Business Plan

How to read this plan

This is the strategic business case that grounds Maxxim's marketing and brand work in commercial reality. It frames how the business makes money, where it grows, what it must prove, and what it risks. It is written for the founding-cohort window (July to September 2026) and the 3 July partner session that opens it, against Master Briefing v2.1, the Gate 1 direction of 27 June, and the newly-known product mechanics from the 29 June product guide. It now rests on a concrete, in-market embodiment: Jewell is a live partner running on Maxxim, the first real proof the channel works.

One discipline runs through every page. The structure of the model is set: three entities, two open doors, a floor-plus-clip revenue model, a land-and-expand funnel, and a partner community as the human moat. The numbers inside that structure are not set. Every percentage, split, floor, clip, cap, and finder's-fee rate is marked [illustrative, pending pilots] and must be locked through the founding-cohort runs before it is stated as fixed or used in public copy. Where Master Briefing v2.1 forbids inventing an answer, the item is carried as [OPEN] rather than resolved here.

Pricing anchors that are real are stated as real: the Growth Diagnostic at $4,950, the premium engagement band of $25,000 to $250,000 with a recent run near $80,000, and the specialist module near $1,500. Ownership is Clent Jewell and Raef Akehurst only, two owners, no board. Paul Clancy is a Client Partner, not a co-founder. Proof clearance is honoured throughout: Pottsville Acupuncture is the only public example, OTR Earthmover is internal only, Hidrive is percentages only, and Jewell's own named results stay under Jewell's banner.

1. Business Model Summary

The shape of the business: three entities

Maxxim is not one company. It is three entities with distinct roles, deliberately separated so that the method, the platform, and the channel each carry their own economics and none can quietly subsidise another. This separation is the integrity of the model, and it is what makes the partner economics testable rather than asserted.

      THREE-ENTITY STRUCTURE

      +-----------------------------------------------+
      |  JEWELL  (method owner)                       |
      |  Owns the 3D Process, standards, quality bar  |
      |  Clent = architect + gate-holder              |
      |  Owns the human 20% (trust, taste, judgement) |
      |  ALSO the live, exemplar partner in market    |
      +-----------------------------------------------+
              |  licenses method + IP             ^
              |  to ...                           |  IP licence fee
              v                                   |  [illustrative]
      +-----------------------------------------------+
      |  MAXXIM  (the platform)                       |
      |  The machine 80%: Max, Sal, Pip at speed      |
      |  Owned by Clent + Raef (2 owners, no board)   |
      |  Charges: platform floor + per-engagement clip|
      +-----------------------------------------------+
              ^  floor + clip                    |
              |  [illustrative]                  |  runs engagements
              |                                  v  on the platform
      +-----------------------------------------------+
      |  PARTNERS  (channel + delivery)               |
      |  Own NEITHER method NOR platform              |
      |  Own the CLIENT RELATIONSHIP + the margin     |
      |  Sign off the 20% that carries their name     |
      +-----------------------------------------------+
              |  delivers to ...
              v
      +-----------------------------------------------+
      |  END CLIENTS                                  |
      |  Any org type, commercialised through partner |
      |  Receive strategy + launched, OWNED assets    |
      +-----------------------------------------------+

The integrity point, now demonstrated. Jewell, the owner's own delivery arm, runs on the same rails and the same floor-plus-clip as any licensed partner, even if it is netted internally. If the owner's delivery does not run on the standard rails, the economics are never really tested. Jewell is therefore both the proof that the method works when a human runs it, and the first live pilot of the partner model itself. The Jewell partner site articulates the seam cleanly: Jewell owns the human 20 percent, Maxxim is the machine 80 percent, one named partner owns each client relationship. Pottsville Acupuncture is the cleared public evidence of the delivery.

The product, as it actually runs

The engagement is carried by three named AI specialists, each one a process and each one human-checked. This is the machine 80 percent in concrete form.

Work is gated and mechanical. Creative unlocks only as the briefs it depends on exist: the logo needs the Logo Brief, brand imagery needs the Brand Guidelines, photo and video need the Photo and Video Brief, collateral needs Brand Guidelines plus a locked logo, campaign pieces need the Design Brief plus a locked logo, and the Social Studio unlocks only once the Social Strategy exists. The gates are not a limitation, they are the moat. They are also why AI proposes and the human decides: the human holds trust, taste and judgement exactly where quality is decided, which is the explicit bet that good, fast and cheap need not be a trade-off.

