Choosing your route: module, provider or build

Station six established the size of the work on your data. This station decides who builds the technical layer above it.

Three routes, and no fourth.

The three routes

Route	What it is	Who it fits
1. A module from your system vendor	The vendor ships compliance inside the product	Whoever’s system ships it for their country
2. A provider or intermediary	A specialist connecting to your system on one side and the authority on the other	Most people — and mandatory in the UAE
3. Building in house	Your team builds the integration	High volume with a permanent team

Route one is cheaper and rarer than it looks. When your vendor says “we support ZATCA”, ask for the name, the version and the line item on the quotation — as in the readiness table. The distance between “we support it” and “it is in the price” is measured in tens of thousands.

Route three is chosen for the wrong reasons nine times in ten. It looks cheaper because its visible cost is zero: no per-document fee. Its real cost is a permanent person tracking format changes and fixing what breaks — a cost that never appears in the comparison because it does not arrive as an invoice.

The calculation that reverses the recommendation

Compute over five years rather than one, and at your real volume rather than a notional one.

Item	Unit	Do not forget
Licence or subscription	Annually	The contracted annual uplift
Document fees	Per thousand documents	The cap, and the price above it
Implementation	One off	The test environment, and training
Change	Recurring	Who pays for a format update?
People	Annually	Who reviews rejections daily?

The last two rows are what always gets forgotten, and what reverses the arithmetic.

“Who pays for a format update?” has to be asked in writing before signing. Formats do get updated, and a contract that turns each update into a priced change request converts the apparent saving into a bleed.

“Who reviews rejections?” matters more. Whatever the route, somebody opens the rejection report every morning — which is station nine. A project without that person is incomplete however good its solution.

Comparing at a real volume

Take your actual figure: how many documents do you issue a month? Then compute all three routes on it.

Three observations recur in every calculation.

• At small volumes — thousands of documents a month — per-document fees are immaterial, and the decision rests on implementation and maintenance.
• At large volumes — hundreds of thousands — per-document fees exceed every other line, and negotiating them is the project.
• Between them lies a band where the gap between routes is smaller than the gap in service quality, so choose on service rather than on price.

The common error is computing month one and multiplying it by five years with no growth. A company growing 20% a year doubles its document volume in four years, and with it the one cost line that scales with volume.

A worked comparison: twenty thousand documents a month

The figures here are illustrative and declared as such — the point is the shape of the calculation rather than its final number.

Line	A provider	A vendor module	Built in house
One-off setup	Moderate	Low	High
Fixed annual fee	Moderate	High (licence)	Zero
Document fees	240,000 documents a year × unit price	Usually zero	Zero
Somebody to watch and maintain	Part of a role	Part of a role	A full role
Cost of a format update	Usually the provider’s	The vendor’s	Yours, every time

Three readings.

One: the people line is what reverses the comparison. A full role per year exceeds the document fees on two hundred thousand documents in most offers. Leave it out and you choose the in-house build on arithmetic that is wrong.

Two: document fees grow with you and nothing else does. So compute year five, not year one.

Three: ask for the unit price at three volumes — yours today, double it, and a quarter of it. A provider who will not price all three is selling you a price for this year only.

What price does not buy

Three things worth paying a premium for.

One: the original error message. A provider that hands you the authority’s rejection text saves hours of diagnosis daily. One that hands you “transmission failed” turns every rejection into an investigation.

Two: a real test environment. Somewhere to try before production, repeatedly, without consuming real documents.

Three: the ability to leave. Extracting your whole archive and migrating. Not writing that into the contract is signing a lock.

What you actually do in this station

• Know your volume: documents a month, split by type and country.
• Compute all three routes over five years at realistic growth, using the five line items above.
• Ask your system vendor in writing: is compliance included, under which line item, at what price?
• Write your recommendation on one page with the numbers and the reasoning, and put it to management.
• Make both questions — updates and rejections — explicit clauses in the contract.

The three commonest mistakes

One: comparing on first-year price. The gap appears in year three.

Two: ignoring the people cost of building in house. It looks free and is the most expensive.

Three: ignoring growth in the document fee. A cap gets signed that is breached in year two at a marginal price nobody negotiated.

The acceptance test for this station

1. Name the three routes and who each one fits.
2. Show your five-year calculation with all five line items and realistic growth.
3. State your system vendor’s written answer on whether compliance is included.
4. Explain why the original error text is worth a price difference.
5. Show your one-page recommendation.

What comes next

You have chosen the route. The next station turns it into a project with a schedule: the implementation project — six months broken down, and what happens in each one.