By Sharon Watkins, Founder and CEO, RadiusPoint
Multi-carrier TEM works when AT&T, Verizon, and regional carrier BANs share one inventory schema, one audit ruleset, and one GL coding model inside a single estate. The moment each carrier keeps its own spreadsheet language, reconciliation stops being analysis and turns into translation, and the errors that matter start hiding in the gaps between files.
Picture a mid-market company that just acquired a competitor with three sites in another state. Overnight it inherited a regional carrier it had never billed against, a fresh set of billing account numbers, and an accounts payable team now approving invoices in two formats it only half understands. Nothing breaks on day one. The break comes a quarter later, when a disconnected circuit keeps billing and nobody catches it, because that carrier’s invoice lives in its own file and answers to no one. The multi-carrier problem is a data problem before it is a cost problem, and RadiusPoint treats it as one by starting with the billing account number rather than the invoice total.
Key Takeaways
- National and regional carriers differ in customer service record formats and charge codes, so your inventory schema has to normalize them before any spend comparison means anything.
- BAN-level ownership prevents orphan regional accounts from surviving quietly after an acquisition or a local site build.
- One estate means one exception queue and one close file, not three carrier-specific closes that never agree with each other.
- RadiusPoint loads multi-carrier BANs into ExpenseLogic so analysts audit different invoice languages against one inventory of record.
Why Multi-Carrier Estates Break First at the BAN Layer
Multi-carrier programs break when each carrier’s BAN set lives in a separate file with no shared owner, status, or cost-center map. A BAN, the billing account number a carrier uses to group services under one bill, is the unit everything else hangs from, and once you hold BANs from several carriers, the weakest link is never the biggest bill. It is the small account no analyst was assigned to watch.
The failure mode is predictable. National carriers get attention because their invoices are large and their portals are familiar. Regional carriers arrive through acquisitions, local builds, or a single site only one provider serves, and they slot into a spreadsheet as an afterthought. That spreadsheet has no concept of status, so a circuit that should have been disconnected keeps paying. It has no owner, so no one is accountable when a charge drifts, and no cost-center default, so finance codes it to a catch-all where the variance disappears into overhead.
This is the complexity threshold where a managed telecom expense management program earns its keep. A single carrier can be run out of a well-kept spreadsheet by a diligent person. Three to seven carriers, each with its own billing format and dispute path, cannot, because the human effort scales faster than the savings. The estate does not need more spreadsheets. It needs one schema every carrier’s data is forced to enter.
Normalize National and Regional Invoice Languages
Normalize product codes, tax lines, and service ID formats from AT&T, Verizon, and regional carriers into one ExpenseLogic schema before you compare spend. Carriers do not describe the same service the same way. One labels a charge as network access, another buries it in a bundled feature code, a third splits it across two line items with a surcharge attached. Compare those raw invoices and you are comparing vocabularies, not costs.
Normalization is a ladder, and each rung turns raw billing into something you can audit. No charge reaches your general ledger, the finance system’s chart of accounts, until it has passed every rung.
| Rung | Layer | What happens | Output you can trust |
|---|---|---|---|
| 1 | Raw invoice | Charges arrive in the carrier’s own product codes, tax labels, and service IDs by EDI, portal, or paper | Source data only, nothing comparable yet |
| 2 | Charge taxonomy | Each carrier’s line items map to a shared set of categories: access, usage, features, taxes, surcharges, and one-time charges | Like-for-like charge types across every carrier |
| 3 | Inventory key | Every recurring charge ties to a specific circuit, line, or service in the inventory of record | Charges with no matching asset surface immediately |
| 4 | GL code | The normalized, inventoried charge carries a default cost center and a general ledger code | One close file that posts cleanly |
Only after the ladder is climbed does a line-item telecom audit become possible across carriers, because the audit ruleset now reads one taxonomy instead of five. A rate reversion on a Verizon circuit and one on a regional carrier’s circuit look identical to the ruleset, which is the point. Regional carriers are not worse billers than the nationals. They simply speak a different dialect, and normalization translates it rather than penalizing it.
An Ownership Model for Every BAN in the Estate
Assign a business owner, a telecom owner, and a cost-center default to every BAN so regional accounts cannot bill without an accountable party. Ownership is the control spreadsheets never enforce and the one that quietly saves the most, because an owned account gets questioned and an orphan account gets paid.
