Made-to-order manufacturing: the bottleneck is the quote, not the shop

Most outdoor living manufacturers are already running made-to-order. Every pergola, veranda, garden room, carport, and louvred awning is built after the order is signed, against a site, a buyer preference, and a regional regulation. The structure has to fit. There is no warehouse of pre-built 4 by 5 metre louvred pergolas in anthracite waiting to ship.

What most of them are not running is instrumented made-to-order. The configurator on the website, the pricing engine, the configurable BOM, and the shop-floor handoff are four separate jobs, often done by four separate tools, often glued together by a salesperson with a spreadsheet on a Sunday evening. That is where the model breaks, and where the manual quoting errors compound, one wrong line item at a time.

We were spending most of our week translating quotes into parts lists. Now the configured spec is the parts list. The Monday backlog disappeared.

This piece is about the difference between MTO on paper and MTO in production. What it actually means in outdoor living. Where the friction lives. What changes when the four blocks share one source of truth.

What MTO means in outdoor living#

The textbook definition of made-to-order is a production model where goods are built only after a confirmed order. The opposite is make-to-stock, where goods are built to a forecast and pulled from a warehouse. There is a middle ground, assemble-to-order, where standard sub-assemblies sit on the shelf and the final structure is built per order.

Outdoor living is mostly assemble-to-order in disguise. The aluminium profiles, the post systems, the louvre blades, the screen fabric, the screws and brackets are stocked or sourced on standing orders. The finished pergola, the finished veranda, the finished garden room is assembled, painted, packed, and dispatched after the buyer signs. The order triggers the cut list, the powder-coat batch, and the install slot.

That model has three structural advantages over make-to-stock, and one structural problem.

The first advantage is inventory. There is no warehouse of finished goods to fund, to insure, or to write down at the end of the season. The capital that would have sat in stock is free to fund spring ad spend or the second installation crew.

The second is cash flow. The buyer pays a deposit before the build starts. The build cost lands later. The gap between the two is, in practical terms, an interest-free working-capital loan, which is the single biggest reason small outdoor living trades can scale without bank debt.

The third is fit. The product is built for the buyer’s site, the buyer’s roof line, the buyer’s preference for anthracite or oak. The buyer is happier with the result. The reseller’s referral rate goes up. The margin holds against cheaper, mass-produced imports because the product is not the same product.

The structural problem is speed. MTO only works if the quote-to-build loop is short enough that the buyer does not lose interest, walk into a showroom down the road, or take the cheaper imported quote that landed on the kitchen table on Tuesday while yours arrived on Friday. First to quote wins is the industry truism, and it is the reason most outdoor living MTO businesses live or die on their quoting speed.

Where MTO breaks#

The bottleneck is rarely the shop floor. The shop floor is good at building things. The bottleneck is the handoff between two parallel translations of the same product.

Translation one. The salesperson talks to the buyer, takes the rough spec on a phone call or a site visit, draws a sketch on the back of an envelope, and types a quote in Word or a CPQ tool. The quote is what the buyer signs.

Translation two. The production engineer takes the signed quote, opens the CAD file or the parts catalogue, and rebuilds the spec as a real bill of materials. The BOM is what the shop floor cuts.

When the two translations drift, the install crew arrives with the wrong post height, the wrong screen colour, the wrong motor. The remake eats the margin. The buyer’s referral disappears. The salesperson spends the next week on a complaint call instead of a new lead.

The drift is structural. Two people, working from two different tools, rebuilding the same product from the same conversation. Even with a careful operator, the drift rate is hard to push below a few percent. With a busy operator, it is much worse.

The visible symptom is the time it takes to send a quote. The invisible symptom is the rework rate. They are the same problem.

1. €133M In quotes generated Mizarstvo Hrovat (Slovenia, modular houses and garden rooms), cumulative figure from the instrumented MTO loop.
2. 10,835 Hours saved Hrovat, on quoting and parts-list work no longer done by hand.
3. 200 Hours per month Industry baseline for manual quoting in a mid-sized outdoor living business, based on 25 quotes per month at 8 hours each.
4. 10× Close rate vs industry Industry pergola close rate sits at 5 to 8 percent. Trades running a configured MTO funnel commonly run 10 to 20 percent.

Full Hrovat write-up sits in the case studies.

The named numbers behind the patterns above sit in the Hrovat case study, where the modular-house and garden-room team rebuilt their quote-to-build loop on a single instrumented MTO stack. The number to anchor against is €133M in quotes, with 10,835 hours saved on work that used to be done by hand. The hours did not go to overtime. They went to the second installation crew, the new dealer market, and the founders no longer working Saturdays.

The four blocks of instrumented MTO#

The fix is not a faster CPQ. It is one source of truth, threaded through four blocks that already exist in the business, often run by four different tools.

Block one. The configurator. The buyer designs the product on the public website in 3D, picks size, posts, roof type, finish, screens, and accessories, and sees the build update in real time. The output is a structured specification, not a free-text description. Every choice is a parameter the BOM can resolve against.

Block two. The pricing engine. A real number, in the buyer’s currency, computed against real cost rules, real margin bands, and real regional installation modifiers. Not a “starting from” placeholder. The number lands on the configured PDF before the buyer leaves the site.

Block three. The configurable BOM. A parameterised parts list. The configurator’s output resolves to the exact posts, beams, louvre count, motor, wiring, screen rails, and fasteners for the chosen build. No production engineer rebuilds the spec. The same parameters that drove the price drive the cut list. There is no second translation, because there is nothing to translate.

