Interactive controls are loading. Phone and email links are available.

Skip to main content
For businesses that build it after the order arrives

Made to Order and Configurator Integration

Blinds, cabinetry, signage, windows, custom furniture, engineered componentry, embroidered uniforms. When the product does not exist until someone orders it, there is no SKU to look up, no stock to check and no shelf price. There is a set of choices, a rule that turns them into a price, and a factory that has to be able to make the result.

This page is about the join between the two. How an option tree stays aligned with what the workshop can actually produce, where pricing rules should live so they are never maintained twice, how to promise a date based on real capacity rather than a hopeful default, and how deposits, drawings, approvals and change orders flow without a spreadsheet in the middle.

What Is Actually Being Sold

No SKU
The product does not exist yet
So the order has to carry a full specification, not a code
1 pricing rule
Held once, used everywhere
A formula duplicated in the website and the ERP will diverge within months
Capacity, not defaults
The only honest way to promise a date
A fixed lead time is a guess that the workshop pays for
Two payments
Deposit then balance, in most cases
Which makes the accounting treatment a design question, not an afterthought

Four Things That Make Made to Order Different

Every one of these breaks an assumption that ordinary retail integrations are built on. Ignore them and the configurator will sell things the factory has to ring the customer about.

The option tree has to match the workshop

A configurator that offers a combination the factory cannot build is worse than no configurator, because it converts a browsing customer into an awkward phone call and a refund. Options need dependencies and exclusions expressed as rules: this fabric is not available above a certain width, that hinge will not take that door weight, this finish adds a week. Those rules belong in one place and flow to the website, the quoting tool and the sales team, rather than living in a production manager’s judgement.

Pricing is a formula, not a price list

Made to order pricing is usually calculated from dimensions, materials, componentry, labour and sometimes installation, with minimum charges and rounding rules on top. The moment that formula exists in both the website and the back office, they begin to drift, and the difference shows up as margin you cannot explain. Hold the calculation once, expose it to every channel that needs a number, and keep a record of the inputs so any historical price can be reproduced when a customer queries it.

Lead time comes from capacity

Publishing a flat lead time is comfortable until the workshop is three weeks deep and the website is still promising ten days. A real date is calculated from the current production queue, the hours a job of that type consumes, material availability for anything imported, and a calendar that knows about state public holidays and your Christmas shutdown. Businesses that get this right convert better, because a specific believable date beats an optimistic vague one.

Nothing is made until it is approved

Custom work needs a gate. A drawing, a proof or a specification summary goes to the customer, the customer approves it, and only then does the job release to production. That approval has to be recorded with a timestamp and a version, because the difference between a variation you can charge for and a remake you wear is usually a question of what was approved and when. Integration makes this fast rather than making it optional.

Six Connections Between Selling and Making

The order in which these are built matters. Pricing and validation first, because everything else depends on the specification being right.

Valid spec

Configuration capture

Every option, dimension and note captured as structured data, validated against the rules before the customer can pay, and stored with the order in a form that can be replayed later. When a customer rings in eight months asking for the same again, or a warranty claim arrives, you can reproduce exactly what was specified rather than reading a free text field someone typed at the time.

One number

Price calculation

Dimensions, materials, componentry, labour, freight and install resolved through a single rule set, with minimum charges, rounding and any account specific pricing applied consistently. The same service answers the website, the quoting tool and the sales team, so the price a customer sees online is the price the office would have quoted, which is not a given in this category.

Job ready

Bill of materials generation

Options translated into the components, quantities, cut sizes and operations the production system needs, rather than a paragraph of description that someone in the workshop interprets. This is the piece that removes the daily rekeying, and it is also where make or buy decisions can be automated, so bought in components generate a purchase requisition at the same time as the job is created.

Real date

Capacity and promising

A promise date calculated from the queue, the job type, material lead times and a working calendar that includes state public holidays and shutdown periods. It is published at the point of order and updated if something material changes, with the customer told rather than left to discover it when the delivery does not arrive.

Signed off

Approvals and change orders

Drawings and proofs issued, approvals captured with a version and a timestamp, and any later change assessed against production status so a change before cutting is free and a change after it is quoted. The record of what was approved is the thing that resolves the argument, and it should be automatic rather than a folder somebody maintains.

Cash flow clean

Deposits and final payment

Deposit taken at order, balance requested at the right trigger, and both reflected properly in the accounting system so the job, the invoice and the bank all agree. For businesses with long lead times this is the difference between funding production from customer deposits and funding it from an overdraft.

