Team Empathy · Internal strategy · 25 July 2026

Redesigning the business around what actually works

Built for the decision on the team

A client-by-client audit and a full financial redesign: what Team Empathy looks like if it is Ben, Nico, Om and Claude, running the clients that genuinely pay, while the new Shopify platform gets built. Version 2 — rebuilt from actual payments received in Xero.

The one-page verdict

Correction — v2, rebuilt from Xero

Version 1 of this document used the Financial Model spreadsheet for the client list and ClickUp task counts for delivery load. Both were wrong. The spreadsheet listed clients who have stopped paying (Modscape, Fulton Swim, Whisk), and 332 of the 374 "completed" ClickUp tasks were closed on a single day — 17 June — in a bulk board cleanup during Abi's handover. Those were historic onboarding tasks, not real work. Every hours figure derived from them was meaningless.

This version uses actual payments received in Xero as the source of truth for who is a client and what they pay. It tells a different and more urgent story.

−44%
Fall in the recurring base since March: $40,564 → $22,812 a month. Five clients churned, four reduced.
$0
What the business pays Ben. ACC has been covering his income — masking the real economics for over a year.
+$3,605
Monthly surplus on the lean model with zero new sales. About a $43k salary, from a standing start.
$8,895
New recurring needed to pay Ben the full $150k on the lean model. At today's cost base it would take $19,293.
The diagnosis, stated plainly

Team Empathy is not a struggling business, but it has a shrinking base wearing an overhead built for a bigger one. March to May cleared $29,633, $35,005 and $24,603 of net cash profit — all of it carried by Nature Baby's upfront payment and two one-off projects.

Underneath that, the recurring base has fallen every single month: $40,564 → $37,664 → $31,435 → $29,472 → $22,812. This is not a revenue-timing problem. It is quiet churn that the good months hid — and it makes the restructure more urgent, not less.

Where the base went

ClientMarchJulyChangeWhat happened
Moreton Bay Skip Bins$6,000$0−$6,000Fixed term finished (Ben confirmed)
Prime Plumbing$5,754$3,450−$2,304Reduced, still solid
KOOSHOO$2,100$0−$2,100Project-based, sporadic. Ripe for a retainer
Modscape$1,725$0−$1,725Stopped after May — Ben was right
Whisk Renovators$1,725$0−$1,725Stopped after April
Brandkit$1,460$0−$1,460Paid every month to June, then nothing. Check this one.
Eftpos Now (POSRite)$3,742$2,304−$1,438Reduced
Drop Bear Adventures$2,400$1,600−$800Reduced
The Haven on K'gari$600$400−$200Reduced
Net change  −$17,752Per month, in five months
Two things to check before Monday

1. Remarkable Auto & Marine has paid $1,725 every month including July. Ben believes they are no longer a client. Either the mental model is out of date, or they are paying for something they aren't receiving. That needs answering this week either way.

2. Brandkit paid $1,460 every month from March to June, then stopped in July. Silent churn or a payment issue — worth one email.

The finding that still makes the lean model work

Even after the churn, the low-touch base is real. Nine of the thirteen paying clients generated fewer than one tracked task a month each, and their dedicated Slack channels carry three to six messages across two months. A&M Detailing, Pink Bins, DCIS, Frontline Security, Will & Wind, Sinclair, Remarkable, The Haven and Drop Bear pay $11,392 a month between them for very little.

Ben's instinct holds: this base can be run by him plus Om plus Claude. It does not need a Head of Delivery.

The recommendation

DecisionCallWhy
Go leanYes Cuts ~$10,398/mo ($124,776/yr). Pays Ben ~$43k even with zero new sales, and needs only $8,895/mo of new recurring to fund the full $150k.
Arrest the churnUrgent The base has fallen 44% in five months. Cost cuts alone don't fix a shrinking top line — this is now the first priority, ahead of new sales.
The Head of Delivery seatDissolve The seat was designed for an agency being actively replaced by Claude skills and the Shopify platform. Nine of thirteen paying clients need almost nothing.
Timing on BrunaBy 10 Sep Decide before Italy on the 12th. Check the trial clause, proper notice, be generous. Right call, clean process.
Pay Ben nowImmediately ACC has been subsidising a structure the business can't carry. Being unpaid distorts every judgement.

Client-by-client audit

How this is measured now

Revenue = actual payments received in Xero, month by month, GST-inclusive as banked (ex-GST shown alongside). This is the only reliable answer to "who is a client".

