Build-A-Bear ERP: Why Configurable Beats Custom for ETO Manufacturers – Rob Jolliffe of Sabre Limited, Part 2

Rob Jolliffe on Engineer-to-Order Manufacturing in Business Central
The Integrate Intelligently Podcast — with Jeff Brickler, Scott Brickler (CADTALK), and Rob Jolliffe (Sabre Limited)

Rob Jolliffe joins Jeff Brickler and Scott Brickler for a deep dive into why engineer-to-order (ETO) manufacturers break traditional ERP production modules, why the Projects module in Business Central is a better fit, and why Business Central's AppSource ecosystem makes it a "build-a-bear" ERP that ages better than most.

Speaker note: three speakers this time. Jeff opens with the question; Rob carries almost the entire technical walkthrough of ETO and Business Central; Scott steps in with his own background (the Northern Kentucky packaging machine manufacturer, and CADTALK's own origin story with Sightline/Visual ERP). Attribution is based on context — Jeff is referred to in third person by the opener ("that's Scott's background too"), Rob is the one describing his own consulting clients throughout, and Scott is the one describing "CADTALK's" origin and positioning. A few short exchanges near the end move fast enough that the Scott/Rob split there is a best guess.

Introduction to ETO Manufacturing Challenges
Jeff: One of the last questions I have — you focus on engineering-to-order, which of course is Scott's background too, and where CADTALK came from. Do you have a particular client story where this was just amazingly effective — where you thought, "man, this fixed-fee approach really worked"?

Rob: Yeah — for the last seven or eight years, just before COVID, we began doing all of our projects remotely, first on Zoom and now on Microsoft Teams, at a fixed fee. Virtually all of my ETO customers have been implemented that way.

Let me segue into ETO for a second, because I love it — it's a fascinating industry to me. There's a real challenge with ETO in a traditional manufacturing ERP module: it assumes the R&D process already happened before the first order came in. Somebody's already done the design, created the bill of materials and routings across however many levels, and it's just sitting there waiting for a sales order. When the order shows up, you release production, MRP plans your purchasing, and away you go. That works great for discrete manufacturing that's repetitive — whether that's ten units a year or a million a week.

The problem with ETO is that when the customer submits a request for quotation, they vaguely describe the problem they're having and the outcome they want.

The Million-Dollar Die Story: A $10 Pin Holds Up a $1M Product
Rob: Let's use an example of a customer who bought an ETO product rather than made one — a company that assembled the little shutoff valves under your sink or toilet. They had a customer, let's say something like Home Depot, ordering a million of these a year. It was all hand-assembled — people in white gloves at little tables, testing, boxing — maybe a thousand a day, but they needed far more volume. So they bought an automated assembly center: vibratory feeds, index tables, robots picking up parts, the whole thing encased in clear paneling with light curtains that would shut it down if you reached across.

Now — who else buys that? Maybe one other valve manufacturer somewhere with the same problem, but they don't make the same valve, so it has to be slightly different, at a different volume. This machine was built for one company, for one reason, and maybe ten years later you're lucky enough to build another one like it.

ETO companies also tend to be small — 50 to 100 people, often founded by an engineer who loved designing and inventing, who hired more engineers, with technicians in the shop. A lot of them don't even own CNC machines, mills, or lathes — they design, have contract manufacturers build the parts, and handle assembly, testing, qualification, and writing the user guides and manuals the customer's employees will use.

Why Production Modules Fail ETO: The R&D Assumption
Rob: That's a very different animal than a repetitive, high-volume manufacturer, and if you try to use the production modules of most ERP systems for it, they just fall apart. Customers push back immediately — "what do you mean I have to create a bill of materials and a routing for every part? I'm only making this once. I just want to send the drawing out and have the guys log their time against the drawing number. I don't want a BOM, just tell me the raw materials off my detail drawing, buy them, and make the part. I don't even need a part number — the drawing number is my item number."

I started realizing this a while back with a die builder client who built transfer dies — these are massive, mounted on a huge press, with a coil of steel running through them. The press comes down, stamps the part, indexes it forward, and repeats — the part goes through several steps inside a single die inside a single press, and pops out finished at the end.

