Auto parts software in Baltimore: the return is a purchase

The stores this is actually about

Baltimore is an old-car city. The average light vehicle on American roads reached 12.8 years old in the 2025 count published by S&P Global Mobility, the eighth consecutive annual rise; passenger cars on their own averaged 14.5 years and light trucks 11.9, across a fleet of roughly 289 million vehicles with a steady 4.5% scrappage rate. A fleet that old is the entire reason the independent parts trade exists, and it is also the reason the trade is so hard to put software around: the median customer walking up to your counter is driving something that left the factory when the iPhone 5 was new.

The businesses serving that fleet are smaller and more numerous than people expect. In the 2023 County Business Patterns file — the Census Bureau's count of establishments with paid employees, which is the caveat that matters in a trade with a lot of one-bay operators — Maryland had 521 automotive parts and accessories stores employing 6,489 people on an annual payroll of $224,998,000. Baltimore City held 24 of those stores with 304 employees and $10,235,000 of payroll. Baltimore County held 72 stores and 831 employees. Prince George's County leads the state with 104.

Two numbers in that file are worth sitting with. The first is that Baltimore City has more tire dealers than parts stores: 25 establishments in the tire retail code against 24 in the parts code, with 159 employees against 304. That is a genuine structural fact about this city rather than a rounding artifact, and it explains why a piece of software sold as an “auto parts system” so often turns out to be a tire system with a parts module bolted on, or the reverse. The second is where the money actually sits. Baltimore City's 14 motor-vehicle-parts wholesalers employ 192 people on $15,102,000 of payroll, which is $78,656 per employee against $33,668 at the retail counter — 2.34 times as much. The counter is where the trade is visible. The warehouse two exits up I-95 is where it is profitable.

Set the state against the country and something else shows up. Maryland's 521 parts stores are 1.39% of the national count of 37,482, while Maryland is about 1.83% of the national population. If the state carried its population's share of stores it would have roughly 687 of them; it has 521, about three-quarters of the expected number. The tire trade shows the same shape a little less sharply, at 279 dealers against an expected 343. Whatever the cause — density, the chains, the proximity of big distribution — the practical consequence for an independent here is that the catchment per store is wider than the national average, which is an argument for shipping and delivering rather than for waiting at the counter.

Three businesses wearing one sign

Every independent parts business I have looked at is really three businesses that happen to share a building. There is the retail counter, where somebody walks in with a symptom rather than a part number and leaves with a wiper blade and a bottle of coolant. There is the commercial book, where twelve repair shops within four miles order on account, take delivery twice a day, return a third of what they order, and expect a monthly statement that reconciles to the penny. And there is the online store, which is not a shop window at all but a database question: does this part fit the car I am standing next to?

Those three have different units of work, different margins, different return behavior and, crucially, different data. The counter needs speed and a human. The commercial book needs credit terms, core tracking, delivery routing and a returns policy that survives contact with a busy shop foreman. The online store needs fitment, and fitment is a problem of a completely different order, which is where I want to start when we get to what nobody models.

What the platforms charge, and what they are charging you for

On 13 September 2026 I checked the published pricing pages of the vendors an independent parts retailer in Baltimore would actually shortlist — counter systems, supplier ordering networks, storefronts, fitment search, marketplaces, and a few repair-shop systems that sell parts modules. Fifty-four URLs, a couple of dozen distinct companies. The result is the familiar one in this series and slightly worse than usual: the closer a product sits to the counter, the less likely it is to tell you what it costs.

The whole of the traditional jobber software tier is dark. epicor.com/en-us/pricing answers 403. mamsoftware.com/pricing, autologue.com/pricing and nexpart.com/pricing all answer 404. So do turn14.com/pricing, rockauto.com/pricing, motorstate.com/pricing, opticat.net/pricing and mitchell1.com/pricing. oreillyauto.com/pricing and napaonline.com/pricing return 403s, NAPA's with a 17-byte body. keystoneautomotive.com/pricing is the funniest of the set: it returns a clean HTTP 200 and a page of 212 bytes, which is roughly the length of this sentence. aftermarketautoparts.org/pricing manages 200 and 114 bytes.