Load uses two weak-but-real proxies: tracked ClickUp completions excluding the 17 June bulk close, and messages in each client's dedicated Slack channel since late May. Neither captures everything — a lot of Maps, content and reporting work never becomes a ClickUp task — so treat load as a relative signal, not a timesheet.

Toggl is not connected to the AIOS (no API key), so Bruna's time data could not be pulled directly. If you want true hours per client, that is the one integration worth adding.

Every client who has paid, March to July 2026

Straight from Xero. Sorted by what they paid in July — the live base.

ClientMarAprMayJunJulStatus
Prime Plumbing$5,754$5,754$3,450$3,450$3,450Live Reduced but steady
Growth Maintenance$3,450$3,450$3,450$3,450$3,450Live Rock steady, never missed
Eftpos Now (POSRite)$3,742$3,742$3,742$2,304$2,304Live Reduced in June
Eftpos Now$2,016$2,016$2,016$2,016$2,016Live Second Eftpos account
Sinclair Plumbing$1,725$1,725$1,725$1,725$1,725Live Never missed
Remarkable Auto & Marine$1,725$1,725$1,725$1,725$1,725Verify Ben thinks they've left — they're still paying
Drop Bear Adventures$2,400$2,400$2,400$1,600$1,600Live AUD, reduced
Frontline Security$1,384$1,384$1,384$1,384$1,384Live Genuinely low touch (2 tasks/120d)
Will & Wind$1,300$1,300$1,300$1,300$1,300Live AUD, steady
A&M Detailing$1,154$1,154$1,154$1,154$1,154Live Zero tracked tasks
DCIS (DC Installation Services)$1,154$1,154$1,154$1,154$1,154Live Genuinely low touch — v1 was wrong
Pink Bins$1,150$1,150$1,150$1,150$1,150Live Steady
The Haven on K'gari$600$600$600$400$400Live AUD, reduced
LIVE BASE (13 clients)    $22,812$19,837 ex-GST — the real floor
No longer paying
Moreton Bay Skip Bins$6,000$3,000$3,000$3,000$0Ended Fixed term complete
Brandkit$1,460$1,460$1,460$1,460$0Check Silent stop — chase it
Modscape$1,725$1,725$1,725$0$0Gone Ben was right
Whisk Renovators$1,725$1,725$0$0$0Gone
KOOSHOO$2,100$2,200$0$2,200$0Convert Project-based. Ripe for a retainer
Fulton Swim School$0$0$0$0$0Gone Ben was right — chatbot stopped
One-off projects (paid upfront)
Nature Baby$33,062$33,062Delivering The hero case study. Convert to retainer at the end
Clearwater Wildlife$11,500Finish Paid March, 6 tasks still stuck
HD Plumbing & Gas$6,900Complete One-off project
Roman Blinds DirectConvert 10 stuck tasks, no revenue. Finish and pitch a retainer

Load signal: tracked work per client

ClickUp completions in 120 days with the 17 June bulk close removed, plus Slack channel volume since late May. Across the entire client book there were 39 genuinely completed tasks in four months — about 10 a month for everything.

ClientReal tasks /moSlack msgs (2mo)Active nowRead
Nature Baby3.3257Correctly the busiest thing in the business
POSrite1.865Active build
Clearwater1.06Stuck — needs closing out
KOOSHOO0.8183High attention, no current revenue
Roman Blinds Direct0.031010 stuck tasks going nowhere
Prime Plumbing0.044$3,450/mo for very little tracked work
The Haven / Frontline / Eftpos / Drop Bear / Will & Wind0.5 ea3–5 ea0–2Genuinely low touch
Growth Maintenance / DCIS / A&M / Pink Bins / Sinclair / Remarkable0.00–3 ea0$11,933/mo combined, almost no tracked activity
What this actually proves

Even allowing generously for untracked work, the paying base is not what is consuming the team. Six clients paying $11,933 a month between them show essentially zero tracked activity, and their Slack channels are near-silent.

The real time sinks are Nature Baby (correctly — it's the case study), and a set of stuck, unpaid, unfinished projects: Roman Blinds (10 stuck), Clearwater (6 stuck), POSrite, KOOSHOO. That is a closing-things-out problem, not a headcount problem.

The MVP delivery standard

For the low-touch base, a defensible and honest monthly minimum: automated rank, traffic and conversion reporting; a monthly Enzo / Google Business Profile pass; one content or technical action; and a short written update. Claude drafts it, Om executes the technical piece, Ben signs off. Given the activity levels above, that is already close to what these clients receive — the difference is it would be systematic, automated and provable rather than ad hoc.