The grid below is the artifact that makes this real. Every BAN gets a row, and every row names the same five facts. When a new carrier arrives through an acquisition, it does not go live in ExpenseLogic until its rows are complete.
| BAN | Carrier | Business owner | Telecom owner | Default cost center | Status |
|---|---|---|---|---|---|
| 8017-XXXX | AT&T | Regional Ops Director | Analyst A | 6120 East | Active |
| 2245-XXXX | Verizon | Facilities Manager | Analyst A | 6120 West | Active |
| 5590-XXXX | Windstream | Site Lead, Plant 3 | Analyst B | 6135 Plant | Under review |
| 3302-XXXX | TDS Telecom | Unassigned | Unassigned | Default | Orphan, flag |
That last row is why the grid exists. An orphan BAN is not a rare event in a multi-carrier estate; it is the default outcome of acquisition-driven growth unless something forces assignment. Building this census is the first onboarding artifact any serious TEM program should demand, and it should be complete across all carriers before a single invoice is loaded.
A BAN is owned when three roles are named: a business owner accountable for the spend, a telecom owner accountable for the service, and a default cost center for coding. A BAN missing any of the three is an orphan account, and orphan accounts are where regional overbilling lives longest.
One Exception Queue Across Carriers
Run one exception queue ranked by dollars and age so a small regional BAN error does not hide behind national carrier noise. The instinct in a multi-carrier estate is to work each carrier separately, because each has its own portal and dispute process. That instinct is what lets the expensive errors slip, because it ranks work by carrier rather than by impact.
A single queue changes the ordering. Every exception the audit ruleset raises, whatever carrier produced it, lands in the same list, sorted by dollars at stake and how long it has been open. Age matters as much as size, because carrier dispute windows are finite and a correct finding raised too late is unrecoverable no matter how correct it is.
The design has three practical effects. It surfaces a large regional error above a trivial national one, because it ranks money rather than brand. It keeps aging disputes visible so nothing ages past its recovery window. And it gives one analyst a single place to work, so the estate closes from one queue rather than three carrier-specific lists that never agree.
How RadiusPoint Reconciles Multi-Carrier BANs in ExpenseLogic
RadiusPoint consolidates carrier feeds into ExpenseLogic, reconciles BANs to inventory, and produces one audited path to payment and GL posting. The platform holds the schema, the ownership grid, and the exception queue described above, and named people work them. That combination of software and a human on the account is the part a spreadsheet cannot copy.
In practice, ExpenseLogic ingests each carrier’s billing, applies the normalization ladder so AT&T, Verizon, and every regional carrier land in one taxonomy, and matches recurring charges against the inventory of record. Charges that do not match an owned asset become exceptions. A RadiusPoint analyst works that one queue, files disputes through the correct carrier path, tracks each to a credit, and only then releases charges to payment and general ledger posting. The estate closes once, in one file, with one set of cost centers.
This model matters most during change, when multi-carrier estates are most fragile. Acquisitions add carriers and BANs faster than internal teams can absorb them, and a provider transition can quietly lose the inventory that took years to build. The safe path is consolidating carriers without losing your inventory of record, so the BAN census, the ownership grid, and the audit history survive the move. One estate, one schema, one audited path to payment, no matter how many carriers the business collects.
Frequently Asked Questions
Do we need a separate TEM instance per carrier?
No. A separate instance per carrier recreates the spreadsheet problem inside software, because you end up with several systems of record that still have to be reconciled by hand. The value of TEM comes from one estate holding every carrier’s BANs against one inventory, one taxonomy, and one queue. Separate instances just give you separate closes.
How should regional carriers be prioritized?
By spend, site criticality, and invoice complexity, not brand familiarity alone. It is tempting to prioritize the nationals because their names are familiar and their bills are large, but a regional carrier serving a critical plant, or one with a genuinely tangled invoice, often carries more recoverable error per dollar. Rank carriers the way you rank exceptions: by impact.
What about wireless sub-accounts under national carriers?
Treat wireless master accounts and their sub-accounts with the same BAN discipline you apply to fixed services. A master account often hides dozens of sub-accounts, each capable of billing for lines assigned to people who left months ago. Give the master and its sub-accounts owners, statuses, and cost-center defaults like any other BAN. Wireless and fixed spend have their own differences, but the ownership rule does not change.
Can carriers be compared on unit cost easily?
Only after normalization. Raw invoice PDFs are documents, not comparable analytics, so any unit-cost comparison drawn straight from them compares two vocabularies and calls the result a number. Once every carrier’s charges pass the ladder into a shared taxonomy tied to inventory, unit cost becomes a fair comparison. Before that, it is a guess wearing a decimal point.