Block four. The shop-floor handoff. The BOM lands in the ERP, the cut list lands at the saw, the powder-coat batch is scheduled, the install slot is booked. The configured spec carries through to the dispatch note and the installer’s tablet. If the buyer’s spec changes between the quote and the build, the change ripples through to every block without anyone retyping it.

The technology stopped being the hard part a long time ago. WebGL is mature. CPQ vendors have existed for two decades. ERPs handle parametric BOMs as a standard feature. What is hard is wiring the four blocks to share one parameter set. That is the work.

What changes when MTO is instrumented#

The change is not on the shop floor. The shop floor was already good. The change is in three places that used to be invisible.

Monday morning. Before, Monday opens with a stack of contact-form lines and a list of weekend voicemails. The first three hours of the week go on triage. After, Monday opens with configured PDFs. Each one has the build, the price, and the BOM. The sales team picks the inquiries with the longest time in the configurator first, because intent is visible in the data. By lunchtime, the orders that are going to convert are already through to production.

The drift rate. Before, three to five percent of jobs needed a remake because the install crew arrived with the wrong part. After, the configured spec is the parts list, so the drift rate falls to whatever rate the configurator’s rule library still misses. With a mature configurator the number is closer to zero than to one.

The founder’s calendar. Before, the founder is in the quoting loop on every job above a certain size, because the configurator only handles the standards and the bespoke work needs a human. After, the founder is in the production-strategy loop, the dealer-network loop, the second-market loop. The thing that used to take a Saturday now takes a click.

The 1-day claim for premade visual sales systems is real. What that buys you is the buyer-facing layer, live, on a public URL, with the configurator and the pricing engine and the PDF in production. The BOM wiring and the shop-floor handoff take the rest of the quarter, because the ERP is the thing that has to be taught the new vocabulary. The faster you start the first leg, the faster the second leg compounds.

When MTO is not the answer#

MTO is a fit for products that have to be specified per buyer. It is the wrong model for products that ship off the shelf.

The unit economics for instrumented MTO hold above 5,000 euros AOV, and they get more attractive every step up the ticket. Outdoor living typically runs between 8,000 and 35,000 euros per build, which is why the category is a textbook fit. Modular houses, kitchens, container rooms, and saunas sit higher and are even more attractive on paper, with the caveat that the configurator has more rules to encode before launch.

The eight checks before you instrument#

If the decision is to move from manual MTO to instrumented MTO, the demo is where the vendors sort themselves out.

1. 01

   One parameter set across the four blocks

   Ask the vendor to show the same parameter, say roof type, flowing from the configurator UI, through the pricing engine, into the configurable BOM, into the ERP export. If any block uses a different vocabulary, the second translation is back. Non-negotiable.

2. 02

   A real configurable BOM, not a flat parts list

   The BOM has to resolve from the configurator's parameters, not be retyped against a quote. Ask for a live demo with a buyer changing the spec mid-flow and the BOM updating without a person in the loop. Non-negotiable.

3. 03

   Premade for your category

   A vendor with a pergola template, a veranda template, a garden-room template, a carport template has spent the years learning the rules already. A generic 3D vendor with no category template is asking you to fund the learning curve. The premade route launches in 1 to 10 days; the generic route quotes 3 to 6 months for the same outcome. Non-negotiable.

4. 04

   A real pricing engine, not a 'starting from'

   Ask the vendor to price a specific configured build on the spot. If they say the pricing module will be configured later, the answer is no. The pricing engine is the thing that turns the configurator into a quote desk.

5. 05

   A designed PDF the buyer will keep

   The PDF should look like your brochure, list the spec line by line, show the price, and read as a serious document. A screenshot and a number is not a quote.

6. 06

   ERP and CRM events on the wire

   The configurator should fire events to the ERP, the CRM, Meta, Google Ads, and GA4 the moment a buyer designs, prices, and submits. Without those events, the marketing layer cannot tell ad spend from organic, and the operations layer cannot tell a serious inquiry from a tire-kicker.

7. 07

   Mobile performance on 4G

   Open the demo on a phone, on real data, not on the showroom Wi-Fi. More than half of outdoor living traffic is mobile. If the 3D model takes eight seconds to load, the buyer is back in their feed before the build resolves.

8. 08

   Three reference trades in your category

   Ask for three named outdoor living manufacturers running the same vendor in production. Call two. The line you want to hear is 'the quote-to-BOM handoff is gone.' If you only hear 'the demo looks good', the reference is vanity.

The first three are gates. If the vendor cannot share one parameter set across the four blocks, cannot demo a configurable BOM that resolves live, and cannot show a category-specific premade template, the next five do not matter. The last five are how good vendors separate from average ones.

The thing MTO is really about#

Made-to-order is, on paper, a production model. In practice, it is a positioning bet. The trades that run MTO well are betting that the buyer is willing to wait two weeks for a structure built to fit their site, in exchange for not getting the cheaper, faster, off-the-shelf product the neighbour has. That bet only pays off if the wait is two weeks, not eight, and if the quote arrives in minutes, not days.

The configurator, the pricing engine, the configurable BOM, and the shop-floor handoff are the four things that make the wait short enough and the quote fast enough for the bet to hold. The trades that have instrumented the four blocks have spent the last three years compounding on the trades that have not. The gap is not the shop floor. The gap is what happens between the buyer’s first click and the cut list.

If you want a longer companion read on the front end of the loop, the pergola configurator buyer’s guide walks through what the buyer-facing layer actually does, and the 3D configurator software buyer’s guide covers the five buying paths and what each one costs.