01 — Executive summary
A supplier the customer can finish buying from.
Concrete Online is a New Zealand ready-mix supplier built around one thing the incumbents do not do: let a customer configure a pour, see a fully itemised price, and complete a confirmed, paid booking without speaking to anyone.
The competitive research is unambiguous. Firth's orderconcrete.co.nz now returns an instant itemised price, so price visibility on its own is no longer the gap. What no competitor does is let a buyer finish — the total remains an estimate, colour and additives carry no visible price, and the delivery date is not real until someone rings back. We close that loop.
We start as a broker — routing orders to whichever supplier will take them, for a fee — and move to reselling once we have the volume and the trust to buy. The configurator, pricing engine, trade surface and staff back-office are already built and working. What we do not yet have is a signed supplier, current pricing, or a payment rail.
On the assumptions set in section 8, the business covers its fixed costs at — orders a week, and reaches — revenue by year five. Year one lands at — before founder pay. Those figures are computed live from the model below — move a slider and this paragraph moves with it, because a plan whose prose disagrees with its own model is worse than one with neither.
Every competitor stops at a quote; we take the order — and because we publish the price in plain text, we are also the supplier an AI assistant can actually name when someone asks who delivers concrete in Auckland tomorrow.
02 — The opportunity
The gap is completion, not pricing.
Buying concrete in New Zealand still runs on the phone. Every major supplier — Firth, Atlas, Allied, Holcim — advertises online ordering, and every one of them stops at a volume calculator or an unconfirmed request.
We captured Firth's flow in July 2026. It is the best in the market and it still ends with a request, not an order: the total is labelled an estimate, colour and additives are priced only by phone, the earliest offered delivery is three days out, and the time of day is decided later by someone calling you.
Three things a buyer cannot get today
- A price that is final. Small loads, out-of-area delivery, weekend surcharges, colour and additives all move the number, and none of them appear until the invoice.
- A delivery slot they chose. Not a date range someone confirms later — an actual window, picked and committed.
- A finished transaction. No competitor takes payment and returns a confirmed booking.
There is a fourth opportunity that is newer and less contested. Buyers increasingly ask an AI assistant rather than a search engine — "who delivers concrete in Auckland next-day with online pricing?" Being the answer requires published prices in plain readable text plus structured data. Incumbents hide prices behind forms, which makes them structurally hard to cite. That advantage is available now and will not stay available.
03 — What we sell
The mix, delivered — and the expertise to pick it.
Ready-mix concrete delivered across Auckland, ordered and paid for online. The product spans the full residential and light-commercial range, and the interface absorbs the complexity rather than passing it to the customer.
| Parameter | Range | Why it matters |
|---|---|---|
| Strength | 10–50 MPa | 20/25/30/40 cover almost all residential work |
| Aggregate size | 20 / 14 / 10 mm, plus 7 mm blockfill | Co-equal with strength — every supplier product is named by both. Smaller stone costs more and shrinks more |
| Finish | Grey, six oxide colours, four exposed-aggregate chips | A decorative chip replaces the coarse aggregate, so it sets the stone size too |
| Volume | 0.6 m³ minimum, 10% waste allowance applied | Exposed aggregate carries a 1.0 m³ minimum, except Maungaturoto |
| Delivery | Date, window, postcode, chute or line pump | Next-day across Auckland; surcharges shown before you commit |
Everything else the pour needs
Concrete is the middle of a sequence, not the whole job. A slab needs hardfill and formwork before it, steel and polythene laid, a pump on the day if the truck cannot reach, a crew to place and finish it, saw cuts within a day or two, and sealing later. Today the customer books every one of those separately and hopes they line up.
