Cross-System ID Reconciliation After M&A
After a merger or acquisition, the same customer, vendor, or employee often exists as separate records with unrelated ID schemes in each company's systems, and there's no shared key to join them on because the two companies never shared a customer numbering system, a vendor code format, or an employee ID structure. Reconciling this by hand means someone comparing names, addresses, and contact details across two full datasets and guessing at matches, a process that doesn't scale past a few hundred records and produces both false merges (two genuinely different customers combined because they share a similar name) and missed matches (the same vendor paid separately under two different vendor codes because nobody connected them), both of which create real financial and operational problems once the combined company starts operating as one.
STARTING PRICE
From €799
Complex tier · Multi-system orchestration, custom logic, and higher-volume or higher-risk processing.
Get a quote →Saves roughly 30-60 hrs on a typical post-merger reconciliation project, plus avoided financial errors from false merges or missed duplicate vendor payments.
How the automation works
We reconcile records across two companies' systems by matching on the combination of signals available, name similarity, address, tax ID or registration number where present, contact details, and transaction history patterns, rather than any single field, since no shared key exists to join on directly. Every match is confidence-scored, high-confidence matches (strong agreement across multiple signals) are presented as recommended merges, while lower-confidence and ambiguous cases are routed to human review with the supporting evidence shown side by side, rather than auto-merged. This is designed specifically to avoid the two costly failure modes of post-merger reconciliation: false merges that combine genuinely separate entities, and missed matches that leave the same vendor or customer paid or invoiced inconsistently under two live records.
Process flow
- 01
Both companies' datasets connected trigger
Customer, vendor, or employee datasets from both legacy systems are connected for reconciliation, scoped to the entity type being merged first.
- 02
Multi-signal matching ai
Records are matched using name similarity, address, tax/registration ID where available, contact details, and transaction pattern signals in combination, since no single shared key exists between the two systems.
- 03
Confidence scoring per match ai
Each candidate match gets a confidence score based on how many signals agree and how strongly, distinguishing clear matches from ambiguous ones rather than treating all matches as equally certain.
- 04
Human review of ambiguous matches output
Lower-confidence and ambiguous matches are routed to human review with the supporting evidence from both records shown side by side, so the reviewer is confirming a specific comparison, not starting from scratch.
- 05
Execute approved merges with ID crosswalk integration
Approved matches are merged or linked with a crosswalk table preserving both legacy IDs against the new unified record, so historical reporting and audit trails from either legacy system remain traceable.
- 06
Deliver reconciliation report and exception list output
A reconciliation report covering match rates, merge decisions, and any records that couldn't be confidently matched or reviewed is delivered for finance and ops sign-off.
Inputs
- Customer/vendor/employee datasets from both companies' systems
- Available shared identifiers (tax ID, registration number) where they exist
- Business rules for what constitutes a confident match
- Transaction or activity history for pattern-based matching
Outputs
- Confidence-scored match candidates
- Human-reviewed merge decisions with supporting evidence
- Legacy-ID-to-unified-record crosswalk table
- Reconciliation report with match rate and exception list
Works with
Prefer a fully custom build instead of an off-the-shelf integration? We scope both options during your free consultation — most jobs like this one work fine on standard connectors, but higher-volume or non-standard systems sometimes need bespoke API work, reflected in the complex tier.
Where this goes wrong if you get it wrong
- A false merge, combining two genuinely different customers or vendors because they share a similar name or address, is often worse than a missed match, because it silently corrupts financial history and relationship data for two separate entities going forward, while a missed match at least leaves both original records intact and findable.
- Vendor and customer records that were entered inconsistently within a single company before the merger even happened (the same vendor already duplicated under two internal codes) compound the cross-system matching problem, since the reconciliation now has to account for pre-existing duplication on each side, not just the cross-company matching itself.
- Auto-merging on a single strong signal, an exact tax ID match, for example, can still be wrong if that ID was mistyped identically in both systems by coincidence of a shared data entry error, or if a parent/subsidiary relationship means the same tax ID legitimately applies to genuinely distinct billing entities.
- Discarding the legacy IDs once a merge is complete breaks historical reporting and audit trail continuity, anyone needing to trace a transaction back to how it was originally recorded in either legacy system needs the crosswalk preserved, not just the new unified ID going forward.
Frequently asked questions
Does this automatically merge records, or just recommend matches?
High-confidence matches are recommended for merge, but the actual merge execution and anything below high confidence goes through human review first; this doesn't auto-merge records without that step.
What happens to the original IDs from each legacy system after a merge?
They're preserved in a crosswalk table linking both legacy IDs to the new unified record, so historical reports and audit trails from either original system remain traceable, not lost when the merge happens.
Can this handle vendors or customers that were already duplicated within one company before the merger?
Yes, pre-existing duplication within either company's own system is factored into the matching, since a cross-system reconciliation that ignores it would carry the existing duplicates forward into the merged company.
How does it decide what counts as a confident match without a shared ID to key on?
Confidence is built from multiple signals in combination, name similarity, address, tax ID where present, transaction patterns, rather than requiring one perfect shared key that doesn't exist between the two companies' systems.