The Referee in the Room: Governing Shared Customer Data Without Squandering Portfolio Trust
Every holding company that has tried to move a lead from one portfolio company to another has run into the same three questions, usually in the same week. Who owns this customer? Whose consent covers this outreach? And who books the revenue if the referral closes?
Left unanswered, those questions get settled by whoever has the loudest operating partner or the most aggressive sales leader. That is not a governance model; it is a tournament. The holding company's job is neither to referee from the sideline nor to seize the data outright. It is to write the rules before the first record moves.
Begin With a Data Taxonomy, Not a Policy Document
Most portfolios draft a privacy policy first and a taxonomy never. Reverse that order. A practical taxonomy has three tiers, and each tier carries different handling rules.
Tier 1 — identity. Names, business and personal email addresses, phone numbers, account identifiers, billing details. This is the tier that triggers nearly every consent and notice obligation, and it should move between portfolio companies only under a signed instrument.
Tier 2 — behavioral and intent. Website activity, pipeline notes, support history, purchase patterns. Individually sensitive, but rarely regulated as tightly as identity data. Tier 2 should flow freely to a shared analytics environment — not to a shared sales floor.
Tier 3 — derived and aggregate. Segment models, propensity scores, benchmark indices, industry-level demand signals. This is the safest and most undervalued currency in a portfolio. Most of the synergy a holdco actually needs can be captured at Tier 3, with far less legal exposure than a raw list transfer.
The taxonomy does two things. It gives operating teams a simple test before any exchange begins, and it gives the holdco's counsel a defensible boundary to point at when a portfolio company asks for "everything."
Name a Customer of Record Before the First Lead Moves
Ambiguity about ownership is usually ambiguity about paperwork. Define two roles and keep them separate.
The customer of record is the entity that holds the contractual relationship, the billing relationship, and the consent. In almost every case, that is the portfolio company — not the center. The holdco can be a steward, meaning it maintains shared infrastructure, quality standards, and cross-sell logic, without becoming the counterparty on the customer's contract.
Recognition as customer of record changes only through a defined process: a re-papering of terms, a fresh notice, or an explicit consent event. A shared login to a CRM instance is not a transfer of ownership, and treating it as one is how portfolios end up on the wrong side of a regulator's question about legal basis.
Consent Boundaries Should Travel With the Data
Consent decays the moment it is copied into a system that does not track it. The fix is unglamorous and effective: a consent ledger that records, for each contact, the purpose, the source, the timestamp, and the scope of the permission granted.
Three operating rules make the ledger useful rather than decorative.
- Purpose limitation is enforced at the point of use. A contact who opted into product updates from one subsidiary has not opted into cross-portfolio marketing. The second use requires its own basis.
- Opt-outs propagate on a fixed clock. If a consumer unsubscribes at one portfolio company, the suppression should reach every other portfolio company within a defined window — 72 hours is achievable and defensible.
- Global Privacy Control signals are honored portfolio-wide. Treat a browser-level opt-out as binding across every property the holdco influences, not just the site where it was sent.
Telephone and text outreach deserve separate attention. The TCPA's consent standards are stricter than email norms, and a lead shared between portfolio companies does not carry a calling permission with it. Where the consent basis is unclear, the answer is written consent, not inference.
State-Privacy Guardrails for a Multi-State Portfolio
There is still no comprehensive federal privacy statute, and the state patchwork has only thickened. California, Colorado, Connecticut, Virginia, Texas, Oregon, Montana, and a growing list of others impose obligations that differ in threshold, definition, and remedy. A portfolio with customers in forty states is subject to a dozen overlapping regimes on any given day.
Trying to track the strictest rule state by state, business by business, is a losing game. Set a portfolio privacy standard at the highest common denominator among the states where the portfolio has material revenue, then review it annually. In practice that means contractual processor terms that travel with the data, documented risk assessments for cross-context behavioral advertising, a clear opt-out path for sale and sharing, and heightened handling for sensitive categories.
One standard, applied everywhere, costs less than a dozen local variations — and it survives an acquisition without a rebuild.
Attribution Rules That Survive a Board Meeting
Attribution disputes are rarely about math. They are about whether the rules were set before the deal was in play. Publish the schedule in advance and the argument ends before it starts.
A workable model has four parts:
- Two ledgers, not one. Track origination (the company that sourced the opportunity) separately from influence (companies that touched the relationship prior to close). Combining them guarantees a fight.
- A published split schedule by motion type. Referral, co-marketing, bundled offering, and shared-event sourcing each get a pre-agreed percentage. The percentages matter less than their existence.
- A synergy credit pool. A small share of cross-portfolio revenue is reserved at the center and allocated by the stewardship council, which prevents every dollar from being claimed twice.
- A dispute window and a forum. Thirty days to contest, then a standing attribution committee rules and publishes the outcome. Precedent accumulates; conflict declines.
Governance Cadence: The Stewardship Council
Assign a named data steward in each portfolio company — typically a operations or RevOps leader, not a lawyer. Convene them quarterly with the holdco's general counsel as an observer. Put four standing items on the agenda: taxonomy exceptions, consent-ledger health, attribution disputes, and one shared metric reported to the board.
After four quarters, a healthy program looks like this: no cross-portfolio transfer without a signed data-sharing instrument; suppression propagation measured in hours, not months; a dispute log that shrinks each quarter; and a board dashboard that reports shared-customer revenue without a footnote caveat.
None of that is glamorous. All of it is what separates a portfolio that genuinely connects its businesses from one that merely announces it will. The referee does not win the game. The referee is why there is a game worth watching.