We know the pour date. That is the thing nobody else has. Once the concrete is booked, every other trade in the sequence can be offered against a date that is already fixed — and the offer is far more useful than a directory listing, because it is scheduled rather than searched.
| Before the pour | On the day | After |
|---|---|---|
| Hardfill and basecourse Formwork and boxing Reinforcing steel and mesh Polythene underlay Plate compactor hire |
Concrete pump Placing and finishing crew Exposed-aggregate washing |
Saw-cut control joints Grind, seal or polish Washout and waste removal |
Pump hire is the first of these and the easiest, because it is already an add-on inside a concrete order and Russel knows operators. The hire companies own the pumps and supply the operator, so we broker a complete service rather than buying and insuring equipment — the model stays asset-light.
The one worth taking seriously after that is the placing and finishing crew. It is the largest labour cost around a pour, the hardest thing for a homeowner to arrange, and the single most common reason a DIY slab goes wrong. It is also where a bad experience would land on our brand rather than a supplier's, so it needs more care than a pump does.
Each of these is a separate supply side to sign, a separate quality risk, and a separate way to be blamed. Do not add a second trade until the first one works. The pour date is the asset; it does not expire, and it will still be there once concrete is running properly.
The differentiator that is hardest to copy is not the checkout — it is the guidance. A homeowner does not know what MPa a driveway needs, let alone what aggregate size fits a 100 mm slab. Our mix helper answers both from the job description, explains why, and prices the difference on the spot. That is the piece incumbents are structurally bad at, because their business runs through a person on a phone.
04 — Market
Auckland first, and we can name our beachhead precisely.
Our launch market is the Auckland ready-mix catchment served by the depot network we can buy from — Takapuna, Silverdale, Kumeu, Wiri, Panmure and Warkworth, with Ruakaka and Brynderwyn covering Northland on separate terms.
Segments, in the order we intend to win them
| Segment | Job profile | Order size | Why they switch |
|---|---|---|---|
| Landscapers & gardeners | Repeat small-to-medium jobs | 2–6 m³ | Speed and self-service. Too small for incumbents to court, big enough to matter to us |
| Private / DIY | Paths, patios, driveways, slabs | 0.6–4 m³ | They do not know the spec and the phone call intimidates them. Highest margin per m³ |
| Small builders | Foundations, slabs, structural | 5–20 m³ | Account pricing and out-of-hours booking. Highest volume, thinnest margin |
Market size, and the direction it is moving
| Market | Volume | Trend |
|---|---|---|
| New Zealand, year to March 2026 | 3.66m m³ | −3.8% year on year, −11% from 2023 |
| Auckland, 2025 | 1,300,642 m³ | −4.5%/yr (1.43m in 2023 → 1.30m in 2025) |
| Auckland share of NZ | 35.7% | Higher than a population proxy would suggest |
Auckland is roughly 1.3 million m³ a year. Our year-five realistic case is under 17,000 m³ — about 1.3% of one region. The size of the prize is not the constraint; execution is.
Ready-mix volume is falling — 3.8% nationally in the year to March 2026, 4.5% a year in Auckland — and construction employment is down 6.8% from its peak. Every growth number in section 9 is therefore a share-taking number, not a rising-tide number. That is harder, and it is the honest framing. It also cuts both ways: in a contracting market, plants have idle capacity and more reason to want demand they did not have to sell for.
05 — Competition
Best-in-market still ends with a phone call.
We captured a real quote from Firth's ordering flow in July 2026 — 1 m³ of 20 MPa standard mix delivered to Papakura — and priced the identical job through our own engine.
Firth's rate includes delivery within a reasonable distance of the plant. Ours does not — we add $130 plus $7/km on top, and bill a 10% waste allowance on the volume. Comparing our bare rate to their delivered rate produced an earlier claim of "12% under Firth". Priced all-in, that reverses.
We reach parity at roughly 20 m³ — the top of the small-builder band, and outside both segments we intend to win first. We are cheapest exactly where we do not compete, and dearest where we do.
