Paid Media Plan

How to read this plan

This is the paid advertising plan for Maxxim, produced in the Deploy phase against Master Briefing v2.1, the Gate 1 decisions of 27 June 2026, and the 1 July refresh that folded in the 29 June product mechanics and the live Jewell partner site. It sits downstream of Marketing Plan and applies Brand Strategy and Audience Teardown. It is deliberately narrow. Paid media is not the engine of Maxxim's near-term growth. The marketing plan is explicit that the model is proof-led and partner-led: the 3 July session, the cleared proof, the live Jewell partner, and the founder relationships do the persuading, and end-client demand generation is the partner's job on the partner's own channels. Paid sits behind organic and founder-led social as amplification, not the primary engine. This document specifies how to run the small, sharp paid layer that supports that motion, and where not to spend.

One frame governs every page. Near term, the paid priority is partner recruitment for the founding cohort, weighted to LinkedIn and aimed at trusted operators: the capacity-capped principals who would license the method or refer clients in. Search intent sits around agency enablement and AI-run marketing, which is where an operator evaluating this category actually looks. The secondary paid job is a small, always-ready end-client Growth Diagnostic intent layer on search that a partner or Jewell can switch on for the universal end-client audience. Everything else is held back.

The same discipline as every upstream document applies. All budgets, splits, caps, and spend figures here are illustrative, pending pilots, and require Clent sign-off before they are treated as fixed. This plan presents structure and approach, not a committed media budget. Real pricing anchors are stated as real: the Growth Diagnostic at $4,950 credited forward, the premium band of $25,000 to $250,000 with a recent run near $80,000, the specialist module near $1,500. Proof respects clearance: Pottsville Acupuncture is the only cleared public Maxxim proof, so it is the only case study used in Maxxim-brand ads. OTR Earthmover is internal only, Hidrive is percentages only, and Jewell's own named results stay under Jewell's banner and are never re-used as Maxxim's public proof in an ad. Ownership is Clent and Raef only. Maxxim's own category lead word is OPEN and is not invented or built into keyword targeting; Jewell's external lead word is growth. The 3 July session promise is OPEN and Clent's to finalise, so recruitment creative invites into the founding table rather than asserting a fixed promise. Australian English throughout, no em dashes, no exclamation marks.

What paid media is amplifying, in product terms

Paid does not sell a promise, it amplifies a showable engine. The 29 June product mechanics give the paid creative its substance and keep every claim inside what is real. This is the same substance the marketing plan and brand strategy lead with, applied to the paid layer.

• Three named specialists, each a process. Max does the words in Claude Cowork, Sal does the pixels in the Design Studio, Pip runs the Social Studio. Naming them makes the engine concrete in a recruitment ad and gives the walkthrough video its content. They stay subordinate to Maxxim and are never the accountable human. The accountable human is always the partner.
• The method is gated, and the gate is the proof. Discover, Design, Deploy, then Deepen, the compounding loop. Creative unlocks only as its briefs exist and a logo is locked. In a paid context, the gate is the visible answer to the trust objection, the thing a considered operator wants to see before they license.
• Human in the loop is the point, not a limitation. AI carries the repeatable 80 percent at speed; the human owns the 20 percent clients pay for. This is the lead message angle for LinkedIn recruitment, tested against the proof-led Pottsville angle.
• Finished, owned outcomes. Client sites are Astro on Cloudflare Pages, images from the Maxxim asset store, an embedded contact form now. The client owns the asset outright. In an end-client Diagnostic ad this is the credible differentiator against agency lock-in.
• A live partner already runs the whole model. Jewell is in market on Maxxim as the engine behind it, which sharpens the recruitment pitch, but Jewell's named results are Jewell's, so paid creative references the partner model as pattern and does not lift Jewell's case studies into Maxxim-brand ads.
• Built lean, on purpose. Only Max plus the engine is built in-house. The marketing consequence carries straight into paid: proof and relationships persuade, not a large media budget, which is exactly why this paid layer is small, sharp, and held back.

