Advisory working session · 1 June 2026 · Michael + Martin
Elisa's separation & Seeda's future
A response to your strategic blueprint, sharpened by evidence we did not have when you built it. The headline: your deck merged two decisions that need to be pulled apart, and the new Slack record makes Elisa's 2026 bonus claim much weaker than the blueprint assumed. This deck reframes the choice, re-quantifies the exposure, and lands on a clear recommendation for Tuesday.
Hard deadline
Tue 3 Jun
Elisa's signed agreement
Disputed claim
A$28,578
2026 "rev share" to Aug 21
Evidence found
4 + 3
Slack proofs + Fathom records
Real risk
Vincent
Parallel claim · must retain
Page 2 · The core reframe
1 · THE REFRAME
Your deck merged two decisions. They must be separated.
The blueprint forces "what do I do with Seeda's future" into the same June 3 box as "how do I respond to Elisa." That creates pressure to make a huge, irreversible decision (close the company) to solve a comparatively small problem (a disputed A$28k bonus). That is the tail wagging the dog. Pull them apart and both get easier.
Decision A · hard deadline Tue 3 June
The Elisa separation
A contained negotiation: notice, transition scope, comp, equity. Real deadline, modest stakes, strong evidence on our side. Resolve it cleanly and on our terms.
Decision B · no deadline · 6–12 month call
Seeda vs Seby (the strategy)
Whether to evolve Seeda toward AI, wind down MMM, or stand up Seby. This is a months-long strategic call. It must NOT be made under the artificial pressure of Elisa's notice period.
Why it matters: the moment you separate them, the radical options in the blueprint (close the company, hardball the bonuses) stop looking necessary. They were only attractive because the deck made the small problem feel like an existential one.
Page 3 · What changed since your deck
2 · NEW EVIDENCE
The 2026 claim is weak, not settled law
Your blueprint treated the A$28,578 as a near-certain liability. It was built before we pulled the Slack record. Elisa's case rests on "no signed amendment, so the original 10% continues." The contemporaneous written record says both she and Vincent accepted the structure change in early 2026.
"Elisa's POV: 2025 rev share to be paid after end of 2025" — Michael's written summary to Elisa, which she confirmed with a ✅ reaction.Slack · M→Elisa DM · 17 Feb 2026
"ok to switch to Profit share bonus from Jan 2026 onwards … let me know if i've got any of this wrong?" — Vincent replied two minutes later: "Thanks for the summary."Slack · M↔Vincent DM · 11 Feb 2026
"I would agree on the suggestion you made to me … Leadership team is paid the same on a fix salary." — Vincent, in the group channel, Elisa present and silent.Slack · M+E+V group DM · 19 Feb 2026
Three Fathom recordings back this up: the Dec 12 Year-in-Review (Elisa herself proposes converting the 2025 bonus to shares/cash), the Dec 15 Vincent catch-up ("we can't afford it"), and your Jan 16 budget session with Martin where the 2026 framework was set as market-rate base plus new incentive plans, with both in the room and no objection.
Honest caveat: there is no single line where Elisa says "I agree the 10% ends." Her ✅ on a 2025-bounded summary plus Vincent's written acceptance is strong, but it is acceptance-by-conduct, not a signed amendment. That is exactly why we settle the principle rather than bet the company on winning it outright in a tribunal.
Page 4 · Re-quantified exposure
3 · THE REAL NUMBERS
The liability waterfall, corrected
Your deck's exposure figure bundled an owed debt with a disputed one. Separating them shrinks the real threat.
| Layer | Amount | Status | Our position |
|---|
| 2025 rev share · 2nd installment | A$18,902 | Owed | Pay it · acknowledged Mar 6, accepted Mar 18, in the P&L |
| 2026 "rev share" claim (to Aug 21) | A$28,578 | Disputed | Defeatable / small-settle · evidence on our side |
| Base salary · transition drain | ~A$10,000 | Conditional | Only if we grant full 3-month transition |
| Vested options @ Aug 21 | 32,810 opts | Good Leaver | Concede · buy-back-at-$1 window expired Nov 2025 |
What's genuinely owed today
A$18,902
The 2025 installment only
What's actually in dispute
A$28,578
The entire fight is this one line
Likely settled cost of the dispute
A fraction, contingent
Not the headline number
Reminder: whatever the 2026 number resolves to, double it in your head for planning. Vincent has the identical claim. That is the real reason to settle on a structure that works for both, not to win a one-off against Elisa.
