HTS code classification cost: what brokers, consultants and software charge

Getting a product classified costs somewhere between nothing and a few hundred dollars per line, and the spread has almost nothing to do with difficulty. It depends on who you ask. A customs broker adding a line to an entry charges around $5 to $15. The same broker doing a formal classification review on a new commodity charges $50 to $200. A trade consultant bills by the hour. CBP itself will give you a legally binding answer for free, and take about 30 days doing it.

Below is what each route actually costs, what you get for the money, and the arithmetic on a real catalog. The reason to care right now is volume: since CBP suspended the $800 de minimis exemption, shipments that used to clear on a manifest need formal or informal entry with a 10-digit code, so brands that classified a handful of products a year are suddenly classifying hundreds. A per-line price that was trivial at 20 lines is a budget line item at 400.

Last updated September 2026.

How much does HTS code classification cost?

Expect $5 to $15 per line when a broker adds a classification to an entry they are already filing, $50 to $200 per commodity for a documented classification review, $150 to $400 an hour for a licensed consultant or trade attorney, and $0 for a binding ruling from CBP that takes roughly 30 days. Software sits underneath all of those as a first pass rather than as a substitute for the licensed opinion on genuinely contested lines.

Those broker figures come from the price lists US brokers publish themselves, so treat them as a market band rather than a quote. The pattern behind them is more useful than the numbers: most brokers fold three to five HTS lines into the base entry fee, which runs about $100 to $250 for full-service clearance, and then meter everything beyond that. If your catalog is narrow and stable, you may never see a classification charge at all. If you import 40 different products in one container, you will.

What actually drives the price

Four things, in rough order of impact.

How many distinct products, not how many shipments. Classification is priced per SKU family, and it is a one-time cost per product if you keep the answer. Brands that pay repeatedly are usually paying because nobody wrote the reasoning down the first time, so the work gets redone at the next entry with a different broker's staff.

Which chapter you land in. Textiles, footwear, machinery parts and chemical mixtures are expensive to classify because the notes do real work. Apparel alone has a chief weight test that is a plurality rather than a majority, closure rules that flip a garment between headings, and a section note that forces retail sets together. A plain steel bracket is cheap to classify. A blended-fiber jacket with a detachable liner is not.

Whether you need a defensible file or just a number. This is the split most quotes hide. A code keyed onto an entry costs a few dollars. A code with a written basis that survives a CBP request years later costs more, because someone had to read the notes and record why they chose what they chose.

How contested the call is. Genuinely close calls are where money should go. A bag that could sit in heading 3923 as plastic packaging or 4202 as a travel article is a real fork with a real rate difference, and paying a licensed opinion or requesting a ruling on that one line is money well spent. Paying $150 for a review of a product with one obvious heading is not.

The math on a 300 SKU catalog

Say you sell 300 distinct products, sourced from three countries, and you have never formally classified any of them.

At $10 a line, having a broker classify the catalog costs $3,000. Pull out the 20 lines that are genuinely hard and pay for documented reviews at an average of $125 and that is another $2,500. Doing it in-house at a realistic 15 minutes per product, including reading the chapter notes, is 75 hours of somebody's time. None of those are unreasonable numbers. They are just numbers most brands never budgeted for, because until recently the parcels cleared without anyone naming the product in tariff language.

Now the other side of the ledger, which is much larger. A cotton knit t-shirt classifies at 6109.10.00 and pays 16.5 percent. The same shirt in man-made fiber classifies at 6109.90.10 and pays 32 percent. If you import $200,000 of that product a year and you are on the wrong side of that line, the gap is $31,000 annually. One misclassified staple product outweighs the entire cost of classifying the catalog, in either direction: you can be overpaying duty you never owed, or underpaying duty CBP will eventually want with interest.

That asymmetry is the actual argument for spending money here. The cost of classification is a known four-figure number. The cost of a wrong classification on a high-volume SKU is an unknown five-figure number that compounds every year it goes unnoticed.

What does a customs broker charge for classification?

Most brokers include three to five HTS lines in a base entry fee of roughly $100 to $250 and charge about $5 to $15 for each additional line. A standalone classification review of a new commodity, where the broker researches and documents the code outside of a live entry, typically runs $50 to $200. Ask specifically what the base fee covers, because that threshold is where surprise charges live.

Worth being clear about what you are buying. Classification is customs business under 19 CFR 111.3, and a licensed broker carries supervisory obligations for the work their staff does. That license is the value. What the license does not do is transfer the legal duty away from you: 19 USC 1484 puts reasonable care on the importer of record, in person or by an agent, and hiring a broker does not move it. If you want the workflow from the broker's side of the desk, our page on customs broker software covers how classification volume lands on brokers rather than on shippers.