1. Paid Media Objectives

Paid media serves the marketing plan objectives, not its own. It has one primary job and one secondary job, and it is sized to a founding-cohort motion, not a demand-at-volume motion.

• Primary business goal. Generate qualified operator interest for the founding cohort of 5 to 8 partners across July to September 2026: warm the recruitment pipeline, profile the 3 July session and the category, and drive session registrations and qualified operator enquiries. This supports marketing plan Objectives 1 and 2.
• Secondary business goal. Capture end-client Growth Diagnostic intent through search, so that when a partner or Jewell wants to switch on demand for a real engagement, the $4,950 credited Diagnostic has a ready paid surface to catch high-intent search. This supports marketing plan Objective 3.
• Explicit non-goal. Paid media does not run end-client demand generation at scale. That is the partner's job on the partner's own channels (marketing plan, Layer 2). Paid does not try to out-spend horizontal AI players, and it does not lead the brand.

Primary KPIs

Benchmarks and why they are ranges

Maxxim has no historical ad-account data, so no cost-per-acquisition or return-on-ad-spend benchmark can be stated as a baseline. The honest position is that the founding cohort is what converts these blanks into numbers. As illustrative external anchors only, B2B LinkedIn lead-generation cost per lead commonly sits in a wide band of roughly seventy-five to two hundred dollars or more for a considered, principal-level audience, and it should be read here as directional, not as a target Maxxim has committed to. The decisive economic number is not a media cost per acquisition at all. It is the value of a signed partner: a single licensed partner unlocks a delivery channel and recurring floor-plus-clip revenue, and again through the Deepen retainer once their clients are live (rates illustrative, pending pilots). Because one conversion is worth so much and the audience is so small and high-trust, a higher cost per qualified lead is acceptable here than in a volume consumer model.

Budget range and timeframe

Budget is deliberately modest and illustrative, consistent with the marketing plan's roughly 70 percent brand-and-community, 30 percent direct-response split. Paid sits inside the direct-response slice and inside the recruitment and enablement lines, not as a separate large budget. No total figure is committed here. The plan is built to scale to whatever modest monthly figure Clent signs off, with the structure below holding at any level. Timeframe is the founding-cohort window: a recruitment push now through the 3 July session, then sustained light spend across the July to September 2026 cohort window, with the search layer running always-on at a low floor.

2. Audience Strategy

Two audiences, mapped to the two layers of the marketing plan and drawn straight from Audience Teardown. Layer A, the partner, is the paid priority and is targeted almost entirely on LinkedIn because that is the operator-to-operator surface. Layer B, the end client, is universal across organisation types and is reached on search through intent, not interruption, because intent is where a low-volume, high-trust product belongs. The map below shows how the targeting layers stack from broad to narrow, and which platform each audience sits on.

   TARGETING + AUDIENCE MAP

   AUDIENCE A: THE PARTNER  (Layer 1, PRIMARY, LinkedIn-weighted)
   target = the capacity-capped operator / B2B principal
   +-------------------------------------------------------------+
   |  BROAD  demographics + firmographics (LinkedIn)             |
   |   seniority: owner, founder, principal, director, partner   |
   |   function: marketing, growth, agency, consulting           |
   |   company size: 1-10 and 11-50 (operator-scale firms)       |
   |                          |                                   |
   |  NARROWER  member traits + interests                        |
   |   marketing agency owners, fractional CMOs, growth          |
   |   consultants, web/brand studios, solo operators            |
   |                          |                                   |
   |  CUSTOM  matched audiences                                  |
   |   the named pipeline + warm list (Alex/3P, Arran, Greg,     |
   |   Christy) as a contact-list match; site visitors           |
   |                          |                                   |
   |  RETARGET  engagement pools                                 |
   |   visited /for-partners or the session page; opened a       |
   |   lead form; watched the Max/Sal/Pip walkthrough video      |
   +-------------------------------------------------------------+

