How far back StateWise can see, and what happens past the edge
Every number on your map is computed from orders. So the honest limit on what this app can tell you is the limit on the orders it can read, and that limit is worth stating plainly rather than discovering later.
StateWise reads about 24 months of Shopify order history. For most sellers and most states, that is enough to answer the question. For some it is not, and this page is about which is which.
Why 24 months, and not 60 days
Shopify's default order permission shows an app the last 60 days. That is a sensible default for most apps, and it is useless here. Economic nexus is measured over the current or previous calendar year in the majority of states, so a 60-day view cannot answer the question the product exists to answer. It could tell you about a two-month slice of one year and nothing about the year itself.
StateWise therefore requests an elevated, approval-gated permission for historical orders, and then backfills 24 months of them. The 24 is our choice, not a limit Shopify imposes: the permission lifts the 60-day cap, and we pick 24 months because that is what the measurement periods in the law actually require. Every window type in use fits inside it, with room to spare.
If that permission has not been granted on your store, the app sees 60 days, and it says so on the map rather than quietly showing you a shorter year. That case is real: a store installed before the scope was granted, or a grant that was never re-approved, lands there.
It reads a deliberately narrow set of order fields to do it, and never a customer's name, email, phone number or street address. That list is enumerated in full, and kept true in one place, on what StateWise reads and what it ignores.
What 24 months covers
Most state thresholds are measured over one of these windows:
| Window | Covered by 24 months? |
|---|---|
| Previous or current calendar year | Yes |
| Previous calendar year only | Yes |
| Trailing 12 months | Yes |
| Trailing 12 months, re-evaluated quarterly | Yes |
| Twelve months ending September 30 | Yes |
| Four sales tax quarters | Yes |
| Four completed calendar quarters | Yes |
| A seller's own fiscal year (Puerto Rico) | Yes, but it is approximated |
Every window type in use fits inside 24 months. So for the question "where do I stand right now, and where am I close?", the floor is not a constraint. That is the question most sellers are actually asking, and it is fully answerable.
Puerto Rico is the one exception worth naming, and it is not a history problem. Its window is the seller's own fiscal year, which cannot be known from order data, so that jurisdiction is measured on an approximation and is labelled as approximate on screen.
Where it runs out
The floor bites on a different question: "did I cross somewhere years ago and not notice?"
That is a real question, and for a multi-channel seller it is often the one an accountant asks first. A threshold crossed in 2022 does not become untrue because time passed. How long an obligation persists after a seller falls back below a line is called trailing nexus, it varies by state, and it is deliberately outside what this app models. We can tell you the test was met and when. What is still live today is a question for your accountant.
If you were selling through Shopify in 2022 and StateWise can only see back to mid-2024, it cannot tell you what happened in 2022. It will not guess, and it will not quietly report a crossing date of "the first date I have data for", which would be a fabricated answer wearing the clothes of a real one.
Two specific cases where this matters:
Rules that were repealed before the window opens. Since January 2023, which is as far back as our rule table reaches, nine jurisdictions have deleted a threshold rule that a seller may already have crossed under. Five of those removal dates are already older than a 24-month window, which means the crossing they would have caused sits outside what Shopify can hand back. Reconstructing those needs your own records.
Long-established sellers. If your store has been running for five years, the 24-month view describes the most recent two of them. Whether that is the whole story depends on how much your volume has changed.
There is a second floor, and it is not about your orders. Our rule table starts at 2023-01-01. We hold each jurisdiction's rule from that date onward and nothing before it, so a period earlier than 2023 has no rule version to be evaluated against. The engine refuses to produce a verdict there rather than reaching forward and applying today's rule to a year it did not govern, which would be the more confident answer and the wrong one. This floor bites even if you upload a 2019 order export: we would have your orders and still not have the rule that was in force.
How to see further
Upload your own order export. Shopify lets you export your full order history from your admin, going back as far as your store does, and StateWise imports that file the same way it imports an Amazon or Etsy export. The 24-month limit is a limit on what the API hands an app, not a limit on what you have. Your export is not bounded by it.
That is also how marketplace history gets in: Amazon, Etsy, eBay and Walmart order exports load the same way, and the per-state figures then include those channels for the whole period the files cover.
What the app tells you when it is at the edge
On the map itself, above the states rather than in a footnote, it states the date its history starts, the window it is working from, and when it last evaluated. It reads like "History from 2024-08-12", followed by the reason that window is what it is, and it says "No order history imported yet" on an empty store rather than implying a zero it has not measured.
That line is the thing to read before you trust a low number. "Under the threshold" and "under the threshold, as far back as I can see" are different statements, and the date is what tells you which one you are looking at.
Excluded orders are handled in the most explicit way the source permits. International, no-address and cross-channel duplicate orders are left out of state figures and shown with a count and reason. Connected Shopify data also identifies digital and local-pickup-only orders exactly, so those receive their own visible exclusion reasons. CSV exports do not expose those two facts consistently, and StateWise does not guess them from a missing address; imported rows are classified only from fields the export supplies. A number that quietly omits things is not a smaller number, it is a wrong one.
The general rule
Where a state's rule is unambiguous, this app is exact. Where the data runs out, it says so and shows you where the edge is. It does not fill a gap with an estimate and present it as a finding, because a monitoring tool that guesses is worse than one that admits its limits: you can work around a stated limit, and you cannot work around a confident wrong answer.
StateWise monitors and shows the arithmetic. It does not file, remit, register, or calculate tax at checkout, and it does not tell you that you owe tax. This page is not tax advice.
Sources. The window types and the repeal dates referred to here are versioned rows in the rule table, each carrying its citation and the date it was last verified, and both are shown beside the verdict they produce. Rules are evaluated against the period being measured, so a 2024 period is evaluated against the 2024 rule rather than today's. The per-jurisdiction table is at /thresholds.
Related: how the numbers are worked out, a historical crossing survives a repeal, what StateWise reads and what it ignores, the marketplace already collected the tax, New York needs both numbers.