Build only what we must: the operating philosophy

The single most important cost and moat decision in the model is what Maxxim does not build. Only the thin layer, Max plus the Maxxim engine, is built in-house. Everything else is deliberately leaned on and swapped for the best tool for each job. This is the Jensen Huang discipline, do not build what others build better, and it is what keeps cost-to-serve low enough for the partner economics to work.

The commercial consequence is direct. Because the stack is swappable and mostly rented, the fixed build burden is small and the variable cost per engagement is the number that decides the model. That is the discipline the whole plan turns on: measure the variable cost per run, do not assume it.

Value Proposition

Maxxim has two value propositions because it serves two customers, and each is the proof of the other.

• To the partner (primary customer): license a proven, gated method and run agency-grade growth at scale under your own name. Keep the client margin, stand behind every result, and move from founder-capped hours to repeatable delivery, without building a method, a team, or an AI toolchain from scratch. You own the human 20 percent; the machine carries the 80 percent.
• To the end client (served through the partner): a complete strategy plus launched assets you own outright, on your own domain, delivered in days not months, with a named human accountable for quality at every gate. The trustworthy middle between slow, opaque legacy agencies and risky, effort-heavy pure-AI tools.

Customer Segments

• Client Partners (license door): trusted operators who license the method and own their client relationships. Named early channel: Alex / 3P (around 36 opportunities), Arran Aitken (tradie database). Jewell is the flagship live partner.
• Referrers (refer door): networked operators who send clients in for an optional finder's fee, with no delivery obligation. Named: Greg Johnson; Christy Kilmartin (refer or specialist, two hats kept separate).
• Specialist Partners: scoped experts (brand, campaign) who deliver a module alongside Sal or Pip and never own the client.
• End clients (the served segment): universal across organisation types per the Gate 1 decision: personal brand, solo operator, SME, corporate, not for profit, and all levels of government. The constant is the method plus AI speed plus an accountable human; the variable is context, tier, and compliance.

Channels

Today the front door is the partner, with Jewell as the first and exemplar partner, live in market. Maxxim reaches end clients through operators who run the 3D Process for their own clients. Clients meet Max, Sal and Pip only through the partner, with an optional read-only portal (Files, a private business assistant called Ari, and Help) for those given one. The north star, sequenced after the partner motion is proven, is an end-user self-serve portal that takes a client through Discovery and outputs their own 3D deliverables, an evolution of the read-only portal from surface into engine. Visibility to end clients is therefore partner-mediated now and direct via the portal later. The maxxim.ai site is the proof surface; the 3 July session is the next deliberate channel to the founding cohort.

Revenue Streams

Five lines, structure set, every economic figure illustrative until pilots.

The recurring spine over time is the Deepen line: the compounding improvement loop after launch, modelled on the Pottsville pattern. It is the piece most agencies never systematise and most tools cannot offer. A gated engagement that lands, then continues as a recurring Deepen relationship, is the shape that carries the highest lifetime value in the whole model.

Cost Structure

The defining cost insight follows directly from the build-only-what-we-must philosophy. Because the stack is mostly rented and swappable, the fixed burden is small and the largest variable is AI and compute per engagement, not a flat monthly line. Contribution margin per engagement is therefore the number that decides whether the model works. Understate it and the whole plan is fiction.

2. SWOT Analysis

Strengths

• A proven, gated method, enforced by the product, not just tooling. The 3D Process has been run with real clients for years through Jewell, with a launched, client-owned site (Pottsville) as cleared public proof, and the platform now enforces the gates mechanically. Implication: lead the partner pitch with the method's track record and its enforced gates, because it is the one asset no enablement platform or self-assembled toolchain can claim.
• The 80/20 architecture fits the moment exactly. AI made the first 80 percent cheap; value moved to the accountable last 20 percent. Implication: Maxxim's structure is aligned with where margin is migrating, so the positioning rides the market rather than fighting it.
• A live, in-market exemplar partner removes the credibility gap. Jewell now runs on the same rails it sells, presents Maxxim as its engine, and articulates the seam publicly. Implication: the founding cohort can be shown a working example rather than a pitch, which directly attacks the central trust objection.
• Low, swappable cost base by design. Only the thin layer is built in-house; the rest is rented best-tool-for-each-job (Cowork, Cloudflare, Recraft, Veo, GPT-image, Buffer). Implication: cost-to-serve stays low, which is what makes a modest floor and clip plausible, and the stack improves for free as the underlying models improve.
• Lean ownership and a fast compounding loop. Two owners, no board, with a tight loop: run, capture friction, decide, reflect in product, retest. Implication: Maxxim can move faster than committee-bound competitors and lock pilot economics quickly.
• Recurring Deepen line plus a community moat in design. The engagement does not end at launch, and the partner community compounds knowledge and deal flow. Implication: the business has both a retention mechanism (Deepen) and an again-and-again mechanism (the community), which most agencies and tools lack entirely.

