Nexus monitoring across 50 states: why this is an ongoing operating discipline
After South Dakota v. Wayfair, Inc., 138 S. Ct. 2080 (2018),[1] every US state with a sales tax can require registration and collection from a remote seller the moment that seller crosses a state-set dollar threshold. A brand's threshold count typically grows year over year as revenue increases and channels multiply, and each new obligation begins on the crossing date itself, not on the next scheduled review.
The complexity compounds with footprint. A brand at $50M in annual revenue typically sells through three or more channels (direct-to-consumer on Shopify, Amazon Seller Central, wholesale accounts on Faire), stores inventory across two or three 3PL nodes, and hired remote employees in four or five states over the past 24 months. Each of those vectors can independently create or extend nexus. Physical nexus from inventory stored at a 3PL facility in a state operates separately from economic nexus triggered by crossing that state’s revenue threshold. Either creates the obligation; the map needs to track both.
The monitoring discipline has three components. First, cadence: how often the map gets refreshed, calibrated to revenue band and footprint velocity. Second, data sources: the five inputs that feed the map with enough specificity to run state-by-state threshold math correctly. Third, reproducibility: the ability to show an auditor exactly how each registration decision was reached, on what data, and when the trigger date was established.
Brands that treat monitoring as a one-time project typically pass the first audit cleanly and surface exposure on the second, when the footprint has changed and the map has not. For a brand registered in 20 or more states, the open questions are whether the current map matches the current footprint, whether thresholds are tracked against the correct measurement basis per state, and whether the documentation would hold under examination. Those three questions define the monitoring function.
The right monitoring cadence by revenue band
The cadence that holds depends on how fast the nexus picture changes, which correlates with revenue band and channel complexity.
| Revenue band | Minimum cadence | What it means in practice | Off-cycle refresh triggers |
|---|---|---|---|
| ~$10M | Quarterly | Threshold-proximity check in each registered state; footprint review for new 3PL placements, channels, or remote hires since last review | Year-end revenue reconciliation against each state’s measurement window |
| ~$30M | Monthly | Threshold-crossing log updated from each source system; proximity alerts set at 80% of threshold in new-registration candidate states | Any new channel, 3PL node, or remote hire; any state threshold change |
| $70M+ | Monthly plus event-driven | Monthly baseline plus an immediate off-cycle refresh for each operational change: new 3PL node, new marketplace channel, remote hire in an uncovered state, acquisition with existing nexus | Acquisition due diligence; marketplace contract execution; 3PL network expansion |
These revenue points represent where footprint complexity typically outpaces a less-frequent cadence, not hard legal cutoffs. A brand at $25M with four active channels and six 3PL nodes may need monthly tracking before that band.
What the $30M monthly cadence looks like in practice: the finance lead pulls transaction data from each source system (Shopify dashboard, Amazon Seller Central reports, Faire exports, wholesale invoices) and runs it against each state’s threshold rule for the current measurement window. For states with a 12-month rolling window (Texas measures the preceding 12 months per Tex. Tax Code §151.107,[6] Minnesota uses a rolling 12-month period per Minn. Stat. §297A.66[8]), the check covers the trailing 12 months, not calendar-year-to-date. For states with an AND structure (New York requires $500,000 and more than 100 transactions per NY Tax Law §1101(b)(8),[2] Connecticut requires $100,000 and 200 transactions per Conn. Gen. Stat. §12-407(a)(15)(A)(v)[3]), both legs of the test need independent verification.
Proximity alerts matter as much as crossing detection. A brand tracking California exposure (Cal. Rev. & Tax. Code §6203[4]) at $430,000 with four months left in the measurement window needs to initiate registration before the crossing, not after. California requires collection before the next taxable transaction following threshold-crossing; waiting until $500,001 creates a retroactive gap.
TaxCloud’s native Shopify and Shopify Plus integration pulls transaction-level data automatically, so the threshold math runs against a current dataset without manual export-and-reconcile cycles between each monthly review.
The five data sources a defensible nexus map requires
A nexus map is only as current as its inputs. A brand with three active channels that tracks only Shopify transaction data is running threshold math on a partial dataset and will systematically understate exposure in states where all three channels contribute to the threshold count.
