The Value Creation ProtocolContents
Part III — The Core Framework Traps
Chapter 13
The ERP Trap → Value-First Platform
Bill Barlas joined me for this trap in the June 2025 series.
The room this chapter is about is the one where the question on the table is how to get people to follow the system. By then, the system has already won its argument with the work.
What this is
Here is the ERP Trap as the methodology defines it:
Forcing business processes to conform to rigid systems rather than enabling natural operational flow.
The name comes from enterprise resource planning software, because that is where the pattern is easiest to see. You don't have to own one to be in it. Any system your people bend their work to fit will do.
Be fair to how it started. Standardize the operation, cut the errors, make the business predictable: on a factory floor, where the work repeats, that is a good idea. What came along with it was an assumption, and the assumption is the trap. It says every organization should contort itself to fit software designed for manufacturing-era complexity.
The trap breaks Emergence over Predictability, the Core Belief that says "Value creation thrives through continuous adaptation and growth." The old default that belief answers is "Build once, maintain forever." A system that can't change is a bet that your work never will.
The alternative is Value-First Platform, and its first sentence is the one to remember:
Systems should enable business rather than constrain it. Choose platforms that adapt to your operations rather than forcing operations to adapt. Prioritize configuration over customization. Honor the actual work people do.
One line to keep two software traps apart, because they get confused: the ERP Trap is one rigid system the work has to bend to, and the SaaS Trap is the context your people rely on, split across many separate tools. They have different ways out. This one's is Value-First Platform. The SaaS Trap's is Value-First Context, and Chapter 6 covers it.
How to tell where you are
Three things you can do this week. None of them needs a consultant.
Listen. Somewhere in your building, someone is saying a version of this: we can't, the system won't let us. Notice who says it and what they were trying to do.
Look. Find the second list. It is the spreadsheet, the shared document or the notebook where the real version of the work lives. Ask its owner why they keep it. The answer will be a precise description of what your system can't hold.
Ask. Ask how the last small change went. A new field, a new step, a renamed stage. Who had to be called? What did it cost? How long did people do without it?
Then two questions about the bigger picture. Was the rollout planned in years? And is there a step in your process that exists only because the software needs it? If the honest answer to either one is yes, you are looking at this trap.
What to do
- Go to the desk where the work happens. Sit with the people who do the work and write down how it really goes, start to finish, second list included. Don't correct anyone. You are there to learn the work.
- Mark every place the work and the system disagree. For each one, decide which is right. Sometimes it is the system. A step that protects a customer, meets a legal requirement or keeps a record other people rely on should stay, and people should be told why it stays. Everything else on your list is the system asking the work to bend.
- Where the work is right, change the system, and configure before you build. Use what the platform already does before anyone writes something custom. Chapter 1 of this book teaches five claims, and one of them is Configuration Over Customization: "Configuration leverages platform investment and improves with the platform." Custom architecture, the same claim says, brings "brittle integrations" and "dependency on capability you don't have."
- Start with the worst one. Pick the single disagreement that costs your people the most and fix that. Then watch what they do with the fix before you touch the next.
- Let the owner retire the second list. When the system holds what the spreadsheet held, the person who kept the spreadsheet will stop keeping it. If they don't, the system isn't finished. Don't delete it for them.
- Decide who can make the next change. Make it someone close to the work. If every change still goes through a vendor or one specialist, your system is flexible for exactly as long as that person is available.
- If you are choosing a new platform, ask the adapting question first. Before features and before price: when our work changes, how does this change with it, and who on our team can do that?
How you know it worked
Not by the go-live. A go-live is a date. It tells you the software is on. It tells you nothing about whether anyone's work got easier.
The people who receive the value of a platform are the people who use it to do their work, and the customers on the other end of that work. Watch and listen there.
- The person who kept the second list stops keeping it, by their own choice, and can tell you why.
- When the work changes, the system changes soon after, and the person who asked for the change is the one who tells you it is done.
- A new hire learns the job, and the system is simply where the job happens. Nobody has to teach them how to get past it.
- A customer gets an answer in one conversation that used to take several people to assemble.
- You stop hearing that the system won't let them. Be careful with this one. Quiet can mean it is fixed, or it can mean people gave up saying it. Ask.
Logins, adoption percentages and tickets closed are counts of activity. Keep them if they help you run the system. They don't tell you it worked.
Where it goes wrong
The trap beside this one is the SaaS Trap (Chapter 6). People leave one rigid system by buying a separate tool for every complaint. Each purchase is reasonable. Together they split the context your people need across many places, and you have traded a wall for a set of islands.
Customizing your way out. You bend the rigid system with custom builds until only the builder understands it. Now the system fits, and you depend on one outside party to keep it fitting. That is the Managed Services Trap from Chapter 12, built into your own software.
Calling it done. A platform project with an end date becomes the next rigid system the day it ends, because the work keeps changing after the project closes.
Throwing out every standard. A system that follows the work is still a system. The steps you kept in move 2 are there for reasons. Letting everyone work however they like isn't Value-First Platform. It is the absence of a platform.
Protecting the money already spent. A common reason to stay put is what the current system cost. That cost is paid whether the system serves the work or not.
Read further
- The AI-Native Shift (Book 3). Chapter 3, "The Complexity You've Normalized", has the listening test for this trap. Chapter 8, "The Platform Question", teaches the configuration discipline as part of choosing a platform and designing on it. Chapter 16, "Handing Over the Keys", explains why there is no finish line.
- A Value-First Guide to Surviving the SaaSpocalypse (Book 1) tells the trap next door, the SaaS Trap, as a story.
- In this book: Chapter 6, on the SaaS Trap, and Chapter 1, where Configuration Over Customization is taught alongside the other four claims.