Weaknesses

• One cleared public case study. Pottsville is the only public proof under the Maxxim banner; OTR, Hidrive, BTC and others are clearance-restricted, and Jewell's named results sit under Jewell's banner. Constraint: the Maxxim-brand proof base is thin in public while every category competitor has more market history, so sourcing additional cleared, named results is the single highest-value action.
• Partner economics are unproven. Floor, clip, cap, and finder's fee are all illustrative until pilots. Risk: the founding cohort is being recruited before the price of being a partner is known, so the pitch must sell the model and the community, not a settled rate card.
• AI/compute is a variable cost not yet measured at scale. Contribution margin per engagement is the swing number and is not yet locked. Risk: a high-depth premium run could erode margin faster than expected; the cap on big upfronts exists precisely to contain this and must be validated against a real run.
• Model quality depends on partner quality. The 20 percent that carries the name is the partner's, so a weak partner can damage the standard the whole model rests on. Constraint: the light quality gate and the community's shared quality bar are load-bearing, not optional.
• Dependence on leaned-on third parties. The low cost base comes from renting the stack, which means exposure to third-party pricing, availability, and terms. Constraint: keep the stack swappable in practice, not just in principle, so no single vendor can hold the economics hostage.
• Several public-facing decisions are still OPEN. The category lead word, the front-door product confirmation, end-client visibility, the fractional CMO model, and the showcase promise are not settled. Constraint: public copy and the session narrative must work around these rather than pre-empt them.

Opportunities

• The empty quadrant is unoccupied. No competitor category combines AI speed, a proven gated method, and an accountable human delivering finished, owned outcomes. Action: plant the flag in that quadrant through the session and the first cleared case studies before a faster incumbent borrows the language.
• Universal audience widens the partner aperture. The Gate 1 decision opens the method to every organisation type, including not for profit and government. Action: recruit partners by vertical strength (a trades partner, a not-for-profit partner, a government-savvy partner) so the network covers ground no single operator could.
• AI fatigue is a tailwind. The market is adopting AI rapidly while tiring of unaccountable output. Action: make human-at-every-gate the explicit answer to the trust gap in all messaging.
• The self-serve portal is a second front door. Once the partner motion is proven, a product-led portal opens a second, lower-cost acquisition path from the existing read-only portal. Action: sequence it deliberately after pilot economics are locked, not before.
• The community itself is a recruiting asset. Partners join a table, not a vendor list. Action: run the 3 July session as the founding meeting of that community so belonging is felt from day one.

Threats

• Agencies adopting AI to move faster and cut price. Defence: compete on the proven method and the enforced gates, not on being the cheapest or merely the fastest; speed alone is copyable, a gated method plus cleared proof is not.
• Enablement platforms moving upmarket into method. Defence: protect the gates and the quality bar as the differentiator, and lock partner loyalty through the community moat before well-funded platforms reach for the same operators.
• Commoditisation of AI-generated sites. Defence: sell finished, owned outcomes and the relationship, not the artefact; the website is one output path, not the product.
• Founding cohort fails to convert or to prove economics. Defence: use the $4,950 Growth Diagnostic as a clean, repeatable first sale so pilot data accrues fast, and keep cohort size small enough to support well.
• Third-party stack risk. A leaned-on tool changes price, terms, or availability. Defence: the swappable, best-tool-for-each-job design means any single tool can be replaced; keep at least a mental substitute ready for each job.
• Key-person concentration. The method, the gate, and final sign-off all run through Clent. Defence: the whole purpose of Maxxim is to lift delivery off Clent's hours; codify standards into the platform and the community quality bar so the gate is a system, not a person.

3. Growth Strategy

Across the Ansoff levers, the right primary lever for Maxxim at this stage is Market Development through a partner channel, with a deliberate, sequenced move into Product Development (the self-serve portal) once the channel is proven. Market penetration and diversification are not the play right now.

Recommended lever: Market Development via partners

Rationale. The proven asset already exists, the 3D method delivered through Jewell, and it now runs live in a partner's hands. The constraint is not the product, it is reach: Jewell's direct delivery is capped by Clent's capacity. The fastest, lowest-risk way to grow is therefore to take an existing, proven offering into new markets through new operators, rather than inventing new offerings (Product Development risk) or entering wholly new markets with new offerings (Diversification risk). Partners are the multiplier that turns one capped delivery arm into many, and the leaned-on stack means each new partner adds little fixed cost. This lever also matches the business stage: pre-scale, proof-hungry, and dependent on real run data to set economics.