The most telling result came from the category's best-known online-parts platform. revolutionparts.com/pricing does not 404; it redirects to the homepage, and that homepage is wall-to-wall with other people's revenue instead of its own price: a parts manager who “adds $15K a month in gross profit,” a dealer group at “$151K a month,” another at “$8M+,” a third at “$4M.” I do not think that is cynical. I think it is an honest reflection of how that market is sold, which is by outcome and by conversation. But it does mean that a Baltimore owner trying to budget a year cannot get a number without a sales call, and that is a real cost even when the software turns out to be good.

Six vendors do publish, and between them they show you the whole meter.

Look at the last row for a moment, because it is the cleanest illustration of what this article is about. Registering a tire with its manufacturer is not a growth feature. It is a federal recordkeeping duty under 49 CFR 574.8, which requires an independent dealer to hand every tire buyer a registration form carrying the full tire identification number and the dealer's name and address. One vendor in this market has correctly identified that as something shops will pay to automate, and prices it at $39 a month per shop, which is $468 a year to not type thirteen characters by hand. That is not a criticism of the price. It is an observation that in this trade, compliance is a product category.

The tier is priced in visitors. The feature you need is a text box.

Convermax publishes the clearest tier ladder in this market, and reading it carefully is the most useful ten minutes an owner can spend before signing anything. Stage 1 is $250 a month for up to 50,000 users a month and 250,000 products, with the phrase “all features included.” Stage 2 is $450 for 100,000 users and 500,000 products, again “all features included.” Stage 3 is $850 for up to 1,000,000 users and 1,000,000 products, and its feature list opens with two words that do not appear on either tier below it: VIN search.

VIN search is not an extra. For an auto parts store it is the entire user interface. A customer standing in a parking lot in Hampden does not know whether their Civic is an EX or an LX, or which of two engines it has, or whether the 2012 facelift changed the bracket. They know there is a barcode on the door jamb. Putting that seventeen-character string into a box is the single interaction that converts, and on this ladder it lives on the plan sized for a million visitors a month.

Run the arithmetic for a real Baltimore store. Say the online side does 40,000 sessions a month, which is a healthy number for an independent with a good local reputation and some marketplace spillover. That fits comfortably inside Stage 1's 50,000. To get VIN search it must buy Stage 3 anyway, which means paying for 960,000 sessions a month it does not have in order to obtain one field on a search form. In money: $3,000 a year becomes $10,200 a year, a difference of $7,200, and the effective rate goes from $0.005 per session at Stage 1's own ceiling to $0.02125 per session at Stage 3 on real traffic — 4.25 times the unit price for the same job.

That $7,200-a-year gap is worth holding onto, because it is larger than the entire fixed price of the custom store we would build. A $6,000 online store is paid for by ten months of the difference between two tiers of one search plugin.

eBay charges four different rates inside one department

The marketplace side has its own version of the same idea, and it is the purest example I have found of classification deciding money. eBay's final value fee for a seller with a Store subscription is 11.5% on the first $1,000 of a sale in eBay Motors > Parts & Accessories. But four sub-nodes of that same department carry different rates: tires pay 9.5%; In-Car Technology, GPS & Security pays 9.35%; Apparel, Protective Gear & Merchandise pays 12.7%; and a seller with no Store at all pays 13.6% on everything up to $7,500. Add the per-order fee of $0.30 on orders of $10.00 or less and $0.40 above that, and an insertion fee of $0.35 once you pass 250 listings in a month.

