Blue Monk
We still run the shop.
Blue Monk builds operations software for ecommerce businesses doing $1M–$50M. Ace runs Threadheads — $25M a year, in-house production, today. Everything we build gets used there before we sell it to you.
No deck. Bring your numbers and we’ll tell you what we’d fix first.
| Figure | What it measures | Source | Dated |
|---|---|---|---|
| $25M | Annual revenue of Threadheads, the business Ace runs day to day | Threadheads accounts, founder-stated | FY25 |
| 7-figure | Annual revenue of Heapsgood, sustainable packaging | Heapsgood accounts, founder-stated | FY25 |
What sits behind that claim
- $25M
- Annual revenue of Threadheads, the business Ace runs day to day
- Threadheads accounts, founder-stated · FY25
- 7-figure
- Annual revenue of Heapsgood, sustainable packaging
- Heapsgood accounts, founder-stated · FY25
The roster
The word doing the work is “still”.
Plenty of firms will tell you they came from operations. That is past tense. We are running an ecommerce business at scale this week, with the same problems you have, on the same platforms. It’s the one claim in this market a competitor can’t buy — they would have to go and build the business first.
- ThreadheadsApparel, in-house DTG/DTF · Australia
- HeapsgoodSustainable packaging · Australia
Two of these we own and operate. That is the whole public list — our other clients haven't signed off on being named, and we'd rather show two we can prove than a wall of logos we borrowed.
The problem
Your numbers disagree with each other.
Shopify says one thing. The 3PL says another. The ad platforms between them claim more revenue than the bank account received.
So someone capable spends a day or two of every week rebuilding the truth in a spreadsheet, and by Thursday it’s stale again. The variance between the tools is wider than most margins, so the number you are deciding on is wrong by more than the decision is worth. The figures, and where each one came from, are beside this.
None of this shows up as a line item. It shows up as a good operator doing data entry, and a decision made on Tuesday’s version of the truth.
| Figure | What it measures | Source | Dated |
|---|---|---|---|
| 10–15 hrs1 | A week spent rebuilding a single version of the truth in a spreadsheet | Polar Analytics | 2025 |
| 15–20%1 | Variance between what the tools in a typical stack each report as revenue | Polar Analytics | 2025 |
| 36%2 | Retailers who say they have accurate inventory visibility | Linnworks, State of Commerce Operations (n=200) | 2025 |
The numbers in that section, and where each one came from
- 10–15 hrs1
- A week spent rebuilding a single version of the truth in a spreadsheet
- Polar Analytics · 2025
- 15–20%1
- Variance between what the tools in a typical stack each report as revenue
- Polar Analytics · 2025
- 36%2
- Retailers who say they have accurate inventory visibility
- Linnworks, State of Commerce Operations (n=200) · 2025
- 1Polar Analytics, on reconciliation hours and cross-tool metric variance, 2025.
- 2Linnworks, State of Commerce Operations, 2025. Survey of 200 retailers.
Fig. 1Where the disagreement actually happens
Shopify
orders, revenue
- From Shopify — API
3PL / WMS
shipped units
- From Shopify — pixel
Ad platforms
attributed sales
- From 3PL / WMS — CSV exportFrom Ad platforms — does not reconcile
The spreadsheet
rebuilt weekly
Four systems, three of them right about something different. Nothing in the stack owns the reconciliation, so a person does it by hand — ten to fifteen hours a week — and the answer is stale by the time it is used.
Hours: Polar Analytics. Map: Blue Monk engagement notes · 2025–26
The thesis
AI made code cheap. It didn’t make judgement cheap.
Writing software costs less than it ever has. That is real, and it is not a reason to hire us. The median rate across the firms we surveyed was $37 an hour, and for a well-specified job that is a fine way to buy it.
What hasn’t got cheaper is deciding what to build. That decision needs someone who has stood on the floor the software is meant to serve, who knows which exception happens weekly and which happens twice a year, and who will still be there when the first version is wrong.