- Shopify direct sales, by state and by period. The primary source for most brands. Shopify’s transaction reports export by state for any date range. For Shopify Plus brands with B2B order flows, those transactions need to be pulled separately; they do not appear automatically in the DTC export.
- Marketplace-facilitated channel volume (Amazon, Walmart, TikTok Shop, eBay). Marketplace facilitator laws require these platforms to collect and remit sales tax on facilitated sales in most states, so the brand does not collect directly on those transactions. Whether facilitated volume counts toward the brand’s own threshold calculation varies by state. California includes facilitated sales in the seller’s threshold count (Cal. Rev. & Tax. Code §6203).[4] Florida excludes facilitated sales from the seller’s threshold (Fla. Stat. §212.0596).[5] Washington includes facilitated sales in cumulative gross receipts (RCW 82.08.052).[7] The brand needs seller-side volume reports from each marketplace to run this correctly, not just the marketplace’s tax remittance summary.
- Wholesale channel volume (Faire, NuOrder, Shopify Plus B2B). Wholesale volume is often excluded from Shopify DTC transaction data but may count toward threshold in states that measure gross sales rather than retail sales. Both Faire and NuOrder produce downloadable transaction reports by state.
- Physical presence inventory by state (3PL placements, FBA snapshots). A current list of active 3PL warehouse locations by state, updated each time the network changes. For Amazon FBA sellers, Amazon’s inventory placement reports show which fulfillment centers hold stock. FBA nexus remains contested outside Pennsylvania following Online Merchants Guild v. Hassell (Pa. Cmwlth. Ct. 2022),[9] but inventory snapshots should be maintained, timestamped monthly, and retained regardless of the brand’s current nexus position on FBA.
- HRIS data: remote hires by state. Every W-2 employee in a state generally creates physical nexus there. For a brand that grew from 25 to 65 employees over three years, with 18 of those employees in states where the brand was not previously registered, the HRIS feed identifies each registration obligation created by a hire. The trigger date is the hire date, not the next registration cycle.
TaxCloud calculates across 13,000+ jurisdictions through a single API integration, meaning rate lookups across all five source-system inputs flow through one connection rather than requiring the finance team to maintain separate jurisdiction tables per channel.
What a defensible nexus workpaper looks like at audit
A nexus workpaper is not a list of active registrations. It is a documented record of how each registration decision was made: what data was used, what threshold rule was applied, when the crossing date was identified, and when registration was initiated. An auditor wants to understand why registration happened when it did, and whether the brand had an unfulfilled obligation in any period before it, not just see a list of active registrations.
Four elements per state define a workpaper that holds up during examination.
Monthly state-by-state threshold-crossing log. A running record of cumulative revenue and, where applicable, transaction count in each state, updated monthly from source systems. The log should show the brand approaching the threshold, the first month the crossing was confirmed, and which review cycle caught it. If a crossing happened mid-month and the monthly close missed it, document why and when the off-cycle review caught it.
Source data references. Each monthly entry needs to be traceable back to the underlying export: the Shopify transaction report for the relevant period, the Amazon seller report for the same period, the Faire export. Source files should be retained and indexed. An auditor who asks how the brand calculated $97,400 in Texas revenue for a given quarter needs a path to the underlying data, not a recalculation.
Methodology notes. A standing section documenting the measurement basis applied per state. Texas counts total revenue including exempt and nontaxable amounts over the preceding 12 months (Tex. Tax Code §151.107; 34 TAC §3.286).[6] Florida counts only taxable remote sales, excluding exempt amounts (Fla. Stat. §212.0596).[5] Washington counts cumulative gross receipts including facilitated and exempt sales (RCW 82.08.052).[7] Applying one total-sales figure across all states produces the wrong answer in each of those three high-volume states. The methodology section shows which basis was applied per state and why.
Registration trigger date. The date the threshold was crossed (or physical nexus was established) and the date registration was initiated. The gap between those dates is what an auditor scrutinizes. A 30-day gap with a documented monthly review cycle is defensible; a 14-month gap with no intervening review is not.