On a single $400 sale those four rates are $50.80, $46.00, $38.00 and $37.40. The spread between the best node and the worst is $13.40 on one item, or 3.35 percentage points. A store doing $250,000 a year through that channel pays $28,750 at the parts rate and $23,375 at the in-car-technology rate: $5,375 a year decided by which branch of somebody else's category tree a SKU was filed under. An LED light bar, a dash camera, a backup sensor kit and a floor mat can each be argued into two different nodes, and nobody at your counter is thinking about that when they list one.

The Store subscription itself is a clean break-even calculation, because it buys exactly 2.1 percentage points on items under $1,000. Basic at $21.95 a month on yearly renewal costs $263.40 a year and pays for itself at $12,543 of parts sales. Premium at $59.95 pays for itself at $34,257. Anchor at $299.95 needs $171,400. Those three numbers are a better description of the stages of an online parts business than any marketing page, and I would rather an owner had them than a demo.

The storefront that upgrades itself

BigCommerce publishes both halves of its meter, which is rarer than it should be, and both halves matter here. The plans are $29, $79, $299 and from $1,499 a month on annual billing. Attached to each is a trailing GMV cap — $30,000 for Core, $100,000 for Growth — and the page states plainly that the plan auto-upgrades when you cross it. Scale caps at $33,333 a month of GMV and charges 0.9% on anything above that. So your plan is not really a choice; it is a function of your own sales, evaluated monthly.

The second half is the one that catches parts sellers, because they nearly all have a processor already. BigCommerce charges an open payment provider fee of 2.0% of eligible order GMV on Core, 1.0% on Growth and 0.6% on Scale, and $0 if you use its embedded payments. On a store doing $480,000 a year online, that 0.6% is $2,880 a year — a charge for continuing to use the merchant account you already had.

Two things fall out of that table. The first is that Stack B is 2.98 times Stack A, and the entire difference is one tier of a storefront and one tier of a search plugin. The second is the ratio that actually governs this business: against $16,007 of software and marketplace subscriptions, the cost of moving the money is $34,512 — $2.16 of fees for every dollar of software. If you have an afternoon to spend improving your margin, spend it on the payment and channel mix before you spend it on the software, and be suspicious of anyone who tells you otherwise. We build software for a living and it is still the right order.

The part no national platform models

Here is where a Baltimore parts store stops looking like a generic retailer. Three separate Maryland instruments land on the same order line, and each one turns on a classification rather than on a price. A platform built in San Francisco or Austin models price, quantity, tax rate and shipping. It does not model any of this, and it cannot, because none of it is a setting.

Two tire fees, one line, and a tax decided by your invoice template

On 1 January 2026 the arithmetic of selling a tire in Maryland changed more than it had in twenty years. The Budget Reconciliation and Financing Act of 2025, Chapter 604 of the Acts of 2025, created a brand-new $5.00 per-tire fee under Transportation Article §22–421, its revenue going to the Transportation Trust Fund, and raised the existing tire recycling fee under Environment Article §9–228 from $0.80 to $1.00, going to the Used Tire Clean Up and Recycling Fund. The Comptroller set it all out in Technical Bulletin No. 61, effective 19 December 2025.

Context for how large that jump is: the recycling fee was first imposed on 1 February 1992 and had sat at $0.80 since 1 April 2005 — almost twenty-one years without a change. It is now indexed, adjustable by the Department of the Environment every two fiscal years against the Consumer Price Index, and capped at $2.00. The rate-setting power moved from the Board of Public Works to MDE, which means the number in your point-of-sale system is now a number an agency can move on a schedule.

A set of four tires that carried $3.20 of fees in December 2025 carries $24.00 today. That is 7.5 times as much, an increase of $5.20 per tire. On a $780 set it is 3.08% of the ticket, which is to say it is now a material line on the invoice rather than a rounding error, and customers notice it.