The build is the easy half. Getting the spec wrong is what costs $60,000.
95%
Three things you can hold us to
We’re still operating.
The people specifying your software are running an ecommerce business at scale today, not recalling one. Ace runs Threadheads — apparel, in-house DTG/DTF, multiple locations — and Heapsgood, sustainable packaging. Every system we sell runs in one of those first.
We only publish numbers we can source.
Every figure carries an attributed source and a date. Where we don’t have the number, the page says so. Engagements start with a paid diagnostic that sets a baseline before anyone commits to a build.
Fixed price. You own what we build.
Quoted as a number before we start. Subscriptions priced on the work, not order volume — triple your orders and the invoice doesn’t move. Code and data are yours, in the agreement. Klaviyo, Recharge, Loop and Gorgias each triggered merchant revolts through billing changes, not bad product.
What we’ve built
Four systems, all of them in production somewhere.
Not a capability list. These exist, they run every day, and most of them ran in our own business before they ran in anyone else’s.
A production ERP
Order intake, artwork, print queue and despatch in one system instead of four spreadsheets and a Slack channel. It runs the floor at Threadheads across multiple locations, every day, including the days it breaks.
3PL and fulfilment integration
Pick, pack and despatch wired to the warehouse so that Shopify, the production floor and the courier agree on what state an order is in. Most of the reconciliation work disappears because there is nothing left to reconcile.
A Shopify customiser
Product personalisation on the storefront that writes print-ready artwork straight into the production queue, rather than a designer re-cutting every file by hand after the order lands.
Internal agents
Small, boring automations. Matching order data between systems that disagree. Flagging artwork that will fail on press before it gets there. Drafting the reply a human still sends. None of it is a chatbot.
How we work
Three stages, and the price of each one is on this page.
Almost nobody in this market publishes a number. We do, because it saves both of us a call. Nearly everyone starts at the diagnostic.
| Stage | What you get | Price | Duration |
|---|---|---|---|
| Operations Diagnostic | Where your numbers disagree, what the gap is costing in hours, and what we would fix in what order. | A$5,000 – A$8,000 | 2–3 weeks |
| Fixed-Scope Build | A quoted number before we start. You own the code and the data from day one. | A$25,000 – A$60,000 | 6–12 weeks |
| Ongoing | A named owner for the systems we built. Priced on the work, not your order volume. | A$2,000 – A$8,000 / month | Rolling, 30 days' notice |
The offer ladder. Prices in Australian dollars, excluding GST.
- Operations Diagnostic
- PriceA$5,000 – A$8,000
- What you getWhere your numbers disagree, what the gap is costing in hours, and what we would fix in what order.
- Duration2–3 weeks
- Fixed-Scope Build
- PriceA$25,000 – A$60,000
- What you getA quoted number before we start. You own the code and the data from day one.
- Duration6–12 weeks
- Ongoing
- PriceA$2,000 – A$8,000 / month
- What you getA named owner for the systems we built. Priced on the work, not your order volume.
- DurationRolling, 30 days' notice
Fixed-scope work is quoted as a single number before anything starts. The ongoing subscription is priced on the work, not on your order volume.
Why the diagnostic is paid
A free audit is a sales document. It was always going to end in a proposal, which is exactly why nobody believes the findings. Charging for it makes it honest. Some clients take the document, fix the things themselves and never come back for the build — that is a good outcome, and we’d rather say so here than pretend otherwise.
Where each service sits on it
All four services enter at the same rung. What changes is what comes after the document — a narrow automation, a fixed-scope system, or a standing technical owner. The services index sorts that by what is actually happening in your operation.
The asymmetry
The cheapest hour of your week is on the operations side.
Put what you spend on advertising next to what you spend on software and development. For most brands in this range it isn’t close. Acquisition takes a double-digit share of revenue. The operational side takes a rounding error.
