A custom build and a standard package are not "better and worse" versions of the same thing. They are two different products, and the test is three questions you can answer before signing anything.

The three criteria

1. Is your core flow different? If your operating flow deviates more than about 30% from what the standard package assumes, custom is unavoidable — modifying the internals of a standard package costs more than building from scratch. If the difference is mostly branding and copy, a standard package plus light secondary development is enough.
2. How often will you iterate? If the core flow changes every month after launch, operating without source access is a slow leak. If it changes twice a year, a standard package is entirely adequate.
3. What does the team look like? With a team that can sustain long-term iteration, custom is an asset. Without one, custom is a liability — and the interest starts accruing on launch day.

The cost table

Read the last row carefully. Three-year totals cross over surprisingly often. When you run the numbers, remember to multiply the subscription by three years — and to price in your own staff time.

The trap in the middle option

The standard-plus-secondary-development route has its own trap: scope creep. The contract says three modified modules; mid-build it quietly becomes seven; the final deliverable sits between standard and custom, resembling neither. The defense is to attach the secondary-development scope as a contract exhibit, with anything beyond it priced through a formal change process — in writing, always. Also note that source ownership for the secondary work must be negotiated separately: licensing of the standard modules and ownership of the custom code are two different legal objects, and blending them guarantees a dispute by year three.

Two hidden costs, chronically underestimated

Data migration. Cleaning the old platform's data is the line item most often missing from quotes. Duplicate records, missing fields and inconsistent formats never show up at the "let's export and look" stage; they surface all at once during the real migration. Reserve a separate budget block for it.

Requirements communication. The largest hidden cost of a custom build is converting intent into an executable document. "Something like this" in an operator's head becomes a technical spec after one to two weeks of back-and-forth. Nobody enjoys this work, but skipping it is paid back as rework during development, at a much worse exchange rate.

The honest middle option

Standard package plus deep secondary development: standard modules carry the core for speed, while the two or three genuinely different modules are custom-built. Many teams end up exactly here. To decide which modules deserve custom work, apply the 30% flow-deviation test from criterion one.

For how the schedule is staged and where the buffers go, continue to the timeline breakdown. Multilingual and multi-currency support is another routinely missed cost line — see the multilingual practice. For market context: in the first-half 2026 review, standard packages closed about 70% of deals.