Then comes the sentence that turns a tax rule into a software rule:

“If the amount of the tire recycling fee and new tire fee are separately stated in a retail sale, the tire recycling fee and new tire fee are excluded from the taxable price and are not subject to sales and use tax … If the tire recycling fee or new tire fee are included in a lump sum charge for a retail sale, and are not separately stated, the fee(s) will be subject to Maryland sales and use tax as part of the taxable price of the tire.” — Comptroller of Maryland, Technical Bulletin No. 61

Read that as an engineer and it says something startling: the taxability of $6.00 is decided by your invoice template. Not by what you sold, not by whom you sold it to, not by where it shipped. By whether the renderer put the fee on its own row or folded it into the price. Six percent of $6.00 is $0.36 a tire, $1.44 on a set of four, and for a shop moving 6,000 tires a year it is $2,160 a year that exists or does not exist depending on a formatting decision made once, years ago, by somebody configuring a cart.

That is a real risk in e-commerce specifically, because online stores love a clean price. Every “all-in pricing” feature, every “include fees in displayed price” toggle, every road-hazard-and-fees bundle that makes the checkout tidier is, in Maryland, a decision to make $6.00 a tire taxable. The tidy option costs money. I have not seen a single platform that warns anyone about this.

Three more details from the same bulletin are worth building into any system that touches tires. The fees are imposed on sellers, not buyers, which is why the Comptroller states flatly that nonprofits, federal, state and local governments and diplomatic customers are not exempt — so the exemption certificate that correctly suppresses the 6% sales tax must not suppress the $6.00. That is exactly the sort of rule a generic tax engine gets wrong, because it treats a customer as exempt or not exempt rather than exempt from a particular instrument. Second, the timely-filing discount of 0.6% applies to the recycling fee only — the bulletin says so explicitly — so a shop remitting $36,000 a year of fees earns a discount of $36.00, one tenth of one percent of the money it moved, because the discount lands on one sixth of it. Third, a no-charge replacement tire carries no new fee, but a tire sold at a reduced price under a guarantee does, at the full $6.00: the fee does not prorate the way your warranty adjustment does.

The core charge is taxable even when you refund it — unless it is one of four truck parts

Now the object that comes back. A core charge is, in every practical sense, a deposit: you add it to the invoice so the customer has a reason to bring you the dead alternator, and you give it back when they do. Maryland's sales tax regulation even contains an exclusion that sounds written for exactly this, at COMAR 03.06.01.08, which keeps out of the taxable price “a deposit posted to insure the return of an item taken from the vendor which the vendor expects to have returned, and which will be refunded in full upon the return of the item.”

Read the emphasis. That exclusion contemplates the vendor's own reusable thing going out and coming back: a keg, a pallet, a gas cylinder, a reel. A core is the opposite vector. It is the customer's dead part coming in, and it never belonged to you. Same refundable dollar, opposite direction of ownership, and the Comptroller's own guidance draws the line exactly there:

“The taxable price is not reduced by prompt payment discounts, manufacturers' rebates, commissions or ‘core charges,’ or deposits to insure a trade-in, even if refunded. … The tax does not apply to core charges made in connection with sales of the following re-manufactured truck parts and their components: air brake systems, engines, rear axle carriers and transmissions.” — Comptroller of Maryland, Business Tax Tip #8

“Even if refunded” is the whole sentence, really. And behind it sits the statute: Tax-General §11–101 defines taxable price to exclude “the value of a used component or part (core value) received from a purchaser of the following remanufactured truck parts: an air brake system; an engine; a rear axle carrier; or a transmission.” Four rows. That is the entire list. Your catalog has tens of thousands of SKUs and the tax treatment of their cores is decided by a four-item enumeration written into the tax code, qualified by an adjective — truck — that the section does not define.

It is not an oversight, either. In the 2016 session, House Bill 540, titled “Sales and Use Tax – Exemption – Trade-Ins and Core Charges,” proposed to exclude from the taxable price “the value of used components, parts, or batteries received from the purchaser of a component, part, or battery for an automobile, a truck, or a marine engine.” It got a hearing on 23 February and never came out of the House. In the ten regular sessions since, no one has filed it again. The legislature was asked to make every core a deposit, and it declined.