   AUDIENCE B: THE END CLIENT  (Layer 2 SUPPORT, search-led)
   target = UNIVERSAL org type, served through a partner
   +-------------------------------------------------------------+
   |  INTENT  keyword search (Google), high-intent only          |
   |   "growth diagnostic", "marketing diagnostic", "[service]   |
   |    + [town]", brand defence on "Maxxim" / "Pottsville"      |
   |                          |                                   |
   |  BROAD-TO-NARROW is by INTENT, not demographics, because    |
   |   the org range is universal (personal brand -> government) |
   |                          |                                   |
   |  RETARGET  site visitors who viewed the Diagnostic /        |
   |   Pottsville proof but did not enquire                      |
   |                          |                                   |
   |  NOTE  NFP + GOV carry procurement/probity; paid does NOT   |
   |   chase them. Partner-led + panels reach them (see S3).     |
   +-------------------------------------------------------------+

Audience A: the partner (LinkedIn, primary)

Audience B: the universal end client (search, support layer)

Privacy note carried up front. Both audiences depend on consent-respecting tracking. The end-client retargeting pools and the matched lists only operate behind a compliant consent banner and consent mode (Section 9). This is a measurement constraint, not a footnote, and it shapes how small the addressable retargeting pools will be early on, consistent with the privacy-respecting posture of the Astro-on-Cloudflare site.

3. Channel Plan

The paid channel mix is intentionally minimal: one primary platform for recruitment, one support platform for intent capture, and a thin retargeting layer. The flow below shows how an illustrative budget divides between the recruitment job and the intent-capture job, with the heaviest weight on LinkedIn because that is where the operator audience is and recruitment is the priority. Every figure is illustrative, pending pilots and Clent sign-off.

   CHANNEL + BUDGET FLOW  (illustrative %, pending sign-off)

   MODEST MONTHLY PAID BUDGET  [illustrative, no $ committed]
            |
            |  paid sits BEHIND organic + founder-led social
            |  (amplification, not the primary engine)
            v
   +--------------------------------------------------------+
   |  LAYER 1: PARTNER RECRUITMENT (primary)                |
   |                                                        |
   |   LinkedIn Ads .................. ~60%                 |
   |     - Sponsored Content (single image + video)         |
   |     - Lead Gen Forms (in-platform capture)             |
   |     - retargeting of /for-partners + session visitors  |
   |                                                        |
   |   LinkedIn = the operator-to-operator surface;         |
   |   B2B principal targeting; the recruitment workhorse   |
   +--------------------------------------------------------+
            |
            v
   +--------------------------------------------------------+
   |  LAYER 2 SUPPORT: END-CLIENT DIAGNOSTIC INTENT         |
   |                                                        |
   |   Google Search ................. ~25%                 |
   |     - high-intent diagnostic + brand-defence terms     |
   |     - switch-on local terms for a partner's client     |
   +--------------------------------------------------------+
            |
            v
   +--------------------------------------------------------+
   |  RETARGETING + TEST RESERVE                            |
   |                                                        |
   |   Retargeting (LinkedIn + light Google Display) ~10%   |
   |   Test / contingency reserve .... ~5%                  |
   +--------------------------------------------------------+

   TOTAL = 100% of a modest, illustrative paid budget.
   Ramp: learn first (Section 7), then scale the winners.

Channels NOT recommended (and why)

Stating what not to run is as load-bearing as what to run, given a modest budget and a proof-led model.

