Custom ecommerce build: a store that fits your rules

A custom ecommerce build is for the store whose rules no standard model holds: a price list per client, orders sold by the case, or a configurator that changes the total. Discovery is one week at $2,500, the Build starts at $8,000, and the plan stays yours whether we build it or not.

What this includes

Every custom ecommerce build we run covers the same ground, whatever the rules turn out to be:

• The rules your store runs on, written in plain words before anything is bought
• The smallest build that carries those rules, sitting next to the parts that already work
• Price lists that differ by client, held in the store instead of in somebody else's spreadsheet
• Orders sold by the case, the pallet or the standing order, with freight and minimums counted
• A configurator that shows a price the moment the buyer changes an option
• Stock and account terms that stay right when two people buy the last item
• A handover: the accounts, the files and a short guide for whoever comes next

The storefront that comes with it behaves the same on a phone, on a laptop and on the tablet sitting on your counter, because your buyers use all three, often inside the same order. Behind it, stock, prices and account terms move together, so the total one buyer sees is the total the next one sees.

The work sits inside the headless commerce practice we run from Miami, and it usually replaces the shop you have, which is what moving a live store involves.

We do not sell a platform, and we take no commission on your sales. What you pay for is the store, and the store is yours: the pages, the catalogue, the price lists and the rules the whole thing runs on. If repairing what you already own is enough, that is exactly what the Discovery week is for, and it is the cheapest week we sell.

The three rules no standard store holds

Two things send a shop down this road: the platform cannot express a rule you sell on, or it can, at the price of a workaround your team redoes every week. Three rules come back most often.

A price list per client. Wholesale accounts, hotels, restaurants, resellers and your own staff buy from the same catalogue at different prices. The price should follow the account that logs in, the quantity bracket and the terms you agreed with each one. On a store built for a single public price, that list ends up in a spreadsheet, and somebody rechecks every order before it ships. The spreadsheet is not the problem: the second check is, and it lands on the day you are busiest.

Orders sold by the case, not by the unit. A shop that sells to restaurants sells a carton of twelve, a half pallet, or a standing order that repeats every Tuesday. The store has to count stock in cases, apply a minimum order, add the freight, and hold an account whose terms are not current. A catalogue that only knows single units is exactly where those orders go wrong, and the mistake is usually found in the warehouse, after the delivery has left.

A configurator that prices as the buyer chooses. Made-to-measure, made-to-order or assembled from options: the buyer picks a size, a finish, a delivery window, and the total has to move with the choice. What we build records the choice with the order, so whoever makes the product reads what was actually sold instead of chasing an email thread.

None of this is exotic. It is the ordinary way a business sells, and it is the part a template cannot be talked into. Combining parts instead of buying one large system is what people call composable commerce: what the MACH Alliance calls composable sets out the idea without the sales talk. Most of our builds start from the store engine we build on most often, which is open source, so your next hire can read it and nobody can switch it off or raise its rent.

The quiet cost of leaving all this alone is time: the hours spent correcting prices by hand, the order that ships wrong, the buyer who leaves when the last item goes twice. Add those up over a month and you have the number to compare against the Build. If it is small next to the build, we say so, and if your rules are ordinary, a shop off the shelf is cheaper, and we say that too.

How it works

1. Discovery, 1 week, $2,500. We follow one order from the buyer's screen to the warehouse and write down every rule that breaks the store you are on: who pays what, how the cases are counted, which options change the price. You keep that document whether or not you build.
2. Architecture, 1 week, $2,500, credited to the Build. What we would build, what we would leave alone, and a price for each step. If repairing the store you own is cheaper, we say so in writing and stop there.
3. Build, two to ten weeks depending on scope, from $8,000. Work is released every week on a test address you can open any Friday. You watch the store grow instead of waiting for a launch date, and a change sent back early costs less than one sent back at the end.
4. QA and launch, 1 week, included. We move the shop on a date agreed in advance, planned outside your busiest weeks, and the addresses that bring you customers today keep working afterwards.
5. Care, from $1,500 a month, only if you want it. Upgrades, small changes and a monthly look at speed and orders, with 30 days notice to stop. Nothing renews by itself, and the store is already yours.