Two identical $250 cores, $15.00 apart, and the difference is the word truck. If your system stores a core charge as a number on a line rather than as an object with a classification, you cannot produce that table, and neither can your accountant at the end of the quarter.

There is a second-order problem that online stores hit and counters mostly do not. A core charge generates two taxable events that have to net to zero: tax charged when the core goes out, tax reversed when the core comes back. If your store issues the core refund as store credit, a gift card, or a fresh credit memo rather than as a credit against the original sale, the second event never happens and the reversal never reaches the return. The customer is made whole on the $85 and left short on the $5.10, and the difference sits in your liability account looking like revenue. We have found exactly this in two builds now, both on well-known platforms, both created by an entirely reasonable refund workflow that was never designed with a taxable deposit in mind.

While we are here: Maryland is one of only five states — with Alabama, Alaska, Delaware and Montana — that never adopted the Battery Council International model battery recycling law, the one that gives most states a statutory take-back duty and a statutory core deposit. So in Maryland the battery core charge is pure private contract. Nothing in state law sets it, refunds it or requires you to take the old battery. That is the mirror image of the next section, and the contrast is the point: two objects that both come back across the same counter, one governed by nothing at all and one governed by a criminal statute.

The one core you may not be allowed to buy

A catalytic converter is not, legally speaking, an auto part when it comes off the car. Maryland's Business Regulation Article, Title 17, Subtitle 10 — the junk dealer and scrap metal processor subtitle — defines “junk” or “scrap metal” to include a list of nonferrous metals and then a short list of used articles made of either ferrous or nonferrous metal. There are nine items on that second list. The first one is catalytic converters. The rest of the company it keeps tells you everything about the legislative intent: metal bleachers, hard-drawn copper, beer kegs, cemetery urns, grave markers, propane tanks, cell tower batteries, and utility property such as manhole covers and street signs. This is a theft statute.

“Junk dealer or scrap metal processor” is defined at §17–1001(g) as, simply, “a person who does business buying or selling junk or scrap metal.” And the exemption a parts store would reach for first, at §17–1002(a)(2), is written more narrowly than it looks. The subtitle does not apply to a person doing business other than junk business whenever the person buys or transports junk used in the operation of the business, or transports for disposal or sale junk accumulated by the business.

That word whenever is the most software-shaped word in the whole subtitle. It is not a status you hold; it is a test applied to a transaction. Taking in cores that came off jobs, and sending the accumulated pile to a recycler, sits comfortably inside it. Running a standing program that pays walk-ins cash for converters does not, because at that point the store is doing business buying junk. Your accounting system models this as a company-level setting — are we a scrap dealer? no — and the statute models it as a per-row predicate. Between those two models is where compliance gets lost.

Whether any particular store is inside the subtitle is a question for its own lawyer, and I am not one. What is not in doubt is what the law demands of anyone who is inside it, because Chapter 484 of the Acts of 2024, effective 1 October 2024, sharpened it specifically around converters. Under §17–1009(b)(2) the record of each purchase must state the date and time; a description including type and grade, and the weight if payment is by weight; and now, if the scrap includes all or part of a detached catalytic converter, the vehicle identification number of the vehicle from which it was obtained. Plus the amount paid, the registration plate number, make and model of any vehicle used, the seller's name and address, the signatures of both the seller and the employee who accepted it, and the seller's date of birth and driver's license number or, failing that, identification details from a state-issued photo ID “including the sex, race, any distinguishing features, and approximate age, height, and weight.”