Page 5 · Decision A — recommendation
4 · DECISION A · ELISA
Modified Option 1A: do the transition, resolve comp on our terms
Take the 3-month transition because the client retention is worth it (Coffee Club renews 15 July, Inspirations Paint 31 Oct), but do not capitulate to her reading of the bonus. Five moves:
Pay
2025 installment A$18,902. It is owed and acknowledged. No games, no leverage plays.
Settle
2026 claim → contingent on profitability, Dec 2027 sunset (your email structure). A good-faith partial amount, payable only if the business reaches profit. Holds the principle, gives her a path.
Counter
Defined-deliverables transition, not all-or-nothing. Tie her pay to outputs: documentation, the two renewals, Vincent/Sophia knowledge transfer. Caps the "rest in France on Seeda's dime" risk she hinted at.
Concede
Good Leaver + expired buy-back. Both are cheap goodwill and legally probably correct. Fighting them looks petty and weak.
Protect
Vincent above all. Whatever you offer Elisa, Vincent gets the same logic. The contingent structure works for both. Lose Vincent and the business ends.
Net: she stays to Aug 21 doing defined, useful work; you pay what's owed; the disputed money becomes contingent and modest; Vincent is held steady; nobody gets torched.
Page 6 · Two things we must not do
5 · RED FLAGS
Two moves in the blueprint that would backfire badly
3C
"Close Seeda, start Seby" to escape the obligations — this is illegal
Closing a company specifically to avoid paying owed wages, bonuses and equity, then restarting the same business with the same founder and Vietnam team under a new name, is illegal phoenix activity under Australian law. ASIC and the ATO prosecute it: personal liability, civil penalties, potential criminal exposure. And it is now in writing in a deck that has been emailed and is discoverable. Take 3C off the table entirely — do not action it, reference it, or let it shape the thinking.
!
"Withhold both bonuses as leverage" — conflates two very different debts
The A$18,902 is a debt you confirmed in Slack (Mar 6), Elisa accepted (Mar 18), and it sits in your own May 21 P&L. Withholding an acknowledged debt to gain leverage on a disputed one is bad faith and legally weak — it converts a winnable position into a Fair Work / breach claim with costs against you. Pay what you've acknowledged; contest only what's genuinely disputed.
Page 7 · The Vincent hinge
6 · THE REAL ASSET
Every path runs through Vincent
Your own note nailed it: "anything you do with Elisa sets precedent good to bad with Vincent." He is the bigger strategic question than Elisa, and the blueprint barely centres him.
- He is staying (your read), and he does the core technical work the business runs on.
- His claim is parallel but he's been reasonable in writing — "we can't afford it," "thanks for the summary," even "deduct that from my bonus" over the Daniel mess. He is negotiating in good faith.
- He has been the swing voice for fairness, not maximalism — he proposed equity and fixed-salary structures himself.
The trap: any aggressive move against Elisa (immediate exit, withholding, closure) reads to Vincent as "this is how Michael treats a departing co-founder." 3C explicitly nullifies his ESOP too — he would walk, and that collapses the company. The contingent-settlement path is the only one that keeps faith with both.
Page 8 · Decision B — Seeda's future
7 · DECISION B · STRATEGY
Evolve toward AI, keep the entity, no closure
The honest read: MMM is commoditising (open-source PyMC is making the core tech cheap), Seeda is sub-scale and churning, and your real edge now is the Claude automation you've built. Directionally, leaning into Seby/AI over the next year is probably right. But that's evolution, not demolition — and it's on a long clock, not June 3's.
2/3A
Keep Seeda the entity · build Seby inside or alongside it
Pro: keeps the brand, the tax-loss credits, the billing infrastructure and the relationships. Let MMM run and wind down naturally as customers churn; stand Seby up with clean books in Xero. Fast to launch off the Claude engine and the Vietnam team. Clean reputation intact.
1B
Immediate exit
Loses client handover, spooks Vincent, opens a revenue hole. Only right if Elisa were genuinely toxic — the record doesn't show that.