Their big problem was communication between the engineering team loading the bill of materials and the purchasing team buying the parts — all done over email. Engineers would change a drawing, purchasing would miss the change, and they'd fail to buy some exotic, long-lead-time guide pin. Now you've got a million-dollar die. We still talk about this all the time: a million-dollar finished die, missing a ten-dollar pin with a six-week lead time because it's fabricated in Japan — so they're air-freighting it, treating it as an emergency, because it has to ship to General Motors.

The Valve Assembly Example: Building One-of-a-Kind Systems
Rob: That's where I cut my teeth on ETO — the valve assembly company, the transfer die company — and I decided the Projects module was a better home for this than the production modules. I'm a big fan of Business Central, and Projects in Business Central have something called planning lines, which can act like a BOM: they can carry an item number and a quantity. They're not quantity-per, but that's fine, because these are usually one-off systems — I'm selling a system, not a repetitively manufactured part.

So I could load the bill of materials into the project module, but the module itself isn't really designed to facilitate the purchasing process, version control, or the handoff from engineering to purchasing. So we've built some IP around the Projects module in Business Central for that — and honestly, we'd love to work with you all on bringing that data in. Our customers often ask, "can we look at CADTALK to import bills of materials?" and I always say let's absolutely put that on the table as part of the project. Sometimes it moves forward, sometimes it doesn't.

Why Projects Beat Production Orders for ETO in Business Central
Rob: From a customer's point of view, what we want is to take a bill of materials, load it into the project in phases or sections, and communicate that to procurement and planning, who decide make-versus-buy for each item based on what equipment they have. We're trying to create a controlled process between the design team and the planning team, so there's one source of truth — the bill of materials living in the project — instead of Excel spreadsheets and email threads.

On top of that, we've built a set of macros into Business Central to handle ETO-specific tasks — shipping components from the project to a job site, since you're often not shipping this as one box but breaking it into pieces. Projects also handle on-site installation and provisioning well, since these companies typically don't just sell the product — they also deliver, install, provision, debug, train the customer's employees, and handle final signoff. It's all progress-invoiced, percentage-of-completion accounting, because projects can stay open for a year on something really complex. It's a genuinely different way of implementing ERP — in some ways simpler than traditional production manufacturing, but in other ways it takes a lot more consulting to help a company design their project templates, because it's flexible enough that you can get lost in the options.

Scott: What you're describing is exactly the environment I came from too. I came from a custom packaging machine manufacturer here in the Northern Kentucky area — their first machine actually pressed Ivory soap for Procter & Gamble, which is where the little logo came from. They also built machines that load bottles of Budweiser into cartons — some parts reusable, a lot of it bespoke to the customer. It's a bit in between what you're describing, but I've worked with a lot of the project-based examples you're using.

It's interesting, because the idea for CADTALK came directly out of that environment — that was the original problem. We were using an Infor ERP at the time — I think it was actually called Sightline.

Rob: I'm familiar with Sightline — I worked with Visual ERP. As soon as you started describing it, I knew it was Visual, just from the way you were describing it. Visual has some interesting features around projects, like how it handles leg assemblies, that make that kind of thing more possible.

Scott: That's interesting, because I think the reason CADTALK has always worked mostly with production is that's how most people actually implement ERPs — we had to go after the market as it exists. A lot of people don't fully understand ETO, so a lot of ERP implementers just default to production builds, and then we have to work within that. I'd agree with you that Projects is the right way to do ETO, but I've rarely met people who've actually done it that way, even though it's the right approach.

Rob: We've rescued a few customers who started out trying to run ETO through production orders, came to us saying it wasn't working, and asked about switching to the Projects module — once they saw we had an app on Microsoft's AppSource for Business Central.

Jeff: Talk about that a little more.