Then comes the provision that decides whether a store can run this at all:

“A junk dealer or scrap metal processor may not purchase a catalytic converter from an individual unless the individual, at the time of purchase, provides identification as: (i) a licensed automotive dismantler and recycler or scrap metal processor; or (ii) an agent or employee of a licensed commercial enterprise.” — Md. Business Regulation §17–1009(d)(2)

In plain terms: a covered buyer cannot take a converter from a member of the public at all. Not with a VIN, not with a photo ID, not with a signature. The DIY customer with a converter in a bin liner is not a permitted counterparty. For a checkout, that is not a validation rule, it is a customer-type gate — and it is the only one I have found in this series that makes an entire transaction unavailable to retail.

Four more features of §17–1009 are pure systems design, and they are the reason I think this subtitle is more interesting to a builder than to a lawyer.

The records “shall be kept in electronic form.” A 2015-vintage statute mandating a database is unusual and useful: it means the paper ledger under the counter is not a compliant fallback. A copy of each record must then be transmitted to the primary law enforcement unit “by the end of the first business day following the date of the transaction,” and here is the part that will make any integrator wince — it must be sent “in a format acceptable to the receiving primary law enforcement unit.” That unit is defined at §17–1001(k) as the State Police, a police department or a sheriff, as designated by a resolution of the county or municipal governing body. So the export schema for your compliance feed is not set by a state agency or a standards body. It is set, county by county, by a local resolution and by whatever that department happens to accept. An operator with a store in Baltimore City and a second one over the county line can face two recipients and two formats for the same record.

The copy that goes to the police is also not the same record you keep, and the difference is delightful. The retained record carries “the amount paid or other consideration.” The transmitted copy carries, at §17–1009(b)(4)(iii)3, “whether the amount paid or other consideration for the junk or scrap metal exceeds $500.” Not the amount — the boolean. The statute has specified a projection with a computed column, which is either a thoughtful privacy decision or an accident, and either way your exporter has to implement it. It also wants an electronic scan or photocopy of the photo ID attached, which means your compliance feed carries images of identity documents out of your building on a daily schedule, which is a sentence that should make anyone think carefully about retention, access and encryption. Maryland's privacy law has its own views about identity documents sitting in a database.

And then, buried at the end of the same paragraph, §17–1009(b)(4)(iv): those requirements “may not be construed to require a junk dealer or scrap metal processor to incur a substantial financial burden to comply.” I have read a lot of Maryland statutes for this series and that is the first time I have seen a legislature put a cost ceiling on its own software mandate. It is not a number, it is not enforceable in any obvious way, and it exists because somebody in 2015 understood that the cost of a compliance integration falls hardest on the smallest operator. A law enforcement unit may also waive the electronic requirements — extending the deadline by a day, permitting written records, or accepting fax or mail.

The enforcement end is short and sharp. A law enforcement agency with reasonable cause to believe an item is stolen may issue a written hold notice for up to 15 days, during which the item may not be processed or removed from the premises — which is an inventory state your system has to be able to represent, because a held converter is neither sellable nor shippable nor writable-off. Violating the section is a misdemeanor carrying up to $500 for a first offence and up to $5,000 or a year for a subsequent one. And there is an exemption worth knowing: the record and reporting duties do not apply to items acquired from a licensed dealer, from a government body, or from a commercial enterprise with a valid business license that has a written contract with you, provided you have given the police that enterprise's name and address and the type of scrap covered. That is a real answer for a store with a dozen repair-shop accounts: a signed contract per account, filed once, converts a daily reporting duty into a relationship.

Maryland tried once more, in 2022, to write a dedicated converter transfer rule — Senate Bill 546, “Vehicle Equipment – Transfer of Catalytic Converter – Regulation” — and it died. What survived is this one, a theft statute with a VIN field bolted onto it in 2024.

The ship-to address decides whether the part is legal

The other converter problem belongs entirely to the online store, and it is the cleanest example anywhere in retail of a product whose legality is a function of the delivery address. An aftermarket catalytic converter certified to federal EPA standards may be sold and installed in most of the country. A handful of states require converters carrying a California Air Resources Board Executive Order number instead: California, where EO labeling has been mandatory since 1 January 2009, along with New York, Colorado and Maine, each with its own wrinkle — Maine's requirement applies to vehicles built after 2001, Colorado's turns on vehicle age and warranty status.