So price is not the wedge. Completion and guidance are. What a buyer cannot get anywhere else is a finished, itemised, confirmed order and an answer to "what mix do I need" — and neither depends on being cheapest:
| What the buyer meets | Firth | Concrete Online | Read |
|---|---|---|---|
| 20 MPa rate, ex GST | $347.00 delivered | $305.00 + delivery | Not comparable as published |
| All-in, 1 m³ delivered | $547 | $593 | We are $46 dearer |
| Small-load fee under 3 m³ | $200 flat | $80–90 tiered | ~$115 cheaper — but on a dearer base |
| Earliest delivery | 3 days | Next day | Dramatically faster |
| Delivery time | Phone call, later | Four pickable windows | We commit, they don't |
| Colour & additives | Phone, price unknown | Priced live on the page | We commit, they don't |
| Booking outcome | A request | Confirmed booking + deposit | The wedge |
The wider field
- Firth — the benchmark. Instant itemised pricing, strong volume calculator, national scale. Stops at an unconfirmed request.
- Atlas, Allied, Holcim — established plants and relationships. Ordering is a phone number.
- ConcreteDirect (US) — proves the fully-online model works at scale in another market. Not a competitor here, but evidence the model is not exotic.
Australia already has this business. MixHub matches postcode to supplier with published upfront pricing, and Found Concrete describes itself as "a broker, not a supplier", takes payment online, and is supplied by Boral, Hy-tec, Concrite, Advanced Readymix and Gunlake. Majors do wholesale to brokers. That is the single best answer to our highest-rated risk.
The uncomfortable truth is that our advantage is a flow, not a patent. Any incumbent could build this. Our bet is that they will be slow — because completing an order online cannibalises a sales desk, and that is an organisational decision, not an engineering one.
One quote, one suburb, one grade, one volume, one day. Three of our four named competitors have no captured data at all. Before this is shown outside the partnership we need all four majors mystery-shopped at three suburbs × three volumes, capturing the all-in delivered total, plus one real trade-account rate. About a day's work, and it replaces the weakest evidence in the plan.
06 — SWOT
An honest read of where we stand.
Strengths
- The product exists and works — configurator, pricing engine, trade surface, back-office, all built and demonstrable today
- Deep supplier-side knowledge of concrete, pricing structures and how plants actually operate
- Genuine cost advantage: buying at trade and selling under the incumbents' retail still leaves a spread
- Structured for AI answer engines from the start — plain-text prices and schema markup
- Almost no fixed cost. No trucks, no plant, no yard
Weaknesses
- No supplier agreement signed. Without one there is no business
- Every price is a four-year-old placeholder
- No payment rail, no dispatch automation, no shared database yet
- No brand and no trading history — a real obstacle with trade buyers
- Single-point technical dependency: one person can currently fix production
- Part-time founders against full-time incumbents
Opportunities
- AI answer engines are an uncontested acquisition channel today
- Trade accounts are badly served digitally — no incumbent offers self-service ordering on account
- The mix-helper expertise generalises: it is the moat, not the checkout
- Auckland proves the model, then NZ-wide, then Australia on the same engine
- Suppliers with idle capacity have every reason to want incremental demand they did not have to sell
Threats
- An incumbent ships a real checkout and the wedge closes
- A supplier learns the demand pattern and goes direct
- Capacity refused at peak, exactly when customers judge us
- Liability for a failed pour landing on us rather than the plant
- Trade terms consuming working capital faster than growth funds it
- Price war we cannot win against a vertically integrated producer
07 — Business model
Broker first, reseller later — the Webjet path.
Decided. We start as a broker: sign as many suppliers as will have us on non-exclusive terms, route each order to one of them, and take a fee. We do not yet have the trust — from suppliers or from customers — that reselling requires, and broking is far lighter on capital and liability.
The destination is reseller, the way Webjet moved from agency to inventory. Buying at trade and selling at our own published price is where the margin and the control are. But that comes after we can demonstrate volume, not before.
That order also matches the vision: a routing layer that any supplier plugs into and any buyer orders through. Broking is not a compromise on the way to reselling — it is the thing that builds the network the reselling model needs.
Built for many, launched with however many sign. We may well open with a single plant. The architecture does not assume that — suppliers are data, so adding one is a row rather than a rebuild. Independence from any single supplier is the reason to build it that way while there is still only one, not after the dependency has already formed.