Specific opportunities under this lever

• Founding cohort of 5 to 8 partners, July to September 2026. A deliberately small, supportable group that proves the method works in others' hands and generates the pilot data to lock floor, clip, and cap. The named pipeline (Alex / 3P, Arran, Greg, Christy) seeds it, with Jewell already live as the reference shape.
• The live Jewell partner as the pattern. The Jewell flow (client enters through the human, Discover is human, Design and Deploy hand off to Maxxim at scale, one named partner owns the relationship) is the concrete reference every new partner is onboarded against.
• Vertical pods after the cohort. Recruit partners by vertical strength so the network spans trades, local business, not for profit, and government. Pods let partners refer across strengths, creating work no single operator would win alone. The first wedge is trades and local SMEs, where a partner already commands the audience (Arran's tradie database) and Pottsville's allied-health pattern is the proven shape.
• The self-serve portal as a second, product-led front door. Sequenced after the partner motion is proven, this is the Product Development move: the same method, productised for direct end-user delivery, opening a lower-cost acquisition path.

Prerequisites for this growth path to work

• The $4,950 Growth Diagnostic adopted as the universal, repeatable front-door sale, so every partner has one clean first transaction.
• Pilot economics locked from real runs: contribution margin per engagement, cost-to-serve per partner, client LTV, break-even partner count.
• At least two additional cleared case studies beyond Pottsville, to close the proof gap that gates partner confidence.
• The community rhythm operating from the founding meeting onward, so partners stay for the people and the shared deal flow, not only the tool.
• A quality gate light enough to scale but firm enough to protect the standard the model depends on.
• The stack kept genuinely swappable, so cost-to-serve stays low as partner volume grows.

4. Revenue Model Analysis

The two open doors

Every partner relationship runs through one of two doors. This is the locked structure of how Maxxim makes money from the channel.

      THE TWO-DOOR PARTNER MODEL

      An operator arrives. They choose a door.

      DOOR 1: LICENSE TO DELIVER          DOOR 2: REFER FOR A FEE
      -------------------------           -----------------------
      - Owns + runs the client            - Sends a client in
      - Keeps the client margin           - No delivery obligation
      - Delivers via Maxxim (Max,         - No method access
        Sal, Pip) + leaned-on stack       - Optional one-off
      - Signs off the 20% (their name)      finder's fee on
      - Pays: PLATFORM FLOOR                conversion, set vs
              + PER-ENGAGEMENT CLIP         GROSS MARGIN
        (clip scales by lead source)       [rate illustrative]
        [floor, clip illustrative]

      Plus a scoped lane:
      SPECIALIST MODULE -> paid a module fee (~$1,500),
      never owns the client; Maxxim clips packaging
      margin ONLY when Maxxim sources the work.

Money flows by lead source

Who pays whom depends on who sourced the client. This is the mechanism behind the clip scaling, and it is exactly the four-flow model in the master briefing.

      MONEY FLOWS BY LEAD SOURCE

      (1) PARTNER-SOURCED CLIENT
          End client -> pays Partner (full engagement)
          Partner   -> pays Maxxim FLOOR + smaller CLIP
          Partner keeps MORE (they brought the client)

      (2) MAXXIM-SOURCED CLIENT
          End client -> pays Partner (delivery)
          Partner   -> pays Maxxim FLOOR + larger CLIP
          Maxxim keeps MORE (it generated the lead)

      (3) REFERRED CLIENT
          End client -> pays delivering Partner
          Maxxim/Partner -> pays Referrer a one-off
                            FINDER'S FEE on conversion

      (4) JEWELL-DIRECT (live exemplar partner)
          End client -> pays Jewell (full engagement)
          Jewell    -> pays Maxxim FLOOR + CLIP
                       (netted internally, same rails)

      In every flow:  Maxxim -> pays Jewell an IP LICENCE
                      Owners (Clent + Raef) -> profit share
      All splits [illustrative, pending pilots].

Pricing ladder (real anchors)

Current model: strengths and weaknesses

Strengths. Floor-plus-clip blends predictable recurring revenue (the floor) with upside that scales with partner success (the clip), and aligns Maxxim's interest with the partner's. The clip scaling by lead source is fair and incentive-correct: whoever brings the client keeps more. The cap on big upfronts stops a single large deal from making the platform look extractive. The $4,950 front door lowers buyer risk because it is credited forward. The low, swappable cost base is what makes a modest floor and clip credible in the first place.