Maryland is not one of them. A Baltimore store can sell an EPA converter to a Maryland customer without a second thought. But Baltimore is two and a half hours from New York's state line and a short drive from Delaware and Pennsylvania, and an online catalog sells to whoever loads the page. The same SKU, at the same price, in the same box, is an ordinary sale to an address in Dundalk and a problem at an address in Queens. That is not a tax rule or a shipping rule; it is a product-availability rule keyed to the ship-to state, and it has to run before the add-to-cart button, not after. Stock carts have a shipping-restrictions feature for hazardous goods. Almost none of them will let you express “this variant is purchasable only outside these four states, and its CARB-compliant sibling is purchasable everywhere at a higher price.”

The same logic lands on a handful of other SKUs for a different reason. An airbag module is a hazardous material in transport: under PHMSA's rules a safety device of that kind ships as UN3268, Class 9, when it has been tested and certified to the relevant special provision, and as a Division 1.4G explosive when it has not. A lead-acid battery is Class 8. A lithium-ion traction pack is UN3480 and is barred from passenger aircraft outright. Aerosol brake cleaner is limited quantity. Every one of those is a parcel your carrier treats differently, and the old ORM-D consumer-commodity marking that used to smooth the whole category over for ground shipments is gone. In practice this means a parts catalog needs a transport classification on the SKU alongside its fitment, its core flag and its state availability — four orthogonal attributes, none of which is price.

The fitment problem, measured

Everything above is compliance. This one is the business itself. A parts catalog's real difficulty is not how many parts it holds but how many vehicles it has to describe, and that number is larger than most people guess.

I counted it from a source anyone can check: the EPA's fuel economy database, which lists every light-duty vehicle sold in the United States with its make, model, year, engine displacement, cylinder count, drive type, transmission and fuel. Restricted to model years 2001 to 2026 — roughly the fleet still on Maryland roads, given that 12.8-year average — it holds 31,051 rows covering 86 makes and 4,423 make-and-model pairs. Those collapse into 19,655 distinct year-make-model combinations, and expand into 29,679 distinct configurations once drivetrain is included: 1.51 configurations per base vehicle, on average.

Averages hide the pain. A Ford F-150 alone accounts for 200 year-and-model rows and 372 configurations across that period, a ratio of 1.86. A Honda Civic is 83 rows and 211 configurations, a ratio of 2.54 — two and a half distinct engineering vehicles behind every line a customer would recognize as “a 2014 Civic.” A Toyota Camry is 118 and 175. And the model year 2026 alone added 1,172 new rows, so the table grows by about a thousand every autumn whether or not anyone updates your catalog.

Two honest caveats. The EPA file is light-duty only, so it undercounts the heavy end of a parts store's book, and its model names are EPA's rather than the manufacturer's, so it will merge trims a catalog would keep apart. Both errors point the same way: the real number is bigger. The industry's own answer to this is the Auto Care Association's ACES and PIES standards and the vehicle configuration database behind them, which is the right answer and also a licensing relationship, an import pipeline and a permanent mapping job between somebody else's vehicle IDs and your own SKUs.

Which is why VIN lookup is not a luxury feature. A VIN collapses all 29,679 of those configurations into one, using a number the customer can read off a door jamb. When a platform puts that behind its top tier, it is not gating a convenience. It is gating the only interface this trade actually has.

Two smaller Maryland notes, and one that did not happen

First, software itself. Since 1 July 2025 Maryland charges 3% on data or information technology services and software publishing, so most of the subscriptions in the tables above arrive with a tax line that did not exist two years ago, and a bundled invoice can attract the 6% rate instead of the 3% one. It is worth itemizing what you buy for the same reason it is worth itemizing a tire fee.

