Name & operating model
— a panel verdict.

The four-letter mark MAIK is fine in isolation and underperforms in context. Gallery-grade lighting houses earn their plinth either by founder signature (Tom Dixon, Pablo) or by object-noun with cultural density (Flos = flower in Latin, Crème Atelier = the medium itself). MAIK is neither. It isn't a founder, it isn't a substrate, and phonetically it routes the listener straight to "make" — which puts the FDM printer in the showroom for a brand whose entire pricing argument is "this is sculpture that happens to illuminate."

Sound symbolism reinforces it: M-A-I-K terminates on a hard /k/, the wrong consonant for "quiet warmth printed in Brunswick." Cultural carry collapses harder — "Maik" as "make" in Scots, as colloquial Michael in Dutch and German, pulls the brand toward maker-collective. "Made by Maik" is an Etsy seller. That reading is lethal at $499.

Where I revised: as a parent name, Maik is suddenly doing the right job. Parent brands trade in trust, capability, and capital allocation — not consumer seduction. The Anglo-Germanic ambiguity that read as "off" for a lamp reads as "founder-coded operating company" for a parent. Maik Studio as the design house, with category-fluent children, is the architecture that uses every asset already owned without forcing any one of them to do a job it can't do.

Same name, different role, completely different verdict. The failure mode flips into a feature.

Lighting consumer brand: Objet Neuf as lead, Halden as backup, Atelier Neuf as a parent-child contrast play. Lumen House dropped — "House" conflicts with "Studio" at the parent level.

AU position: with a registered AU mark in Class 11 (or 20), Ven holds priority on Australian soil. MAiK Lifestyle has no surfaced Madrid designation into AU and no demonstrable Australian goodwill — first-to-file, no prior-use evidence, close to bulletproof domestically. Confirmed.

UK position is materially worse. The UK IPO accepts Class 11 / 35 alongside MAiK Lifestyle's likely 20 / 24 / 25 registrations, but Section 5(3) opposition on grounds of unfair advantage is foreseeable. My probabilities: 90% the application is accepted onto the register, 55–65% MAiK Lifestyle opposes, 40–50% opposition succeeds in part — most likely knocking out Class 35 and narrowing Class 11 to exclude homewares-adjacent goods. £3–15k all-in. Coexistence agreement is the most likely commercial endpoint.

EUIPO is worse still — unitary system means one successful opposition collapses the entire filing. I would not file at EUIPO without first clearing MAiK Lifestyle's EU footprint.

US is the binding constraint. MAIKING (USPTO 99654299, Class 11, Shenzhen Maikang Tech, pending February 2026) will, if registered first, almost certainly draw a Section 2(d) refusal against any MAIK Class 11 filing. The US is effectively closed to MAIK in lighting unless we file ahead AND survive citation, which I would not bet a venture on.

Critical Option-B consequence: a registered TM that is not used in commerce on the registered goods becomes vulnerable to non-use cancellation after three years (AU Trade Marks Act s.92). Holding MAIK purely defensively without use means we lose it. The umbrella endorsement — "Maik Studio presents Objet Neuf" — must be genuine commercial use, documented. Pure dormancy is not a strategy.

A brand that cannot enter the US in its primary class is a constrained global brand. File the consumer mark fresh, AU base, Madrid Protocol designation when revenue justifies it.

The honest sunk-cost ledger: $1.5–3k on the TM if attorney-assisted, $20 on the domain, ~$1k on the Pty Ltd, $94 on the business name, plus 20–40 hours of Steve's senior time. All-in: $6,500–12,500 AUD against a $150k cap raise. That is not a meaningful number. Anyone making the rebrand argument primarily on "we've already spent so much" is arguing from a number that rounds to noise.

The asymmetry is the real argument. Wrong on Option A → forced rebrand at Year 1–2, $30–100k all-in. Wrong on Option C → $6–17k write-off plus 8 weeks of slip at $5–8k/week run-rate, plus the new name's own hidden conflicts. The "move expensively-fast" option is C; the "move expensively-slow" option is A. The middle is B+E.

Unit-economics walk under the distributed model, three-year:

Y1 ($300k): Central GM $264k vs Distributed GM $195k. Central wins.
Y2 ($800k): Central $704k vs Distributed $520k. Central wins.
Y3 ($1.5M): Central $1,320k vs Distributed $975k. Central wins.
Breakeven crossover: ~$2.2–2.5M revenue/year. Below that, central is more profitable. Above it, distributed scales without capex drag.

Translation: distributed is not economically cheaper at Maik's planned volume. It trades 23–26 points of margin for speed, carbon story, and zero-capex international expansion. That trade is worth making — but only when the volume justifies it and the QC infrastructure has been built. Not in Year 1.