• Meta (Facebook and Instagram) for partner recruitment, not recommended now. The operator audience is reached better and more precisely on LinkedIn by seniority and firmographics. Meta's strength is consumer reach, which is not the Layer 1 job. A single small Meta awareness test to a matched operator list may be run from the test reserve, but Meta is not a core recruitment channel.
• TikTok, not recommended. The audience is B2B principals making a considered licensing decision, not a short-form consumer audience. The budget does not support a third social platform, and the brand voice (no exclamation marks, claim-led, Sage) does not fit the platform. Explicitly do not use TikTok.
• YouTube paid (in-stream and pre-roll), not recommended now. Video-first awareness at scale is premature for a founding-cohort motion and a thin proof base. The Max, Sal, and Pip walkthrough video should run as organic and as LinkedIn-native video to a tight audience, not as broad paid pre-roll.
• Google Display prospecting and programmatic, not recommended. Broad display manufactures impressions, not qualified operator interest, and risks brand-safety exposure for a young brand that lives on credibility. Display is used only for warm retargeting, never cold prospecting.
• Performance Max, not recommended now. It needs conversion volume and a clear value signal to optimise. Maxxim has neither yet. Revisit only once the search layer has accrued meaningful conversion data and the Diagnostic value is confirmed.
• Paid end-client demand generation at volume, not recommended by design. Layer 2 demand is the partner's job on the partner's own channels. Maxxim-run paid stops at intent capture and partner support. This also holds the OPEN end-client visibility and ownership questions where the marketing plan leaves them, rather than pre-empting them with a spend commitment.

4. Campaign Structure

A lean account structure a media buyer can stand up directly. Two ad accounts, a single naming convention, and a clean prospecting-versus-retargeting split inside each platform.

• Account structure. One LinkedIn Campaign Manager account and one Google Ads account, both under Maxxim, with conversion tracking unified through the Maxxim tag setup (Section 9). No separate partner sub-accounts at this stage; partner-run local search campaigns, if any, sit in the partner's own account and are out of scope for this Maxxim-run plan.
• Naming convention. Maxxim_[Platform]_[Objective]_[Audience]_[YYYY-MM]. For example, Maxxim_LinkedIn_Recruit_OperatorCold_2026-07, Maxxim_LinkedIn_Recruit_Retarget_2026-07, Maxxim_Google_Diagnostic_Intent_2026-07, Maxxim_Google_BrandDefence_2026-07. The convention feeds clean reporting and the lead-source attribution the revenue model depends on.
• Campaign to ad set or ad group segmentation. On LinkedIn, separate campaigns for cold prospecting (firmographic plus interest), matched-list and lookalike, and retargeting, so budgets and bids are controlled per audience temperature. On Google, separate ad groups by keyword theme (diagnostic intent, local service, brand defence) so message match is tight and quality score holds.
• Audience segmentation within campaigns. Prospecting and retargeting never share a campaign. Tiers run warm to cold: matched named-pipeline list and site converters first (highest priority), then lookalike and interest prospecting, then broad firmographic. Retargeting is its own line with its own creative.
• Geographic and scheduling settings. Australia first, aligned to the founder network; widen only as the cohort proves out. Schedule LinkedIn delivery to business hours and weekdays, where principals engage, during the learning phase, then open if data supports it. The search layer runs continuously at a low floor so high-intent Diagnostic queries are never missed.

5. Creative Requirements

This plan specifies formats, dimensions, and variant counts; it does not write the copy or design the assets. Copy direction references the Brand Strategy messaging pillars and the locked voice (no em dashes, no exclamation marks, claim-led emphasis). Where the assets are produced by the engine, Sal makes the pixels in the Design Studio and Max drafts the words, with a human signing off, which is the model marketing itself. Ad production stays inside the gate: campaign pieces need the Design Brief and a locked logo, so paid creative is briefed against those, not ahead of them.

Creative testing plan

Test in priority order, one variable at a time so the read is clean. First, message angle on LinkedIn cold: the engine-for-your-judgement (80/20) claim against the proof-led Pottsville angle, because the winning angle shapes everything downstream. Second, format: single image against the Max, Sal, and Pip walkthrough video, to learn whether demonstration outperforms a static claim for this audience. Third, CTA and capture path: native Lead Gen Form against a click to the session landing page, to learn where qualified operators convert more cheaply. On Google, test headline theme (Diagnostic outcome versus named-human accountability versus speed) once volume allows. Audience tests (matched list versus lookalike versus interest) run in parallel at the campaign level, not blended into the creative test. Launch every ad group with at least two to three variants; never run a single creative.