Money: where it actually goes, and the scenarios

The last five months, cash basis

Pulled live from Xero. Cash basis — money when it actually landed. Net profit is before any salary to Ben.

MonthIncomeCost of salesExpensesNet profitWhat happened
March$55,060$11,743$13,686+$29,633AirOps Transformation + Maps strong
April$64,261$13,212$16,049+$35,005Best month. Nature Baby upfront ($34,552)
May$57,123$14,388$18,136+$24,603Transformation continues + $5,750 commission paid
June$27,998$9,955$26,960−$8,911Upfront money gone. Wages spike to $16,983 (Abi payout) + $3,891 ads
July (to 25th)$20,940$14,754$11,060−$4,872No project revenue at all. Nico COGS $7,168
The honest read on June and July

June's loss was largely one-off. Strip the Abi payout (roughly $9,000 above a normal wage month) and June was close to break-even. That is not a business falling over.

July is the real signal. $20,940 of income is the business running on retainers alone with nothing sold on top — and it still loses money before paying Ben. That is the structural problem, and no amount of good months hides it.

Cash position

$108,727
Total across all accounts
$60,011
Profit account
$32,025
OPEX account
$24,290
Tax savings — owed to IRD, not spendable
~$77k
Genuinely free cash. About 5 months of lean runway at $16,232/mo.

Income (−$3,396) and GoCardless (−$4,202) are currently negative and net off the total.

Cost base: today versus lean

Today — normalised monthly

Wages & Salaries (Bruna, $90k)$7,500
Contractors [COGS] delivery (Om, Milda)$5,700
Service Providers — Nico$4,300
Ads (Meta)$3,000
Maps / Enzo subscriptions$1,300
Contractors [Admin]$1,200
Insurance, accounting, phone, office, vehicle$1,100
Admin subscriptions$950
Delivery subs (AirOps etc)$900
Sales subs (Ahrefs, SEMrush)$468
Banking / GoCardless / Stripe$150
Marketing subscriptions$62
Total before Ben's salary$26,630
With Ben on $150k$39,130

Lean — Ben + Nico + Om + Claude

Wages & Salaries$0
Contractors [COGS] delivery (Om-heavy)$4,000
Nico — new deal (USD $2k + equity)$3,470
Ads (Meta) — keep the lead flow$3,000
Maps / Enzo — keep, it earns$1,388
Contractors [Admin]$1,000
Admin subscriptions$1,000
Insurance, accounting, phone, office, vehicle$1,000
Delivery subs — AirOps out$600
Sales subs (Ahrefs, SEMrush)$468
Banking / GoCardless / Stripe$241
Marketing subscriptions$65
Total before Ben's salary$16,232
With Ben on $150k$28,732
$10,398
Monthly cost removed by going lean
$124,776
Annualised saving
$13,232
Lean cost if ads are paused too
$28,732
Lean break-even including Ben on $150k

The five scenarios

Revenue is the live July recurring base, ex-GST ($19,837). Costs as above.

ScenarioRevenue /moCosts /moBen paidNet /moVerdict
A. Status quo, no new sales$19,837$26,630$0−$6,793 Burns ~$82k/yr and still doesn't pay Ben. Not survivable.
B. Status quo, Ben on $150k$19,837$39,130$12,500−$19,293 Needs $19,293/mo of new recurring just to break even. Very hard.
C. Lean, no new sales$19,837$16,232~$3,605+$3,605 The floor. Even selling nothing, Ben draws ~$43k/yr. Today he draws $0.
D. Lean + $10k new recurring$29,837$28,732$12,500+$1,105 Full $150k salary and just profitable. Lou alone does this.
E. Lean + $20k new recurring$39,837$28,732$12,500+$11,105 The target. $150k salary plus ~$133k/yr retained to fund the platform build.
The single most important comparison

To pay Ben a full $150,000 salary, the business needs new recurring revenue of:

$19,293/month at today's cost structure   versus   $8,895/month on the lean model.

Ben's pipeline forecast is $10,000–$20,000 of new recurring. That forecast clears the lean bar and misses the current one. The restructure doesn't just cut cost — it converts a plausible pipeline into a paid founder.

The caveat that matters most

All five scenarios assume the recurring base holds at $19,837. It has fallen every month for five months. If churn continues at even half its recent rate, scenario C stops working by summer. Cutting cost buys time; it does not fix the top line. Arresting churn and converting the stalled projects (Roman Blinds, KOOSHOO, Nature Baby) is the first job, not the second.