It reverses the previous draft, which had us reselling from one or two suppliers and evolving to a marketplace. Broking first means more supplier conversations before launch, not fewer — the cascade is only real with several signed — and it means our revenue is a fee we negotiate rather than a spread we set.
How the money moves
Decided. The customer pays the supplier, not us. We invoice the supplier our fee. No bond.
- No bond. The existing terms of use add a refundable bond to every order. It would cover the cancellation and washout exposure, but it also puts an unexplained extra charge in front of a first-time buyer, which costs more conversions than it saves in claims. Cancellation is handled by the supplier's own policy instead.
- We never hold the money. No cashflow gap, no chargebacks, no bad debt, and GST as agent on a commission rather than as principal on the whole order. It is the lightest possible start.
- Our revenue becomes an invoice to the supplier. That is the trade: we swap payment risk for the risk that a supplier disputes attribution or simply stops paying. Which makes the supplier agreement — not the customer checkout — the document that protects the business.
"The customer pays the supplier" must not become "the customer is handed to the supplier to pay". The entire wedge is that a buyer can finish here — if we send them elsewhere to complete payment, we have rebuilt the handoff we criticise Firth for, just with a nicer front end.
The way to have both is a marketplace payment — Stripe Connect direct charges or the equivalent. The card is entered on our page, the checkout completes on our page, and the funds settle straight into the supplier's account with our fee split off automatically. We never touch the money and never hold a licence risk, and the customer never leaves. confirm that whichever supplier signs will accept a Connect account.
Revenue = concrete + delivery
Contribution = revenue − supplier cost − delivery − card fees
Delivery is passed through, so it sits in revenue and cancels out of contribution. Stating the second line as "revenue" — as an earlier draft did — is the kind of error a financially literate reader catches immediately and then discounts everything after.
How much spread do we actually need?
The retail uplift is the single most consequential number in the business. It decides whether the volume required to pay ourselves is achievable or fantasy.
The read is that our current 15% is conservative. We are pricing 12% under the market leader and still requiring nearly twice the volume that pricing at their level would need. There is room to move retail up without losing the price story — a decision worth taking deliberately rather than by default.
08 — Unit economics
What one typical order actually earns.
Every number in this section and the next is live. Move an assumption and the economics, the projections and all four charts below recalculate. Nothing is hardcoded — this is the model, not a picture of it.
Assumptions
Payment processing is the only meaningful variable cost beyond the concrete itself, at Stripe's New Zealand rate of 2.65% plus 30c. Delivery is passed through at cost, so it inflates revenue without adding margin — which is why contribution as a percentage of revenue looks thinner than the concrete margin actually is.
1 — The GST basis contradicts itself. The supplier sheet is ex-GST. The demos label the retail price "GST incl" while the written plan treats the same figure as ex-GST. Both cannot be true, and if the demos are right then the entire uplift is consumed by GST and we are selling at trade cost. Every figure on this page is ex GST.
2 — The supplier cost is an assumption, not a fact. The written plan used $275.50/m³, which is the 25 MPa trade sell price — wrong grade and wrong column. The $240.50 default is better sourced but still only what Atlas quoted a cash account in 2021. We have bought nothing. Worse, it sits $25 under a trade rate our own target customers already hold — a landscaper with an Atlas account buys at $265.50 direct, so our retail is a 15% markup on a price they have. One conversation with a supplier settles this, and nothing else in the model matters more.
3 — Every price is from 2021. Four years of cement inflation sit between these figures and reality. The sliders exist so that replacing them is a two-minute job, not a rewrite.
09 — Projections
Growth is driven by orders, not by price.
Growth is driven by orders, not by price. Set the first-year order count and the growth rate you believe, and the five-year picture follows.
Growth assumptions
| Projection | Year 1 | Year 2 | Year 3 | Year 4 | Year 5 |
|---|
Where the break-even actually sits
The business is viable at genuinely small volumes. The risk is not that it fails outright — it is that it settles at a size that does not pay the people running it.
