What an ongoing data refresh obligation can cost

Define the delivery before costing it

A monthly refresh sounds small until a source system changes two days before delivery. The useful question is whether the business can keep supplying the same agreed package without borrowing capacity from customer work. Count the human decisions around an export as well as the export itself. An automated job can run successfully while emitting a new field that was never approved for disclosure.

W3 C’s DCAT vocabulary describes update frequency and version relationships. Its Data on the Web Best Practices recommends version indicators and change history. These are useful ways to name a delivery; they do not prescribe a commercial refresh service or its price. Use them to distinguish a corrected historical release from a new period of records.

For a proposed contract, record the reporting period, included fields, exclusions, delivery date, correction window and permitted format. Assign a person to approve each release and someone to take over during absence. A promise whose only owner is a script is not a complete capacity plan.

Worked example: twelve monthly releases

This hypothetical USD budget concerns one license, one recipient and twelve monthly deliveries over one year. It assumes a bounded maintenance-record package that has passed its own rights and privacy review. The entered annual receipt is $36,000 if the specified deliveries are accepted and paid under the hypothetical agreement. It is not a market quote or prediction.

The seller estimates loaded internal hourly costs. There is no upfront seller referral fee to VOID and the illustration enters zero other external license fees. Tax, financing and revenue recognition are outside the calculation. Preparation is $6,000 once; one schema-change incident adds $3,000 during the year.

The arithmetic and the capacity test

Ordinary delivery cost is $1,400 × 12 = $16,800. Including initial preparation and the entered schema incident, the modeled first-year cost is $16,800 + $6,000 + $3,000 = $25,800. The difference from the conditional $36,000 receipt is $10,200 before tax. This is a contribution illustration for these assumptions, not profit the business can book now.

There are 168 ordinary staff hours in the twelve-release plan. Put those hours on an actual calendar. If the quality reviewer is needed for a plant shutdown in April, increasing the spreadsheet margin will not make that person available. A smaller field set or a different cadence may be the only deliverable proposal.

Run a second estimate for the work the recipient calls a correction. Re-exporting a failed file in the same agreed schema can be modest; rebuilding a mapping after a vendor migration can be a separate project. Define that boundary in the commercial discussion before either side assumes it is included.

A source-system change gets its own decision

In the hypothetical incident, the new system adds technician notes and merges two fault codes. The release owner pauses those fields, records the old and new schema versions, and prepares a metadata-only explanation for review. The $3,000 incident allowance funds investigation and remapping; it does not authorize new material or promise the recipient will accept it.

The completed change decision is: retain the original field scope, preserve the old fault-code interpretation in the manifest, and request a revised delivery date if review cannot finish. A request for all new notes is a scope-change proposal, even when the exporter makes including them easy. Its preparation, review and commercial effects need another decision.

What remains when renewal does not happen

For year two with no renewal, this illustration enters zero new receipts, stops future monthly releases and budgets $800 once for delivery closure and retention review. It does not erase surviving duties in the actual agreement. Identify which credentials to revoke, what manifest evidence to retain and who handles later correction requests.

The no-deal case is different: if preparation finishes but no license is signed, the entered loss is the $6,000 preparation cost. No monthly deliveries or schema incident are assumed in that case. Do not keep producing a feed for an unsigned opportunity simply because the exporter is ready.

Use earnings to vary your own receipt and deductions, then attach the refresh budget to offer comparison. VOID can help identify potential fit and coordinate a permissioned introduction; a sample, recurring delivery commitment and license remain separate approvals. The next useful artifact is a schedule the operating owner is willing to fund.