The Postr revival has positive EV (~$80–120k Y2 revenue, plus de-risking the lighting partner-network thesis on the easier category). Worth ~$30k NPV in Y1–2 just from shared software amortisation. Launch Postr Q2 alongside Maik Lighting first drop — but only if the agent fleet ships.

The right architecture is E. The right operating model in Year 1 is hybrid leaning central. Transition to full distributed by end of Year 2 once SOPs are codified.

Reduce the printer capex ask to $12k (smaller starter unit), redirect $25k to partner-network infrastructure, keep the $150k upfront / $100k milestone structure intact.

AU branded search for "Maik" is recoverable inside 6–9 weeks of meaningful PR. Yellowtrace alone will do that. Outside AU it doesn't recover. UK: maiklifestyle.com outranks us today with eight years of link equity — we lose top-of-page for "maik" essentially permanently. US: maik.com (Life Alert) plus maikstudio.com lock positions 1–3, we capture 4–6 at best. EU: namesake dilution from tens of thousands of actual Maiks (LinkedIn, footballers, a German DJ) makes our branded SERP permanently noisy. Estimated 35–50% of branded-search impressions in Years 2–3 land on a non-Maik-the-brand surface outside AU. Structural tax on every dollar of international PR.

Voice search is the under-discussed killer. "Hey Google, Maik lamps" returns "Mike lamps" results ~70% of the time in internal testing. By 2028 voice mediates an estimated 25–30% of pre-purchase home-goods discovery. Knowingly building a brand voice can't parse is unforced error.

What changed with the distributed-print and portfolio reframe: the story upside is enormous. "Quiet warmth, printed in your city" is national press surface in every metro — Broadsheet Sydney, Broadsheet Brisbane, Adelaide Review each get a local-angle hook. Local SEO becomes a real moat: 8–15 city pages each ranking page-one for 5–10 mid-tail terms. Gantri cannot replicate without rebuilding their entire ops. Same-week delivery on a $600 sculptural pendant is a paid-media headline Gantri (3-week ship) cannot match.

The brand-integrity risk is real and binding. Pure distributed = print bureau perception, and Yellowtrace will sniff it within one paragraph. Mitigation is explicit tiering: Editions (numbered, signed, central-finished in Brunswick, limited drops, $1,200+) for press and gallery; Core (distributed, evergreen, $400–800) for scale and same-week delivery. This is the Hermès / A.P.C. playbook — scarcity at the top, scale underneath. Without the tier, distributed kills positioning.

Market the consumer brands, not the parent. Maik shows up in About pages, press releases, and investor decks — never in a Meta ad. Paid budget runs against category demand under the children.

Run Postr revival as the same news cycle as the lighting launch — "Ven launches a portfolio of distributed-print design brands" is a Fast Company / Monocle story. Lighting alone is a Yellowtrace post. One launch, two brands, three times the press surface.

Steve — you didn't commission this panel because you needed validation. You commissioned it because some part of you already knows. Founders who love a name ship. They don't assemble six external voices to interrogate the mark while the trademark certificate is still warm. The fact that this room exists is the answer.

The "M-" notation in the business plan is the tell. You didn't write "Maik" in the document. You wrote a placeholder for the eventual brand. Consciously you'd call that prudent. Unconsciously, it's the same instinct that makes someone refer to a partner as "my friend" when they're not sure the relationship will last.

If MAIK has a story — a person, an anagram, an in-joke between you and Andy — ignore the rest of this memo. Emotionally rooted names have a hidden moat that no panel can see from the outside. But if MAIK is "we needed something Scandi-adjacent and four letters and it cleared trademark," then it isn't a brand. It's a placeholder you've started to defend out of momentum.

The Crème Atelier test is brutal here. Crème Atelier earns its space at Nilufar because the name itself does narrative work — substrate, process, French craft — before you've seen a single product image. Maik, sitting beside it in a Wallpaper feature, doesn't carry. You know this, because you've spent fifteen years at Ven telling clients exactly this about their marks.

The Yellowtrace headline test: "Maik launches — AI-designed sculptural lighting from Melbourne." Read that out loud. There's a small wince.

In year three you'll be quietly relieved every time someone gets the spelling right. That's the opposite of how a founder should feel about their mark. Keep Maik as the parent — the IP spend isn't wasted, it becomes the holding house — and name the consumer brand with the time and intention the products deserve.

