The Excel-to-ERP gap
Walk into any 50-person manufacturer in Tuas, Loyang, Woodlands, or Joo Koon and the operating system you will find is often not on a server. It is a spreadsheet, usually called something like Master Pricing Final v3.xlsx, and the formulas inside it are maintained by one person. When she goes on leave, the business slows down. When she resigns, the business discovers that the system was never really a file. It was a person.
For years that was fine. The spreadsheet got the company from S$2M to S$10M of revenue. Then orders multiplied, customers wanted real-time status, suppliers in three countries demanded structured purchase orders, and IRAS started pushing the market toward e-invoicing. Suddenly the spreadsheet is not a productivity tool anymore. It is the constraint.
The natural next move is to get an ERP. Three quotes come in. SAP Business One: six figures and a long implementation. Odoo: cheaper, but full of partner and customization risk. A no-code platform: fast in demo, weak when it meets real B2B pricing. The owner looks at the numbers, decides the master Excel can survive another year, and the same problem returns six months later.
This is the Excel-to-ERP gap. Most Singapore SMB manufacturers live in it for three to five years. They are too complex for spreadsheets and no-code, but too lean for enterprise ERP. Start Canyon exists for this gap: custom workflow systems, usually S$10k to S$30k for the first useful build, shipped in weeks rather than quarters.
Why off-the-shelf rarely fits this segment
SAP Business One, Dynamics 365 Business Central, and NetSuite are serious platforms. They are also designed around a level of process maturity that many Singapore manufacturers do not yet have. They assume dedicated finance ownership, internal IT capacity, implementation governance, and budget for consultants. For a 70-person factory where one admin handles half the operational exceptions, the implementation cost can become larger than the software problem.
Odoo sits closer to the SMB band. The product can be strong, especially when the process fits the modules. The problem appears when the business needs customer-specific pricing, subcontractor workflows, multi-language supplier updates, or unusual product configuration. The partner customizes heavily, the client ends up with a half-standard and half-custom system, and future changes become dependent on the partner staying interested.
No-code tools work for lightweight coordination. They rarely survive manufacturing pricing logic. Customer tier, material grade, dimensions, finish, lead time, MOQ, freight term, and approval margin are not a simple form. Once sales or operations has to open the old sheet to double-check the no-code tool, adoption is already slipping.
What custom means in the S$10-30k band
Custom does not mean bespoke-everything. A useful custom build is narrow. It replaces the few surfaces where Excel is causing daily drag, while leaving stable systems alone. If finance is working in Xero, Million, Sage, Globe3, or an existing SAP instance, we do not rebuild finance just to call the result an ERP. We integrate, export, or hand off cleanly.
A typical first Start Canyon build replaces three to five concrete things from the master Excel:
- A pricing engine that handles customer rates, product variants, quantity tiers, freight terms, and approval margins.
- An order pipeline that gives sales, admin, and management one shared view of every active job.
- A supplier loop where subcontractors can update their own assigned work without asking your production manager for the latest spreadsheet.
- A document flow for quotes, purchase orders, delivery orders, and exports that does not depend on Word templates.
- A mobile surface for sales reps, field buyers, or shop-floor staff who should not wait until they return to a desk.
That is why the price can stay sane. The first build does not pretend to replace every department. It removes the bottleneck that is costing the company the most time, speed, or control.
A concrete example
This is the kind of system that belongs in the gap. It is not SAP. It is not no-code. It is a workflow system with ERP-like discipline, focused on a measurable operational constraint.
How to scope the build without wasting money
Before any code is written, we run a paid discovery session. The output is a written build plan: how the workflow actually runs, where the exceptions live, what the first system should include, what it should not include, the fixed scope, the timeline, and the price. If you do not like the plan, you still keep the document.
The reason paid discovery matters is simple: free discovery rewards vague shopping. Manufacturing software fails when the scoping phase avoids the hard details. Who approves exceptions? Which customer rate overrides the normal discount? Which supplier can see which order? What happens when the job is split across two subcontractors? These are the details that decide whether staff use the system.
A typical engagement after discovery is six to twelve weeks of build, weekly demos, staging access from week two, then two to four weeks of adoption: training, data migration, parallel-run with the master Excel, and cleanup of the edge cases that only appear when real staff use real data.
Singapore-specific considerations
Three Singapore realities should shape the build. First, InvoiceNow and PEPPOL are becoming part of the operating environment. Even if e-invoicing is not the first problem, the system should not be designed in a way that makes it painful later. Second, grants affect buying behavior. PSG is narrow and usually tied to pre-approved solutions. EDG can be broader, but it requires a stronger business case. Third, generation transition changes the sale. The second generation may be pushing for modernization, but the founder still wants proof that the system will not become another failed IT project.
The best first system respects all three. It is small enough to be trusted, specific enough to be useful, and built on standard technology that the company owns.
Where to start
If the master Excel is the bottleneck and the SAP quote is the wrong shape, start with the workflow that hurts most. Quoting. Supplier updates. Production status. Customer order visibility. Pick one. Map it properly. Build the smallest system that changes daily behavior.
That is the S$10k to S$30k path between Excel and SAP. It is not the right answer for every manufacturer. It is the right answer when the business has outgrown spreadsheets but does not need a six-figure ERP implementation to move forward.