6. Landing Page Strategy

Targeting and creative cannot fix a weak landing page, so message match and conversion readiness are flagged as dependencies, not assumptions. Paid points to pages on maxxim.ai (Astro on Cloudflare Pages) that must exist and convert before spend ramps. These dependencies sit with the website build; see Website Strategy.

Conversion tracking on the page. Every landing page carries the Google tag, the LinkedIn Insight Tag, and event tracking for the key actions (enquiry submit, session register, Diagnostic enquiry), all gated behind consent (Section 9). The embedded contact form now live at apps.maxxim.ai is the current capture surface, so its submit event is the anchor conversion; the newsletter and chat apps come later and are not depended on here. Landing-page A/B testing is deferred until traffic justifies it; early on, fix message match and load speed first, then test the single highest-leverage element (the headline and the primary CTA) once volume supports a clean read. The overriding principle: the best targeting will not rescue a page that does not match the ad or does not make the next step obvious.

7. Budget Allocation and Pacing

Pacing follows a learn-first discipline: never full spend from day one. The table shows an illustrative three-month shape across the founding-cohort window, with the recruitment push concentrated around the 3 July session. All figures are percentages of a modest, illustrative monthly paid budget; no dollar total is committed, and the whole table requires Clent sign-off. The split holds at any modest budget level Clent approves.

• Daily spend limits. Set per-campaign daily caps so a single campaign cannot run away during the learning phase. LinkedIn cold prospecting carries the largest cap; retargeting and brand defence carry small caps because their audiences are small. The search layer runs at a low daily floor continuously.
• Ramp-up schedule. Month 1 deliberately runs at roughly half to sixty percent of steady-state spend to gather data without overspending on unproven audiences and creative. Scale only what clears the optimisation triggers in Section 10. This protects a modest budget from being burned before the learning is in.
• Seasonal adjustments. The one fixed peak is the 3 July session; concentrate recruitment spend in the run-up. There is no consumer seasonality to chase in a partner-recruitment motion. Beyond the cohort window, sustain the search floor and pulse recruitment spend around quarterly community sessions.
• Reallocation triggers (illustrative). If LinkedIn cost per qualified lead exceeds an agreed ceiling for fourteen days, shift budget to the best-performing retargeting pool or pause and refresh creative. If Google Diagnostic intent shows no qualified enquiries after a defined click threshold, reduce the search floor and hold the saving in the reserve. If a clear winner emerges, move the test reserve behind it. No reallocation rule commits a number Clent has not signed off.

8. Bidding and Optimisation Strategy

Bidding starts manual or capped during learning, then moves to automated targets only once each campaign has enough conversions to optimise honestly. With no historical data, automated bidding has nothing to learn from on day one, so the early discipline is control.

• Bid adjustments. On Google, apply location bid nudges to match a partner's service area when local campaigns are active, and a remarketing bid uplift for site visitors. On LinkedIn, lean on audience tiering rather than granular bid modifiers. Device and time-of-day adjustments only after data shows a clear pattern.
• Learning-phase expectations. Be explicit with the AM and Clent that low conversion volume means a longer, noisier learning phase than a high-budget account. Automated bidding generally needs a sustained run of conversions before it stabilises, and Maxxim will reach that slowly. This is expected for a founding-cohort motion and is a reason to keep early bids manual and budgets capped.

9. Tracking and Attribution

Measurement is specified so a media buyer can set up tracking from this document alone, and so paid reconciles cleanly with the marketing plan's two-layer attribution. The single most important discipline is unchanged from the marketing plan: the lead-source rule tags every engagement partner-sourced, Maxxim-sourced, or referred before delivery begins, because that tag drives the clip and the finder's fee. Paid attribution feeds that rule; it does not replace it.