3B
Dual stream (MMM + Seby in parallel)
Your deck is right — the worst option. A solo founder running two different sales cycles is unsustainable. Maximum complexity, minimum clarity.
3C
Clean slate (close + restart)
Illegal (see page 6), ends Vincent, reputational ruin. Never.
Page 9 · Synthesis matrix — re-scored
8 · RE-SCORE
Your six options, re-scored with the evidence in hand
The new Slack record collapses the "financial & legal" advantage the radical options seemed to have, because the disputed money is now small and contingent. That changes the picture.
| Option | Cash preserved | Legal safety | Revenue continuity | Vincent retained | Founder sanity |
|---|
| 1A modified Pick | ◑ | ● | ● | ● | ◑ |
| 1A as Elisa proposed | ○ | ◑ | ● | ● | ○ |
| 1B immediate exit | ● | ◑ | ○ | ○ | ◑ |
| 2 data pivot | ◑ | ◑ | ○ | ◑ | ◑ |
| 3B dual stream | ◑ | ◑ | ◑ | ◑ | ○ |
| 3C clean slate | ● | ○ | ○ | ○ | ◑ |
Read: once the disputed money is small and contingent, 3C's only edges (cash, "legal isolation") evaporate — and it scores zero on the two things that actually keep the business alive: legal safety and Vincent. Modified 1A wins on every column that matters.
Page 10 · The plan + what I need from you
9 · NEXT 7 DAYS
The plan
Today
Holding email to Elisa. Agree the operational + 2025 items; defer 2026 bonus, Good Leaver, and SHA to Monday; ask to move sign-off from 3 → 6 June.
This weekend
Confirm Martin's recollection of the Jan 16 budget in writing, and pull Vincent's contract to size the parallel claim.
Mon–Tue
Counter to Elisa: pay 2025 · settle 2026 contingent (Dec 2027 sunset) · defined-deliverables transition · concede Good Leaver.
This week
One legal consult (AU employment + corporate): Good Leaver mechanics, enforceability of the contingent bonus, and a clean line away from anything phoenix-adjacent.
Separate clock
Open the Seeda-vs-Seby strategy track on its own months-long timeline. Not tied to June 3.
10 · OVER TO YOU, MARTIN
Four things I genuinely need your judgement on
Q1Do you agree to decouple the two decisions — settle Elisa cleanly now, and run the Seeda/Seby strategy on its own clock?
Q2Is the contingent-on-profitability bonus structure (your Dec 2027 sunset idea) something we can actually make stick in a separation agreement, or does it need a different mechanism?
Q3What's your read on Vincent — how do we present this to him so the Elisa settlement reassures rather than rattles him?
Q4Who's your go-to AU employment/corporate lawyer for a fast consult this week? I want a name, not a search.
Appendix · On your blueprint
A · WHAT I LIKED
What your deck got right (and why I'm building on it)
- The June 3 threshold framing. There is a real decision point and naming it stops drift. Correct and useful.
- The liability waterfall. Three layers plus equity is the right way to see the money — I've simply corrected which layer is owed vs disputed.
- The synthesis matrix. Scoring options across competing objectives is exactly the right tool; I kept it and re-scored with the new evidence.
- Flagging the Vincent precedent and the ESOP exposure. Both are the crux, and your instinct to surface them was right.
- "Involve lawyers." The instinct is correct — I've kept it as a concrete step, just pointed it at the right questions.
B · WHAT I ARGUE AGAINST
Where I push back on the blueprint
- It's a NotebookLM options dump, not your considered advice. Your email reply (contingent + Dec 2027 sunset) was wiser than the deck. I've treated the deck as the menu and your email as the judgement.
- It merges two decisions. Forcing the Seeda/Seby strategy into the June 3 box manufactures pressure to do something drastic. Decouple them (page 2).
- Option 3C is presented as a "pro." It is illegal phoenixing and a path to personal liability and reputational ruin. It is not an option (page 6).
- "Withhold both bonuses as leverage" conflates an owed debt with a disputed one. Doing that to the acknowledged A$18,902 invites a costs claim against you (page 6).
- It overstates the financial threat because it predates the Slack record. The disputed money is smaller and contingent, which removes the entire case for the radical paths (pages 3–4).
- It under-centres Vincent. He's the asset and the constraint; every option should be stress-tested against "does this keep Vincent?" (page 7).