Pipeline cross-check

Against the open deals already tracked, the $10–20k assumption is credible rather than optimistic:

DealExpected /moStatus
Lou — Ageless You$5,0006-month proposal, Ben rates it highly
Lou — restaurant brands$5,000Second proposal, same relationship
Roman Blinds Direct$3,000–$5,000Already active in delivery, 10 live tasks
Eden Orchards~$5,000Transformation with break-even guarantee
KOOSHOO~$2,000Conversion from project
Absolute Essential (Bo)TBCPitch incoming
Realistic landing$10k–$20k Only $8,895 is needed to fund the full $150k salary on the lean model

The lean design and the sequence

The framing that resolves it

"You're trying to scale the old way while building the new way." — Steven

That is the whole diagnosis. The team was hired to run an agency that is actively being replaced by Claude skills and the Shopify platform. Training someone into a dissolving role, using systems that don't exist yet, was never going to work — and that is a role-design problem, not a character problem.

Who does what

PersonSeatWhy it works
BenVision, relationships, sales, strategy, building in Claude His actual genius. Selling, shaping offers, and building the systems — not managing a delivery team.
NicoCTO / co-founder — the Shopify platform A true co-builder with equity, not an employee needing direction. USD $2k/mo baseline plus 20–30% of the software.
OmTechnical delivery across all clients An A-player technically, always wants more work, and per the Milda handover can absorb most of what Milda did.
Claude / the AIOSReporting, research, drafting, automation Already replacing the task-level work of Om, Milda and Abi via skills. This is the leverage that makes lean possible.
AlexOptional co-builder on the AIOS Self-taught, ships without hand-holding. The profile of the only kind of person worth adding right now.
The hiring principle, stated once so it holds

Right now Team Empathy needs co-builders, not employees. Anyone who has to be trained into a system that hasn't been built yet will fail, and it will look like their fault when it is really a sequencing error. Hire again only when there is a working system to hire into.

The sequence — Simon's model, in the right order

Now — Sep
Ben + Nico + Om + Claude
Build the Shopify platform. Nail Nature Baby. Hold the Gold base on automation. Convert the zero-revenue clients. Start paying Ben.
Step 2
Prove the ROI
Nature Baby plus the next five or six Shopify clients, each on the break-even guarantee. Three case studies with attributable revenue.
Step 3
Delivery owner
Only once there is a documented system and proven results to hire into. This is the seat Bruna was hired for — twelve months early.
Step 4
GM
Once delivery runs itself and revenue justifies it, bring in the GM to scale. Then Ben is doing only vision, relationships and product.

What gets kept, cut and automated

AreaDecisionDetail
Enzo / Maps softwareKeep$1,388/mo and it directly earns the Maps retainers. Clear positive margin.
Ahrefs + SEMrushKeepCore to research, audits and the reporting that proves ROI.
AirOpsExitUSD $3k/mo enterprise plan with a year lock-in, too complex for this market. Replaced by Nico's platform.
Meta adsKeep, watch$3,000/mo. It is producing the pipeline that makes the model work. Pausing saves $3k but starves the funnel.
Admin contractorsTrim$1,715 → ~$600. Most of this is absorbed by the AIOS.
ReportingAutomateThe single highest-leverage build: automated monthly client reports remove the main reason a delivery manager existed.

The next 14 days

#ActionWhy
1Put Ben on the payroll, even partiallyEvery decision is currently distorted by not being paid. ACC is masking the real economics.
2Walk Simon through this on MondayGet his read on the sequence. His own model — Ben + Nico + GM + delivery owner — is the destination; this is the order to reach it.
3Convert the stalled work: Roman Blinds, KOOSHOO, Clearwater21 stuck or active tasks earning $0 recurring. Converting these beats any cost cut — and arrests the churn.
4Close Lou — both proposals$10k/mo on its own clears the entire lean break-even including Ben's full salary.
5Build automated low-touch reportingThirteen clients, $19,837/mo ex-GST, on autopilot. This is what makes the lean base genuinely hands-off — and it doubles as the ROI-attribution engine.
6Reprice or release DCIS$70/hour is the worst economics in the book.
7Decide on the Head of Delivery seat by 10 SepBefore Italy on the 12th. Check the trial clause, proper notice, be generous. The seat is gone; the decision should still be clean.
The bottom line

Team Empathy has a genuinely good business hiding inside it: $11,933 a month from six clients showing almost no tracked activity at all. That base, plus Om, plus Claude, plus a founder who is finally paid, funds the build of the thing that actually matters — the Shopify platform and three ROI case studies.

The lean model doesn't shrink the ambition. It is the only version that can afford it.