Weaknesses. None of the rates are proven, so the model's actual margin is unknown until pilots run. The clip depends on clean attribution of lead source, which needs a simple, documented rule before the first Maxxim-sourced deal to avoid disputes. The premium band is wide ($25K to $250K), so the cap mechanism carries a lot of weight and must be tested against a real high-depth run where AI/compute cost is highest.

Model fit

The model fits the market well. It mirrors the buyer behaviour Discover identified: partners want margin and a repeatable engine, end clients want a low-risk first step and owned outcomes. Floor-plus-clip is the standard, well-understood shape for platform-plus-channel businesses, which lowers the explanation cost when recruiting operators. The recurring Deepen line is the piece most agency models never systematise and most tools cannot offer, so it is both a fit and a differentiator. Jewell running live on the same rails is the strongest possible evidence that the model works before the cohort has even filled.

Recommended adjustments

• Treat the $4,950 Diagnostic as the deliberate margin-and-data engine of the founding cohort, not just an entry price. It is the one clean, repeatable sale that produces the run data needed to lock everything else.
• Set a simple, written lead-source attribution rule before the first Maxxim-sourced engagement, so the clip scaling cannot be disputed later.
• Pilot the cap on big upfronts against an actual premium run (the ~$80K band) to confirm it protects margin without making the platform feel extractive to a high-value partner.
• Measure AI/compute per run from the first pilots, using the per-action cost logging that already exists for owners, so contribution margin is a real number and not an assumption.
• Consider a founding-cohort risk reversal: a low or waived platform floor during the pilot window so a partner can prove the model on a first engagement before committing to ongoing cost. Any terms are [illustrative, pending pilots] and require Clent sign-off.

New opportunities

• The self-serve portal as a second revenue surface (product-led, direct), sequenced after the channel is proven, evolving the existing read-only portal.
• Community-originated deal flow as a revenue multiplier: cross-partner referrals create engagements no single member would win, lifting clip volume without new acquisition spend.
• Specialist packaging at scale: as the cohort grows, Maxxim-sourced specialist modules become a margin line in their own right.
• Deepen at scale: as each partner's early engagements mature into recurring Deepen relationships, the compounding line becomes the durable base of platform revenue.

5. Key Metrics and KPIs

Most baselines are TBD because Maxxim is pre-scale and the brand motion is early. That is expected. The point of the founding cohort is to convert these from blanks into numbers. Targets below are directional and require Clent sign-off before any is treated as a commitment.

The unit economics to lock through pilots

These four are the spine. Until they are real, every revenue percentage stays illustrative.

Leading indicators (predict future performance)

• Founding partners signed. Target 5 to 8 by end of September 2026. The primary near-term proof of the channel.
• Growth Diagnostics sold per active partner. Directional target of at least one repeatable first sale per partner inside the pilot window; this is the pilot-data engine.
• Diagnostic-to-engagement conversion rate. The land-and-expand health signal; baseline TBD, to be set from cohort runs.
• Engagement-to-Deepen conversion rate. The share of landed engagements that continue as recurring Deepen relationships; the forward signal of durable LTV.
• Cleared case studies in hand. Target at least two additional beyond Pottsville, since proof gates partner confidence.
• Community participation rate. Share of partners active in the weekly table and monthly review; a forward signal of retention and shared deal flow.

Lagging indicators (confirm past performance)

• Platform recurring revenue (floor). The predictable base; grows with active partner count.
• Engagement clip revenue. Confirms delivery volume and lead-source mix.
• Deepen / retainer revenue. The compounding line; the truest measure of durable value.
• Partner retention / renewal. The again-and-again metric; the community moat shows up here.
• End-client satisfaction and asset-ownership delivered. Confirms the promise was kept at the gate.

6. Strategic Priorities

Five priorities for the engagement period, force-ranked. The ranking is dependency-driven: each earlier priority is a prerequisite for the ones below it. Prove the channel and lock the economics first, because nothing public can be committed until those are real.

Ranking rationale

Priority 1 is first because without partners there is no channel and no pilot data. Priority 2 is second because the data those partners generate is what unlocks every economic number. Priority 3 is third because it is the moat that makes the channel durable rather than a one-time recruitment. Priority 4 is fourth because proof reduces the largest competitive risk and accelerates 1 and 2. Priority 5 is last because the self-serve portal is the right move only after the channel is proven; doing it earlier would split focus before the model is locked.

7. Risk Assessment

Seven risks across categories, each with a response. A risk without a mitigation is just a worry.