10 — Funding
The requirement is small, and mostly working capital.
Fixed costs are near zero — no plant, no trucks, no yard, no premises. The build is done and self-funded. The requirement is the year-one operating shortfall plus the float between paying suppliers and being paid.
| Requirement | Amount | Notes |
|---|---|---|
| Year 1 operating result | — | Live from the model. Before any founder pay |
| Supplier float (private) | Low | Card payment clears before or on delivery |
| Supplier float (trade on terms) | unmodelled | We pay the supplier before a 20th-of-month customer pays us. Grows with revenue |
| Incorporation, legal, insurance | to quote | Shareholders' agreement, public liability, professional indemnity |
Trade working capital is the single most likely way a profitable version of this business runs out of money. Every trade account we win extends the gap between paying the plant and being paid on the 20th. The financial model does not yet include it, and it must before we offer terms to anyone.
11 — Operations
Cascade to the cheapest supplier who can take it.
- Order placed → Stripe authorisation hold on the deposit, not captured.
- Notify the cheapest covering supplier with a one-tap Accept link.
- Accept → capture the deposit → SMS the customer → release the others.
- Declined or no answer within the window → auto-forward to the next cheapest.
- Nobody accepts → human fallback: call round, or offer the next slot.
- First order from a new customer: we phone to confirm. We text on every order, and call whenever something about the site is unclear.
Nothing is charged until an order is real, which is what allows us to promise a confirmed booking without holding inventory. Later, routing should score suppliers on reliability — acceptance rate and on-time record — not price alone.
The dispatch cascade is designed, not built. It is the most operationally significant part of the model and currently exists only in this document. Until it is built, confirmations depend on a person checking with the plant — workable at a handful of orders a day, not past that.
Capacity
Whether a confirmation is genuinely true depends on whether suppliers pre-commit capacity. Pre-guaranteed loads make the promise real immediately; without them, the cascade leaves a window where a customer believes they are booked and is not. This is the most important thing to negotiate in the first supplier agreement.
12 — Go-to-market
Be the answer, then be the habit.
- Answer-engine optimisation first. Published per-m³ prices in plain text, schema.org Product and Offer markup, an
llms.txt, and written answers to the questions people currently ring to ask. This is cheap, compounding, and uncontested. - Search, on the specific questions. "How much concrete for a 4x6 driveway", "what MPa for a house slab", "what aggregate size for a 100mm slab" — our mix helper already answers all three better than anything ranking today.
- Trade by relationship. Landscapers and small builders switch on a recommendation, not an ad. This is direct outreach, not a campaign.
- Word of mouth on the small jobs nobody wants. A 1 m³ pour costs $140 less with us than with Firth. That is a story a customer repeats.
open Paid acquisition is unproven for us and unbudgeted. Partnerships with hire yards and trade merchants are plausible but unexplored.
13 — Team
Three founders, and one of them is not in the operation.
The venture needs things that rarely sit in one person: a working product, the supplier relationships to put concrete behind it, and someone running the day to day. Three of us, split across those.
| Part of the business | Share | Covers |
|---|---|---|
| Product & technology | 25% | The site, the pricing engine, automation and dispatch — everything the customer touches. No operational role and no capital contribution. |
| Supply & operations | 37.5% | Supplier relationships and negotiation, concrete expertise, pricing input, plant-side operations |
| Operations & customers | 37.5% | Day-to-day running, the phone and the inbox, orders and customer service |
No owner draws a salary. Everyone takes their share of clean profit instead, which means the split above is the compensation — there is no second lever. It also means nobody is paid at all until the business is profitable, so the year-one line in section 9 is the number that matters most to the people in this table.
ConcreteOnline Limited already exists — the customer terms of use were drafted for it around four years ago and are still published. That is an advantage: the entity, the domain, the brand and a professionally drafted broker agreement are all in place.
