This map is part of the integration cutover reconciliation guide. The hub is about inventory before a legacy system is retired; this page is about joins: which identifier connects one system's record to the next, who owns each, and what a fully joined case looks like.
The map: system, record, linking key, owner
Every cutover spreads its evidence across tools chosen for other reasons. The table reads top to bottom in the order evidence is created.
| System | Record it holds | Key that links it to the next record | Usual owner |
|---|---|---|---|
| Mapping repository (shared drive, wiki or code repository) | Field mapping specification, one file per version | Version label, such as v2 | Implementation lead |
| ETL or integration tool, such as Azure Data Factory | Pipeline run history and row-level rejects | Run ID and start time, with the version label as a parameter | Integration engineer |
| Reconciliation workbook or reporting database | Counts and balances tested after each load | Run ID plus the name of the count tested | Reconciliation or finance analyst |
| Project tracker, such as Jira | Defect or cutover ticket, comments, field history | Ticket key cited in the workbook row | Project manager |
| Approval record (ticket, form or cutover runbook) | Exception sign-off | Approver name, date and mapping version | Controller or business owner |
| Legacy application and new ERP | The business records being moved, and code tables | Order and line identifiers, status code values | Business system owners |
Azure
JiraThe right-hand column matters. Six owners means six people who must agree that their part of the trail exists before anyone can describe it accurately.
Five joins, and what breaks each one
Each join can fail quietly. A partner cannot repair any of them, but can ask whether each one held.
- Mapping version to load run. Breaks when the version is not passed into the run or written in its name. The run then cannot be tied to a rule set.
- Load run to import log. Breaks when run history expires. Microsoft documents that Azure Data Factory only stores pipeline run data for 45 days unless diagnostic logs are routed through Azure Monitor, so a spring trial load may have no log by autumn.
- Import log to reconciliation count. Breaks when analysts label workbook tabs by date instead of run ID. Two loads on one day become indistinguishable.
- Count gap to ticket. Breaks when the ticket is opened from a chat message and the workbook never cites its key.
- Ticket to approval. Breaks when an exception is accepted in a meeting or by email and the ticket is simply closed. The decision exists, but not beside the evidence.
One fully joined case (Illustrative)
A fictional 380-person parts distributor's second trial load, followed end to end. Identifiers are invented.
| Step | System | What the record says | Key carried forward |
|---|---|---|---|
| 1 | Mapping repository | Mapping v2: legacy status H after 2019 with a pickup location becomes pickup hold | v2 |
| 2 | ETL tool | Run started Mar 9 with parameter mapping v2; 248,300 lines loaded, no rejects | Run ID R-0309 |
| 3 | Reconciliation workbook | Run R-0309, credit holds: 1,750 in ERP against 500 on the credit list | R-0309, "credit holds" |
| 4 | Project tracker | Ticket CUT-212 comment cites R-0309 and the 1,250 remaining lines | CUT-212 |
| 5 | Approval | CUT-212: 1,250 lines to load as review hold under mapping v3; controller and operations director, Mar 12 | Approvers, date, v3 |
With all five keys present, someone who was never on the project can answer: what rule was used, what it produced, how far off it was, who noticed, and who accepted the rest. That is the test of a joined case. The step-by-step story of the same exception, including who argued for each rule, is in the worked exception example.
Where platform behavior cuts the joins
Two tool behaviors deserve a direct question to the implementation director, because they remove links even when nobody deleted anything:
- Run history retention in the ETL tool. Ask whether runs were exported or logged elsewhere. The 45-day limit in Azure Data Factory is a documented example; other tools have their own settings, and the company should confirm them.
- Ticket history in the tracker. Jira search results can be exported, including to CSV, which gives the ticket list. Field history is a separate matter: Atlassian's data pipeline export writes issue history to its own issue_history file keyed by issue ID, and only includes history after the export's start date. A CSV alone should not be assumed to carry the order of decisions.
Cutovers driven by acquisitions are especially exposed, because the acquirer often retires the integrator's tooling soon after go-live; see the buy-and-build IT integration guide and the page on what happens when an MSP's client is acquired. For a different cutover risk, losing closed operating history rather than migration evidence, see the CMMS migration guide.
What to ask before an introduction
Keep the conversation at metadata level. For each row of the map, ask the implementation director:
- Does this record exist for every trial and production load, or only some?
- Which key does it carry, and was that key used consistently?
- Where does it live today, and when does that system lose its license, server or project space?
- Who owns it: the company, or the integrator under the statement of work?
The workflow map template is a neutral way to write the answers down, and the referral brief template turns them into a one-page introduction.
The introduced company needs 50+ full-time employees at peak (contractors excluded), documented operating history, rights to license and an authorized contact. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 90% of the data is in English. SourceX licenses approved business data directly from companies for AI training and evaluation and manages agreement, rights review, data preparation, delivery and payout. Demand, the rights result and any offer remain unknown until SourceX reviews the opportunity.
Next step
Refer a company. Earn up to $100,000. Default reward: 25% of Actual Supplier Payout, capped at $100,000 cumulatively per referred company. Eligibility and payment conditions apply, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.
When most of the five joins held and the owners of each system are known, register as a partner, confirm the company on who qualifies and read the program terms before you send the introduction.