The Average-Customer Tax: What Enterprise Platform License Fees Actually Pay For
Every business paying an enterprise platform license fee is paying a tax it has never itemized. What that fee buys is software designed for the average customer of the platform. Not for you. For the mean of every company the vendor sells to.
We call it the average-customer tax, and once you see it, you see it everywhere.
Platforms are built to serve the middle of their market. That is not a criticism; it is the business model. The reference architecture, the default workflows, the roadmap priorities, all of it optimizes for what most customers need most of the time. The trouble is that no actual business is the average. Your operation differs from the mean in exactly the places that make you competitive, and that is where the tax gets collected.
Where the tax shows up
It starts with the workaround layer: the spreadsheets, side databases, and swivel-chair processes that bridge what the platform does and what your operation actually needs. Every one of them is unpaid labor the license fee was supposed to eliminate. Then there is the customization treadmill, configuration that fights the platform's grain and gets rebuilt at every major upgrade because it was never supposed to work that way. Add the modules and seats priced into the contract because the bundle was designed for the average customer's needs rather than yours, and finally the differentiation ceiling: the capabilities that would actually set you apart cannot be built inside the platform at all, so they wait. Sometimes for years.
The tax is measurable. Count the people-hours per week spent in spreadsheets and manual reconciliation that exist only to compensate for platform gaps. Compare paid capability against used capability, module by module and seat by seat. Track the engineering time spent re-fitting customizations after each platform release. And write down the backlog of capabilities the business wants that the platform cannot hold, because that backlog is the price of the ceiling.
When to pay it, and when to stop
Sometimes the tax is worth paying. If your process in a given area genuinely is close to the market average, a platform is the cheapest excellent answer. Payroll is usually payroll. General-ledger accounting is usually general-ledger accounting. The average-customer tax is only a problem where your business is not average, and being honest about which parts of your operation those are is most of the work.
What has changed is the math on the other side. Custom software used to be the expensive option, which made the tax rational almost everywhere. Built with AI-accelerated, specification-driven delivery, scoped against only the work that actually differentiates you, custom is now frequently cheaper than the tax it replaces. The build-versus-buy line has moved, and most license renewals are still being signed as if it hasn't.
The next time a platform renewal crosses your desk, do the itemization before you sign. Add up the workaround hours, the unused seats, the upgrade drag, and the differentiating capabilities waiting behind the ceiling. The number you get is not the cost of the platform. It is the cost of being treated as average. Whether that price is worth paying is a decision, and it deserves to be made as one.
Related reading
Beyond Vibe Coding: Build Products, Not Technical Debt
Beyond Vibe Coding: Build Products, Not Technical Debt is the talk Chris Coughlan gave at the NVSBC ...
EngineeringThe Real Moat in Agentic AI Isn't Speed. It's Specs.
There's a flawed premise running through most conversations about agentic AI for software developmen...
EngineeringWhy Spec-Driven Delivery Beats Prompt-Driven Delivery
The temptation with agentic coding tools is to skip the spec and go straight to the prompt. The mode...
EngineeringAccelerating AI Innovation: Building an Enterprise-Grade AI POC with {responsible} Vibe Coding
Every company today feels the urgency to infuse AI into their products and operations. But moving fr...
We're ready when you are. A real read on whether we are the right fit.
If we are, we will tell you what the work looks like. If we are not, we will tell you that too, and point you toward someone who is. Sometimes the right answer is "not yet."
Technology and AI solutions, delivered smoothly. Engineered with Conntinuum.