Where manual tracking breaks down
Brands typically start monitoring nexus in a spreadsheet: one tab per state, revenue by month, a formula that flags when the running total crosses $100,000. That approach holds at five states and one channel. It begins to fail somewhere between 15 and 25 states, and the failure mode is not immediately visible because the spreadsheet still produces a number. It produces the wrong number.
The core problem is measurement-basis divergence across states. Three examples from high-volume states illustrate where the math goes wrong.
Texas measures total Texas revenue, including exempt and nontaxable amounts, over the preceding 12 months (Tex. Tax Code §151.107; 34 TAC §3.286).[6] A brand with $480,000 in taxable Shopify sales into Texas and $30,000 in separately stated nontaxable shipping charges sits at $480,000 in a Shopify-only export but at $510,000 against the correct Texas measure: over the $500,000 threshold and unregistered.
Florida measures only taxable remote sales, explicitly excluding exempt amounts (Fla. Stat. §212.0596).[5] A brand with $95,000 in taxable goods and $20,000 in exempt goods shipped into Florida sits at $115,000 on a gross-sales pull but at $95,000 against the correct Florida measure: under the $100,000 threshold and not yet obligated.
New York uses an AND test: $500,000 in cumulative gross receipts and more than 100 separate transactions, measured over the immediately preceding four sales tax quarters (NY Tax Law §1101(b)(8)).[2] A brand at $520,000 in cumulative receipts from 90 transactions has not yet triggered a New York obligation. Connecticut uses the same AND structure at $100,000 and 200 transactions, measured over the 12 months ending September 30 (Conn. Gen. Stat. §12-407(a)(15)(A)(v)).[3] A spreadsheet that handles OR logic correctly in most states and AND logic incorrectly in New York and Connecticut produces the wrong answer in two of the most commercially significant states in the country.
The multi-channel problem compounds this. Marketplace volume from Amazon, Walmart, and TikTok Shop needs to be included or excluded from each state’s threshold calculation, then reconciled against Shopify DTC revenue, Faire wholesale volume, and physical presence inventory data. Doing that correctly across 25 states and five data sources in a spreadsheet is nonreproducible: the next person who opens the file cannot verify the methodology, and neither can an auditor.
The operating model in steady state
A nexus monitoring operation running reliably at $50M to $80M in annual revenue looks different from one built at $15M and patched forward. The steady-state model has a clear three-way responsibility split, a defined monthly close sequence, and a compliance provider that absorbs the pieces finance and ops cannot efficiently run in-house.
Finance owns the map and the methodology. The finance lead (head of finance, controller, or VP of finance) is accountable for the accuracy of the nexus map, the measurement basis applied per state, and the workpaper that documents each registration decision. Finance does not need to be the system of record for every data source. It owns the conclusions drawn from those sources and the decisions that follow.
Ops owns the source-system feeds. The ops or logistics lead is accountable for current 3PL placement data, FBA inventory snapshots, and the HRIS record of remote hires by state. These are operational data points that carry nexus implications. The handoff is a monthly data pull delivered to finance before the nexus review closes.
The compliance provider absorbs the registration apparatus, the filing chain, exemption certificate management, and the audit documentation trail. Once finance identifies a new registration requirement, the provider handles the registration, sets up the filing calendar, ensures collection is configured in each active channel for that state, and maintains the documentation trail that audit defense requires. Finance does not need to track filing deadlines across 30 states or manage the certificate renewal cycle for wholesale customers.
What the monthly close looks like when this model runs correctly: finance receives the source-system data pull from ops by the fifth business day of the month. By the tenth business day, the threshold-crossing log is updated and any proximity alerts or crossings are documented. Registrations triggered by that month’s data go to the compliance provider. The provider confirms filing deadlines for the coming month and any open certificate issues. The map, the registrations, and the filing calendar are reconciled. No state moves to filing without a current registration on record.
The compliance provider’s role at audit is not incidental. The documentation trail connecting each registration to the threshold-crossing date and source data that triggered it is what makes the position defensible under examination, not just accurate in the current month. TaxCloud is built for that operational layer: 13,000+ jurisdictions calculated through one API, consolidated Streamlined Sales Tax (SST) filing across the 24 SST states,[11] and the audit documentation trail that links each registration to the triggering event.