• Conversion tracking setup. Install the Google tag (gtag or GTM) and the LinkedIn Insight Tag on maxxim.ai, plus the Google Ads conversion tag. Configure GA4 as the cross-platform source of truth, with platform-native reporting used for in-platform optimisation. All tags load through a consent-gated tag manager.
• Conversion events. Define and track: 3 July session registration (primary Layer 1 event), qualified operator enquiry (Layer 1), Diagnostic enquiry submit via the embedded contact form (Layer 2), and Max, Sal, and Pip walkthrough video 50 percent view (a soft, retargeting-only signal). Assign a nominal lead value to qualified operator enquiry to let bidding optimise toward quality, recognising the true value is a signed partner, not a media conversion.
• UTM conventions. A fixed structure on every paid link: utm_source (linkedin, google), utm_medium (paid-social, cpc), utm_campaign matching the campaign name (for example maxxim_linkedin_recruit_operatorcold_2026-07), utm_content (creative variant), utm_term (keyword, search only). Consistent UTMs are what let GA4 reconcile the platforms and what feed the lead-source tag.
• Attribution model. For Layer 1 recruitment, use first-touch-plus-confirmed-source, matching the marketing plan: record where a partner first heard of Maxxim and confirm the source at sign, because the decision is near-solo and relationship-led, not a multi-touch web journey. For Layer 2 Diagnostic intent, use data-driven attribution in GA4 where volume permits, falling back to last-click while volume is thin. Neither paid model overrides the commercial lead-source rule that drives the clip.
• Cross-platform reconciliation. GA4 is the source of truth for cross-channel reporting; LinkedIn and Google native reporting are used only for in-platform bid and creative decisions. Expect platform-reported conversions to over-count against GA4; reconcile to GA4 for any spend or return conversation.
• Privacy and consent (fundamental, not a footnote). A consent management banner and Google consent mode v2 must be live before paid tracking is trusted, consistent with the privacy-respecting posture of the Astro-on-Cloudflare site. iOS and browser tracking restrictions mean retargeting pools will be smaller and view-through data weaker than platform dashboards imply; plan for server-side tagging where feasible to improve signal. For any end-client data captured, honour public-sector and not-for-profit privacy obligations where a partner activates those audiences. No tracking that consent does not cover is recommended; if it cannot be measured cleanly and compliantly, it is not run.

10. Reporting and Optimisation Cadence

The paid cadence nests inside the marketing plan's reporting rhythm so paid is one input to the community business review, not a separate report. The funnel below shows the paid path stage by stage with the creative that carries each stage, then the cadence table sets the operating rhythm.

   THE PAID FUNNEL  (creative by stage)

   LAYER 1: PARTNER RECRUITMENT  (LinkedIn, primary)

   AWARE     +--------------------------------------------+
             | LinkedIn Sponsored Content (cold)          |
             | creative: 80/20 claim OR Pottsville proof  |
             | KPI: reach vs operator profile, CTR        |
             +--------------------------------------------+
                          |  clicked / engaged
                          v
   CONSIDER  +--------------------------------------------+
             | Max/Sal/Pip walkthrough VIDEO + retarget   |
             | creative: human signs the gate; the table  |
             | KPI: video view, landing-page visit        |
             +--------------------------------------------+
                          |  showed intent
                          v
   CONVERT   +--------------------------------------------+
             | Lead Gen Form OR 3 July session page       |
             | creative: register for the founding table  |
             | KPI: qualified operator lead / register    |
             +--------------------------------------------+
                          |  qualified lead
                          v
             ==> HANDOFF to founder 1:1 + the cohort
                 (paid stops here; relationship closes it)

   ----------------------------------------------------------

   LAYER 2 SUPPORT: END-CLIENT INTENT  (Google, parallel)

   INTENT    +--------------------------------------------+
             | Responsive search ad on Diagnostic terms   |
             | creative: $4,950 credited Diagnostic;      |
             |           named human accountable          |
             | KPI: CTR, click-to-enquiry                 |
             +--------------------------------------------+
                          |  visited, did not enquire
                          v
   RECOVER   +--------------------------------------------+
             | light retargeting -> Diagnostic page       |
             | creative: Pottsville proof reminder        |
             | KPI: a