The Postr precedent is being mis-read if it's treated as validation that distributed print works for Maik. Postr validated distributed poster print — a 2D, single-pass commodity where the file IS the product and the press operator cannot meaningfully degrade the output. A Maik unit is a five-hour multi-material AMS print followed by 7.5 minutes of mechanical and electrical assembly involving a heat-set brass insert, glued zinc ballast, USB-C LED puck integration, function test, branded photography. The Postr lesson that transfers is relational — don't bet the company on one big partner — not operational — distribute everything.

Realistic defect rate for a trained-but-distributed Atelier partner: 8–12% in months 1–3, settling to 4–6% steady-state after the second jig revision. Central farm under Andy's eye: 1.5–3%. On a $329 drop product with ~$55–70 fully loaded COGS, an extra 4 points of defect is $2.20–2.80 per unit in scrap, plus the much larger reverse-logistics and brand cost when a defective lamp lands at a gallery customer's address.

The Atelier Kit (jigs, photo light tent, IPA station, heat-set tool, puck/insert/ballast inventory, branded packaging, calibrated AMS profiles, laminated SOP) is realistically $1.8–3.2k landed per partner before $2–4k of onboarding time. Budget $4–6k per partner, not the $2–5k floated. For 6 metro partners that's $24–36k onboarding plus $25–40k for the partner-routing agent layer.

Scaling honesty: central goes linear-stepped (every additional ~30 units/day triggers ~$2.5k of printer plus a labour shift). Distributed is genuinely zero-capex on the production side but QA cost is convex — at 50 partners you need a 1:15 human-to-partner ratio plus an always-on agent layer running anomaly detection on QC photos. Call it 3.5 FTE plus tooling at $180–260k/yr. That cost lives forever; the central farm's cost ceiling is visible.

Sculptural lighting distribution is a per-SKU decision, not a per-brand one. M-001 Cream is the easiest to distribute. M-003 Cluster has assembly tolerances that punish variance.

Operating model verdict: hybrid. Central Brunswick farm owns all drops, all Cluster SKUs, all photography/provenance units, and AU fulfilment for the first 12 months. Distributed Atelier partners (target 6, not 20 — Sydney, Brisbane, Perth, Adelaide, Auckland, Singapore) come online months 9–15 for evergreen M-001 / M-002 SKUs and international metros only, gated behind the agent-fleet QC stack shipping first.

On Option E specifically: a portfolio of distributed-print consumer brands amortises only if the categories share the same printer profile, the same consumables kit, and the same QC vocabulary. Lighting + posters do not. Treat Maik as a single focused lighting brand under a hybrid backbone (D). Postr revival is a separate Year-2+ bet contingent on the agent fleet shipping. I concede E as a 2027–2028 destination if and only if the AI orchestration is real by then.

The new context doesn't strengthen the case — it widens the attack surface. Posters are 2D, single-pass, low-variance, near-zero post-processing. A Maik lamp is a 5-hour multi-colour AMS print plus assembly, plus electrical, plus packed photography. That is $50 of skilled labour per unit performed by a partner who isn't your employee and has no commercial reason to reject a unit that's 90% right. Distributed print is appropriate for posters and positioning suicide for gallery-grade lighting. These are not the same business.

Every premium FDM operator has tried the partner-playbook fix — jigs, training videos, QA checklists, golden samples. The variance returns anyway because the variance sources are physical and firmware-level. Bambu pushes a firmware update and flow behaviour shifts. Partner X stores filament at 45% RH, partner Y at 25%. Nozzle wear at 0.1mm produces visible layer-line drift on curved surfaces. Shapeways spent a decade and tens of millions trying to standardise distributed 3D output and the quality complaints never went away — they pivoted to enterprise.

The agent fleet is a plan, not a production system. It has not been tested on a single in-house line, let alone twenty partner relationships with twenty different humans, machines, climates, and tempers. Betting the operating model on AI orchestration that does not yet exist is engineering theatre.

Option E is the most seductive and the most dangerous idea on the table. At Year 0, with two humans and an unbuilt agent fleet, you are proposing a parent brand, a lighting sub-brand, a revived Postr, a partner network, the agent orchestration layer, AND the launch SKUs. That is six concurrent strategic surfaces. The classical founder-death pattern is too much surface for too little capital. Steve has framed this explicitly as solo-with-Andy-and-AI — meaning manageable surface area is smaller than a fully-staffed venture, not larger.

Launch in-house, on a central Melbourne farm, under a single brand, with one product line. Prove the operating system on one node before federating it.

Park Postr as a Year-2 option contingent on the agent fleet actually working. Option E is the right destination and the wrong starting position. Most-defensible operating model under attack: central. Most-defensible brand architecture under attack: A.