It is also a different transaction. Taking shares in an existing company means inheriting its history — any debts, unpaid tax, GST position, prior contracts, guarantees or dormant claims come with the shares. Before any name is added to the register: confirm it is currently registered and in good standing, get the financial position, and list every obligation it already carries. If the history is untidy, incorporating fresh and moving the assets across is cleaner than joining and hoping.
The split is agreed. What is not agreed is who carries legal responsibility. Shareholding and directorship are separate things — a 25% shareholder with no operational role risks their shares, but a director carries personal statutory duties under the Companies Act 1993 whatever their shareholding, and an "officer" carries duties under the Health and Safety at Work Act that cannot be signed away. Personal guarantees on a supplier account are a third, separate exposure.
"No responsibility for operations" has to be structured — it does not follow automatically from holding a quarter of the shares. This needs a lawyer before incorporation, not after.
to agree Who is a director, who signs supplier guarantees, how decisions get made when two shareholders hold 75% between them, and who owns the code and brand. These are worked through in the decision session and should end up in a shareholders' agreement drafted properly.
14 — Risks
What kills this, and the answer.
| Risk | Likelihood | Mitigation |
|---|---|---|
| No supplier will deal with us, or they go direct | High | Start with committed partners; demonstrate we bring demand they did not have to sell for |
| Nobody accepts an order we have confirmed | High | Authorisation hold rather than a charge, human fallback, and negotiated SLAs |
| Trade credit consumes working capital | Medium | Credit checks and limits; earn-in to terms rather than terms on day one |
| An incumbent ships a real checkout | Medium | Move fast, own the answer-engine channel, and win on service and guidance rather than flow alone |
| Liability for a failed pour | Medium | Contracts placing certification and liability with the supplier; our own insurance |
| Price spread collapses across suppliers | Low | Cascade by cheapest, plus a retail buffer |
| Key-person dependency on the build | High | Currently unmitigated. Documentation and a second engineer are the year-two answer |
15 — Roadmap
Six steps to a soft launch.
- Now — the product existsConfigurator, pricing engine, mix guidance, trade surface and staff back-office, all built and demonstrable.
- Settle the plan and the partnershipWork the open decisions, agree the split, do due diligence on the existing company before joining it, and correct the pricing basis.
- Sign one supplierRates, coverage, SLA, liability and capacity. Nothing else matters until this exists.
- Wire the money and the dispatchStripe authorisation and capture, the supplier cascade, SMS and email confirmations.
- Real backendShared database, slot handling, staff authentication.
- Soft launch AucklandPrivate and landscaper segments, then add suppliers and trade accounts.
16 — What we decided
The answers, and the three that need another pass.
These are set out one at a time — with real options and their consequences — in the decision session. Anything already answered there appears below automatically.
1 — Trade terms and the float answer are reconciled. They read as a contradiction and are not. Nothing is committed before payment for new customers and DIY — the default. 20th-of-the-month terms are earned, and only by trade customers who have placed and taken five to ten orders cleanly. So the float only ever applies to a known account with a track record, which is a much smaller and better-understood exposure than terms offered on day one. It still needs modelling — see section 10 — but it is a deliberate, bounded risk rather than an open one.
2 — DIY pricing removes our margin unless the fee replaces it. The pricing answer is that DIY pays "the same price as they would pay at the supplier". That is right for a broker and it kills the retail uplift — so 100% of our revenue becomes the fee the supplier pays us, which is a number nobody has negotiated yet. Section 8 now models it that way. If no supplier will pay a fee, there is no broker business, only a reseller one.
IP: "the company owns everything". Reasonable, and what any future investor expects. But it assigns the pricing engine, the configurator and the mix-guidance work to this company outright — so none of it can be reused on another venture. If the intention is for this to be one project among several, that wants an explicit licence-back for the underlying tooling, kept separate from the Concrete Online brand, data and customer list. Easy to agree now; expensive to unpick later.
Answers live in this browser — move them to another one
The session saves to the browser it was filled in on, so a decision made on one laptop will not appear on another. To move them: open the session, press Copy as text, then paste it here.