What Changes Between Order and Delivery

TaskTraditionalConnected to ProductionNotes
A customer configures onlineEnquiry form and a callbackPriced and validated liveThe rules stop impossible combinations before the money changes hands.
Turning options into a jobRekeyed into the production systemBill of materials generatedRemoves the transcription error that produces a remake at your cost.
Promising a dateA standard lead time for everythingCalculated from the queueState public holidays and shutdown weeks belong in the calendar, not in someone’s head.
Bought in componentsOrdered when the job reaches the benchRequisitioned at orderImported componentry is usually the real constraint on the promise date.
Drawing approvalEmailed and chasedIssued, tracked, recordedThe approval record decides whether a later change is a variation or a remake.
Customer changes their mindVerbal, sometimes after cuttingAssessed against job statusChange before release costs nothing, change after it is a quote, and the system knows which.
Deposit and balanceInvoiced manually, chased manuallyTriggered by job milestonesWorth confirming the GST treatment of deposits with your accountant and building to match.
Job profitabilityKnown at year end, roughlyActuals against the quoteMade to order margin varies per job, so averages conceal the ones losing money.

Where Made to Order Projects Go Wrong

A configurator the factory cannot honour

The most expensive failure in this category, because it happens after payment. Every constraint the workshop applies in its head needs to become a rule: maximum spans, weight limits per fitting, finishes unavailable on certain substrates, minimum sizes, combinations that void a supplier warranty. Get the production lead to try to break the configurator before it goes live, and treat every impossible order that gets through afterwards as a rule that needs writing rather than a one off.

Pricing logic in two places

The website has a formula, the office spreadsheet has a slightly different one, and both are edited by different people. Six months later nobody can say which is right, and margin analysis becomes guesswork. Hold the calculation once, call it from everywhere, version it, and store the inputs and the resulting price with every order so a historical quote can be reproduced exactly. If a rule has to change mid season, effective dating keeps existing quotes honourable.

Lead times that ignore the calendar

Australian production calendars have quirks that generic systems miss: public holidays that differ by state, the summer shutdown that closes suppliers as well as your own workshop, and imported material lead times that stretch around Lunar New Year. A promise engine that does not know about these will confidently commit to dates in the middle of a fortnight when nobody is at the bench. Load the calendar properly, review it annually, and make shutdown dates a scheduled task rather than something remembered in December.

No stored version of what was ordered

If the specification exists only as rendered text on an invoice, you cannot reproduce it, cannot analyse which options actually sell, and cannot defend a warranty claim. Store the structured configuration alongside the human readable summary, keep the version of the rules that priced it, and make both retrievable from the order. This costs almost nothing at build time and is very difficult to add retrospectively.

Consumer law treated as a returns policy

Custom made goods cannot generally be returned for change of mind, and it is reasonable to say so clearly. What does not change is that consumer guarantees under Australian Consumer Law still apply: the goods must be of acceptable quality, fit for the purpose you were told about, and match the description or sample. A deposit that is genuinely part payment is treated differently from a security deposit for GST purposes, so agree the treatment with your accountant and build the system to match rather than deciding it invoice by invoice.

Building a full quoting platform when you needed a form

Not every made to order business needs configure, price and quote software. If you sell a handful of configurable products with a few options each, a well built form on the site, a pricing service and a clean handoff into your existing job system will do the work at a fraction of the cost and complexity. The case for something heavier appears when option counts run into the hundreds, when dealers and reps quote on your behalf, or when a single job carries dozens of configured line items. We will tell you which side of that line you are on.

How Yes AI Approaches Made to Order Work

Time in the workshop, not just the office

We watch how a job actually gets made, what the bench needs on the docket, and which orders cause a phone call back to the customer. Those phone calls are the specification for the validation rules, and you cannot find them from a process diagram.

Right sized recommendations

Sometimes the answer is a configurator, a pricing service and an integration. Sometimes it is a better form and one connection into the job system you already own. We will tell you when the smaller project is the right one, including when your production software already does most of this.

Built, hosted and monitored by us

The integration runs on a managed cloud automation layer we operate, with record level logging and same day alerting. When a supplier changes a component code or a price file format, we notice before your quotes start coming out wrong.

Rules documented and owned by you

Option constraints, pricing formulas, capacity assumptions and calendar rules are written down in language your production and finance leads can check and change. This is the knowledge that usually lives with one long serving employee, and it should not.