Speed is what your customers feel, so we report what Google measures on a phone, on an ordinary connection, before anything goes live. A page that opens in about a second on a phone is the target, and a store we rebuild is measured before and after, in the same conditions both times.

The rhythm matters as much as the sequence: a real page every week on a test address, with the change you asked for already in it. Whoever answers your questions during Discovery is the person who draws the architecture, and the person who reads the code before it goes out, so the project is not handed to a stranger at the end of the month.

What it costs

The prices below are where each phase starts. The scoped quote fixes the exact scope, the price and the deadline, and it comes after the Discovery week, once we know what your rules really are.

• Discovery, one week: $2,500, and the verdict is yours to keep.
• Architecture, one week: $2,500, credited to the Build when you go ahead.
• Build, two to ten weeks: from $8,000, priced step by step, so you pay for finished work.
• QA and launch: included in the Build.
• Care, monthly: from $1,500, with 30 days notice.

Three things move the number: how many price lists you run, how many order sizes and minimums you sell, and how many options the configurator has to price. We tell you which one costs what before you commit, and how each phase is priced is published on our process page, so you can budget before you speak to anyone.

A custom ecommerce build is worth paying for when the workaround costs more than the build. That is the comparison we put in writing: what the manual checks cost you in a month, against a build priced step by step. If the two numbers are close, we say so and you keep your money.

What is not in these numbers: your own time, about two half-days of decisions in total, and any subscription your business already pays for. We write down what you keep paying and what stops, before the Build begins.

Proof

Food By Box ran 2,500 products through the whole sequence. Its Google speed rating on a desktop reached 95, monthly visits grew from 20,000 to 500,000 in 24 months, and no address that brought traffic was dropped along the way. You can see the numbers behind those projects, published one by one with the duration of each piece of work.

Tatiana Miath's store is the other half of the same story. On a phone, a page that took just under five seconds now appears in barely over one, and nothing she ranked for was lost.

Neither store was rebuilt for the sake of it. Both kept the parts that worked, changed the parts that did not, and stayed open while the work happened, which is why their addresses survived the move.

Michael Agnassia draws the architecture and reads the code that goes out, after 15 years in ecommerce. who writes the plan and reads the code is on our about page. The studio is at 78 SW 7th Street in Brickell, Miami, and we work on site across South Florida, in English and Spanish.

You can repeat the checks we publish on your own store: the speed rating, the visits month by month, the addresses that bring buyers. If a number here does not hold up on your shop, that is worth knowing before we start.

Questions

How do we know a custom ecommerce build is worth the money?

Count what the workaround costs today: the hours spent correcting prices by hand, the orders that ship wrong, the buyers who leave when the last item goes twice. If that number is small next to the Build, we tell you in writing and recommend the repair instead.

What happens to our shop while you build it?

Nothing visible. It keeps selling, we build beside it on a test address, and we move it on one agreed date. Your customers notice the speed, not the work.

Can a headless commerce setup keep the back office we already use?

In most cases, yes. Accounting, warehouse and supplier tools stay where they are: the store sends them the same order in the shape they already expect, and we test that link with a real order before launch.

Do we need a custom storefront, or can we keep the shop we have?

Often you keep it. A fair part of what we build sits behind the part customers see, and only the pages that must behave differently get replaced.

How long before we can click on something?

In the first weeks, on a test address: a working store you can open and try every Friday. The shop you run today stays untouched until the switchover.

What if our rules change after launch?

A new price bracket, a new account type, a second language: the store is yours to change, so it is a small job, priced before it starts and scheduled around your selling season.

What do you need from us to start?

One order, from the buyer's side to the warehouse, and the people who handle it today. The Discovery week runs on your real orders, not on a list of wishes, and it ends with a written plan you can act on even if you never build with us.

Next

The plan is yours whether or not we build it. When the work around each order should run itself, that is our automation practice, where quotes, invoices and follow-ups stop being retyped. Book a call and describe the rule no standard store will hold.