger for recipe royalty), vet the first creators, run the lifecycle in creator-operations, and open the flat sponsorship lane for a few high-value placements. Start with warm Collective applicants.
• Phase 3, compound (marketing owns). Ship the owned-content cadence, open the recipe publishing flow (the royalty leg), and layer tiering and the ambassador bundle onto the creators who perform.
• Owners, by lane. Product owns the sharing surfaces and the badge. Marketing owns content, creators, and free tools. Payments owns the payout rails and the credit model. The handoffs are explicit so no lane waits on another by accident.
• Cadence. A weekly read on the one number and its leading indicators (new free workspaces, activation rate, shares per active workspace, creator-attributed signups), each with a source and a pull date, because marketing's numbers live in dashboards this repo cannot read.

Related: who-we-build-for (the thesis and the five tools) · influencers (the firm creator model) · campaigns-and-coupons · creator-operations.