From Phone Quotes to Connected Production

Five steps. A first configurable product priced and flowing into production is usually live in six to ten weeks.

Map one product properly

Every option, every constraint, the pricing formula, the components it consumes and the operations it takes. One product done thoroughly teaches more than five done at a glance.

Agree ownership of rules and price

Where option rules live, where the pricing formula lives, who may change them, and how changes are versioned. This is the decision that keeps the system honest in year two.

Design the production handoff

What the job needs, how the bill of materials is generated, how purchases for bought in components are triggered, and how the promise date is calculated. Written in plain English and checked by the workshop.

Build and pilot on one line

Live on one product family with the manual path still available, real orders and daily review. We look hardest at the orders that fail validation, because they are telling you about a missing rule.

Extend and measure margin

More products, then approvals, change orders and payment milestones. Job level actuals against quote reported from the start so you can see which configurations are worth selling.

FAQ

Do we need configure price quote software, or can our website do it?

It depends on how much variation you sell and who does the selling. A business with a handful of configurable products, each with a few options and a formula based on dimensions, can usually run a well built form on its own site with a pricing service behind it and a clean handoff to the job system. Dedicated quoting software earns its keep when option counts get large, when dealers or reps quote on your behalf and need controlled pricing, when a single job carries many configured lines, or when approval workflows and revisions are constant. We will look at your actual product range and give you a straight answer rather than defaulting to the bigger build.

How does a configurator produce a bill of materials?

By mapping options to components and operations through rules rather than through description text. A width and drop become cut sizes and a fabric quantity with an allowance, a chosen fitting becomes a specific part number and quantity, a finish becomes an operation with a time and possibly an extra day of lead. Where a component is bought in rather than held, the same rule can raise a purchase requisition so the material is ordered the day the job is taken. The important discipline is that every option must resolve to something the production system understands, which is exactly the exercise that surfaces the constraints nobody had written down.

How do we promise a delivery date we can actually meet?

Calculate it rather than publish it. The inputs are the current production queue, how much workshop time this job type consumes, availability of any material you do not hold, and a working calendar that knows about weekends, the public holidays for your state, and your shutdown period. Then add a buffer you have chosen deliberately rather than one that crept in. Two practical notes: publish a date range rather than a single day if your variability is genuinely high, and tell the customer as soon as the date moves. Customers forgive a changed date far more readily than a silent one.

Can we take a deposit online and the balance later?

Yes, and it is the normal pattern for anything with a lead time. The design questions are what triggers the balance request, whether it is before dispatch, on completion or after installation, and how both payments are represented in your accounting system so the job, the invoice and the bank reconcile. There is also a tax treatment question, because a deposit that is genuinely part payment is handled differently from a security deposit for GST purposes. That is a conversation to have with your accountant, and once the answer is settled we build the flow to match it rather than making it up per order.

What happens when a customer wants to change their order?

The system needs to know where the job is. A change before the job is released to production should be simple and free. A change after materials are cut or a finish applied is a variation with a cost, and it needs to be quoted and approved rather than absorbed. Integration makes this workable by keeping job status visible to whoever answers the phone, recording the original approval with a version and a timestamp, and generating the variation as a document rather than a note. Most disputes in this category come down to what was approved and when, which is precisely what an integrated approval record settles.

Our production system is old and has no API. Is this still possible?

Usually yes, and it changes the pattern rather than ruling the project out. Older production and job systems commonly offer a database you can read, a scheduled import folder, an emailed report or a file export, and any of those can be the basis of a reliable one way flow into production with status coming back on a schedule. The design constraint is that file based flows are batch by nature, so the promise date logic and the customer notifications are built around a sensible cycle rather than pretending to be instant. If nothing at all is available, we will tell you that plainly rather than building something fragile on top of screen scraping.

How do we know which configurations are actually profitable?

By storing the structured configuration with the order and posting actual costs back against the job. Made to order margin varies enormously between jobs that look similar, because one option adds an operation, another wastes material at a certain size, and installation on a difficult site eats a day. Once configurations and actuals sit together, patterns appear quickly: the size range where waste jumps, the finish that always runs over, the customer type whose jobs need three site visits. That analysis is often worth more than the labour saving that justified the project in the first place, and it is only available if the data was captured properly from day one.

Sell It Online, Build It Without the Phone Calls

Book a call. We will map one of your configurable products end to end, tell you what to automate first, and give you a priced plan.

All discussions held in confidence. Australian-based consultants.