Nobody is suggesting you spend less on acquisition. But that account has been optimised weekly for years by people who are good at it. The operations side has had almost no attention and almost no budget. That is usually where the cheapest hour of the week is hiding.
- 1Eightx, “Ad spend as a share of revenue by stage”, 2026. Derived from SEC 10-K filings and 35 client engagements. Self-published by an agency serving this segment.
- 2Median of the five listed DTC and CPG brands, out of twenty sampled, that disclose R&D as a separate line. Mean 2.75%, range 0.13–8.43%. A small sample, and the only directly-filed figure we found.
Fig. 2Where an ecommerce brand's money goes
Median of the five brands, out of twenty sampled, that disclose R&D separately.
All bars share one scale. Full width represents 30% of revenue.
One scale, drawn to proportion. The advertising bars are published as ranges and are drawn as ranges — the hairline marks the lower bound.
Eightx; SEC 10-K filings · 2026 / FY25
Case studies
The long version, including what didn’t work.
Every case study carries a section on the parts that missed. A write-up with no misses in it reads as marketing, and gets discounted as marketing.
The questions you were going to ask anyway
Answered here, so the call can be about your business.
Eight objections, answered in the words we would use on a call. If yours isn’t here, it is a good first question to bring.
Ace runs Threadheads: $25M a year, apparel, in-house DTG/DTF production, multiple locations. He'll be on your call. Ask him something only an operator would know the answer to — that's a fair test and we'd rather you ran it.
Mostly we agree with the study. Pilots fail because nobody measured the baseline, so nobody can tell whether anything changed. We start with a paid diagnostic that sets that baseline before a line of code is written. If the numbers don't support a build, we say so and you keep the diagnostic.
You own the code and the data from day one — it's in the agreement, not a favour we grant on the way out. The systems run on Shopify, Supabase and standard infrastructure any competent developer can pick up. We'd rather you could leave and chose not to.
You can, and for some jobs you should. Cheap development is genuinely cheap now. What you can't buy at that price is someone deciding what to build who has run the operation you're describing. The build is the easy half. Getting the spec wrong is what costs $60,000.
Fixed-scope work is quoted as a number before we start. Subscriptions are priced on the work, not your order volume. You've watched Klaviyo, Recharge and Loop reprice on you. We're not building that into the contract.
Fair, and we won't pretend otherwise. We build on infrastructure your next developer will recognise, and the ongoing subscription exists so there's a named owner. If you'd rather buy something off the shelf, we'll tell you.
Because a free audit is a sales document. Ours is a paid piece of work with a deliverable: where your numbers disagree, what it's costing in hours, and what we'd fix in what order. Plenty of clients take it and fix things themselves.
Most of the work is asynchronous and the systems don't care where they're deployed. We hold overlapping hours with US and EU clients. If a project needs someone on your floor, we'll say so during scoping, not after you've signed.
Measure 12 of 12
Bring your numbers. We’ll tell you what we’d fix first.
If the honest answer is that you don’t need custom software, that is what you’ll hear, and it costs you half an hour to find out.
Ace takes the call. Brett does the technical scoping. Neither of us is a salesperson, because there isn’t one.
| Item | Detail |
|---|---|
| Length | Thirty minutes, on a video call |
| Cost | Free. There is nothing to buy on the call itself. |
| On the call | Ace, who runs Threadheads day to day |
| Bring | Your numbers. We do not present a deck. |
| After | A written diagnostic at A$5,000 – A$8,000, or the honest answer that you don't need one |
What the 30-minute call actually is
- Length
- DetailThirty minutes, on a video call
- Cost
- DetailFree. There is nothing to buy on the call itself.
- On the call
- DetailAce, who runs Threadheads day to day
- Bring
- DetailYour numbers. We do not present a deck.
- After
- DetailA written diagnostic at A$5,000 – A$8,000, or the honest answer that you don't need one