Is a binding ruling worth 30 days?

For a product you will import for years, usually yes. A ruling issued under 19 CFR Part 177 is the official position of the agency and binds CBP personnel at every port until it is modified or revoked. Nothing else does. Under 19 CFR 177.1(b) even an oral opinion from CBP's own staff is not binding, which tells you exactly how much weight a vendor's confidence score deserves.

The limits are real, though. Rulings are generally available only for prospective transactions, so you cannot use one to fix an entry that has already liquidated, and you have to describe one specific product rather than a category. The practical approach is triage: classify the whole catalog quickly, identify the handful of lines where the rate difference is large and the call is close, and spend the 30 days on those. The rest do not need it. Our walkthrough of how a customs binding ruling works covers what goes in the request.

What you are really paying for is the record

Here is the part that changes how you should think about the price. Under 19 CFR 163.4(a), records relating to an entry have to be kept for five years from the date of entry, and CBP's list of records required for entry expressly includes the HTSUSA number and the description of merchandise. Your classification decisions are producible records, not internal working notes.

So the deliverable is not the ten digits. The deliverable is a retrievable explanation of why those ten digits are right, available years later, when the person who made the call has left and CBP is asking. Plenty of brands technically hold that information and still cannot produce it, because it lives in an email thread from 2024 and a spreadsheet nobody named properly. Getting an answer back out of your own files is a search problem across everything the company has written down more than a filing problem, and it is worth solving before somebody asks.

This is also why the cheapest option is often the most expensive one twice over. A code keyed onto an entry with no written basis leaves you with an obligation and no evidence. When the question comes back, you pay to classify the same product again.

The cost of getting it wrong

Penalties under 19 USC 1592(c) are the reason classification error is not just a duty problem. Negligence caps at the lesser of the domestic value or twice the lawful duties, taxes and fees. Gross negligence caps at four times. Fraud reaches the full domestic value of the merchandise. Note the column most people miss: where no revenue is lost, a negligent error is still penalizable at 20 percent of the dutiable value, and gross negligence at 40 percent. "The duty would have been the same anyway" is not a defense.

The counterweight is prior disclosure under 19 USC 1592(c)(4). Disclose before, or without knowledge of, the start of a formal investigation and a negligence or gross negligence penalty is capped at the interest alone on the unpaid duties, provided you tender them on time. That mechanism only exists if you find the error yourself, which is a direct financial argument for auditing your own catalog rather than waiting.

Where software fits, honestly

Software is the first pass. It is the right tool for turning 300 product descriptions into 300 candidate codes with the reasoning attached, in an afternoon, so that human attention goes to the 20 lines that deserve it. It is the wrong tool for settling a contested classification, and any vendor implying otherwise is selling you a liability rather than a product.

Judge a classification tool on one thing: does it show its work. A tool that returns a code and a confidence percentage gives you nothing to check. A tool that cites the General Rules of Interpretation, the chapter and section notes it relied on, and the product facts that drove the decision gives you and your broker something to review and correct. That written basis is also the record you are required to be able to produce.

If you are running an ecommerce catalog through this for the first time, our page on import compliance software for ecommerce brands lays out the whole cost stack per entry, including the merchandise processing fee floor that rises to $34.58 on October 1, 2026. For a side-by-side of the named vendors in this category, see the best HS code classification software compared.

Common questions

Can I classify my own products, or do I need a broker?

You can classify your own products. Nothing requires an importer to use a broker for classification, and as importer of record the duty of reasonable care is yours either way. A broker becomes necessary when you want a licensed opinion on a contested line, or when you want someone else transacting customs business on your behalf. For a straightforward catalog, the realistic constraint is time and access to the notes, not permission.

How much does it cost to classify a whole product catalog?

Budget $5 to $15 per line for broker classification and $50 to $200 for each product that needs a documented review. A 300 SKU catalog therefore lands somewhere between $1,500 and $5,500 depending on how many lines are genuinely difficult. Software brings the first pass close to zero and concentrates the paid work on the contested lines.

Do HTS codes need to be updated every year?

The tariff schedule is revised repeatedly through the year and the World Customs Organization amends the underlying nomenclature every five years, so codes do move. In practice the bigger source of change is policy rather than nomenclature: exclusion lists, Section 232 derivative lists and Section 301 annexes shift which additional duties attach to a code you already hold. Reviewing your top SKUs by import value once a year catches most of it.

Is classification software cheaper than a customs broker?

For volume, yes, and that is the wrong comparison. Software replaces the per-line data entry, not the licensed judgment. The sensible structure is software for the catalog, a broker or attorney for the close calls, and a binding ruling for the one or two lines where the annual duty at stake justifies waiting a month for certainty.