Business Central vs. Infor: Scoring Manufacturing and Accounting
Rob: I really love Business Central. I came from the Infor space — Infor is fantastically well designed for manufacturing specifically. You really can't use it well outside that niche, and that's fine, because that's not what it's built for — it's called Visual Manufacturing, follows APICS best practices, and covers essentially everything a small-to-medium manufacturer needs. I'd give it a 9.5 out of 10 on core manufacturing capability. I'd give Business Central maybe an 8.5 — I wouldn't give it a 9.5 there. But Business Central gets a 9.5 from me on accounting, and I'm not going to embarrass Visual by giving its accounting score out loud.

The AppSource Ecosystem and the "Build-A-Bear" ERP
Rob: The other piece is AppSource — think of it like an iPhone app store. You want another app, you go to the store, click install, maybe pay a subscription, and you're done. That exists in Business Central; as far as I know, it doesn't really exist in Infor products — NetSuite has something similar, but Business Central's version is far more user-friendly, with roughly nine thousand apps in the store.

That means it's very easy to find genuinely high-quality add-ons that bring Business Central's manufacturing score up to that 9.5 — and often they end up exceeding what Visual offers, because in the Infor world, when Infor builds its own add-on for something, nobody else bothers building a competing one for that same purpose. So a CRM add-on built years ago just doesn't get reinvested in over time and drifts further from what a modern CRM should look like. In Business Central's AppSource, if you release an app and let it stagnate, a competitor will build a better version and overtake you. There used to be a production scheduling and finite capacity planning app we liked and sold, but the vendor stopped investing in it — and now a newer one has surpassed it, at a comparable price, both still built on Business Central's standard routings, production orders, and work centers. They might add an extra field here or there, but swapping one for the other isn't a massive rework — like switching GPS apps on your phone. You might re-enter your favorite addresses, but you don't need a new phone.

That's what makes Business Central so good — if you've outgrown part of it, chances are someone in AppSource has already built a better mousetrap. So when a client says "I want CAD integration," I ask: are your CAD libraries in good shape? Are you using a PLM, or Vault, or SolidWorks PDM? How are you storing your drawings — in libraries, with standardized attribute templates, and are your engineers actually filling those attributes in consistently? If yes, let's talk to CADTALK right now, you're ready. If not, talk to them anyway and find out what to prepare. It's a much easier process than the old days of custom SQL integrations you had to redo every time the ERP upgraded — all of that's gone. You're not buying an ERP anymore, you're buying a platform with an ERP built in, one you keep extending through the app store. I think Business Central is going to be one of those systems people run for thirty years without it becoming technical debt, because Microsoft keeps maintaining and updating it, and the AppSource ecosystem keeps improving around it.

Scott: That sounds like the incentives are genuinely aligned — similar to a conversation we had earlier, where having competitors in that ecosystem means there's a market pressure to keep improving, without the heavy vendor lock-in you'd otherwise get. You have some lock-in to Microsoft, sure, but there's a whole ecosystem of people building the add-ons underneath it.

Rob: Exactly — with my own Projects-based approach, as long as I keep it tightly aligned with what customers actually need, I'm fine. If I don't, another competitor in that ecosystem will build a better mousetrap, sell it to my customers, and I'll be out of business. So I have every incentive to keep adding features, tightening things up, and debugging — same as everyone else in that ecosystem.

What to Look for in a Business Central Partner for ETO
Rob: I think Business Central, as a manufacturing ERP, is very good — but you do have to curate the right add-ons. Some of that curation is obvious; there's a fairly well-known set of six to eight add-ons that dominate in North America, and Europe tends to have its own preferred set. As a general rule, I think you can get NetSuite, Acumatica, Epicor, or Business Central all up to a comparable level of manufacturing capability — it's just done differently in each. But personally, I'd rather have the "build-a-bear" ERP — a configurable system I can build up from components and Lego blocks, rather than relying on custom code.

CAD Integration Readiness and the Role of CADTALK
Rob: That's really the opportunity — go to AppSource, download the CADTALK connector, get your team trained on it, get it installed, and start using it. It's a world away from the old days of custom SQL integrations against a specific ERP release that broke on every upgrade. That's all gone now.