Second, what did not happen. In the 2026 session, House Bill 1429 would have created a Maryland right to repair for motor vehicles and farm equipment, requiring manufacturers to make diagnostic and repair information and tools available to owners and independent repair facilities. It had a hearing on 10 March and did not pass. For the independent channel that is a live question that will be back, and it is the single biggest potential change to what data a Baltimore parts store can get at.

What custom actually costs

We publish our prices for the same reason I have just spent four thousand words on other people's: a number you can put in a budget is worth more than a demo. These are fixed, not estimates, and they are the same numbers on our pricing page.

Set that against the stacks. The layer a custom store can genuinely replace is the storefront plus the fitment search — not the marketplace commissions, not card processing, not your supplier ordering network, all of which stay exactly where they are. That replaceable layer is $3,948 a year in Stack A and $13,788 a year in Stack B. A $6,000 online store is therefore paid for in about five months against Stack B and about eighteen months against Stack A; a $12,000 build in about ten months and about three years respectively.

If you are on Stack A and happy, those eighteen months are a real argument for staying put, and I will say so on the call. The case for building gets strong at exactly the moment the platform ladder asks you to buy a tier for one feature.

What we would actually build

The design follows from everything above, and it starts with one decision: an order line is not a price and a quantity. In the systems we build for this trade, a line is a small object that knows what it is, and four things hang off it that a generic cart does not have.

A core is its own record, not a number. It has an amount, a classification that decides whether it belongs in the taxable price, a state machine — charged, due back, received, inspected, credited, expired — and a link back to the original sale so the credit reverses the right tax against the right invoice. The converter case gets its own subtype carrying the VIN, the counterparty type, the ID capture and the export job, and a customer-type gate that simply refuses the transaction when the counterparty is an individual. If the store has contracts on file with its commercial accounts, the system knows which accounts are covered and which are not.

A statutory fee is its own kind of line, separate from tax and from price, that renders itself as its own row because rendering is what decides its taxability. It carries its own effective dates, so the day the Department of the Environment moves the recycling fee the system takes a new value rather than a new deployment, and it ignores the exemption certificate that suppresses sales tax, because that certificate has nothing to say about it.

A fitment record is a relationship between a SKU and a set of vehicle configurations, not a text field. It imports from the industry standard rather than being retyped, it supports VIN decoding at the front, and it is honest with the customer about the difference between “fits” and “fits, with this note.”

And a SKU carries its own transport and legality attributes: hazard classification for the carrier, state availability for the ship-to check, core group, and the supersession chain that tells you this part number replaced two older ones. Supersession is the quiet one that wins arguments with customers, because it is the difference between “we don't have it” and “we have the part that replaced it.”

Around that sit the ordinary things: a counter screen fast enough to keep a queue moving, a commercial account view with terms and a statement that reconciles, a delivery run, and a returns flow that a shop foreman can use one-handed. None of that is exotic. The exotic part is that the four attributes above are usually bolted onto a data model that never expected them, which is why they break at the edges — at the refund, at the exemption, at the state line, at the tier upgrade.

Build, or keep paying

I would rather you kept a subscription that works than paid us to rebuild it. The honest test is short, and it is the only list in this article:

• Keep buying when the platform's meter is roughly proportional to what you earn from it, when the features you need are on a tier you have already outgrown financially, and when nobody in the building is doing a monthly job by hand to patch a gap.
• Build when a tier upgrade is buying you one feature, when compliance is arriving as a paid module, when the same data is being retyped into two systems, or when a rule that decides money — a fee line, a core, a state restriction — lives in somebody's head instead of in the software.

Most parts businesses we talk to end up somewhere in the middle: keep the supplier ordering network, keep the marketplace channels, keep the accounting package, and own the storefront and the counter logic where the trade's actual peculiarities live. That is a $6,000 to $12,000 decision, not a rip-and-replace.

When you should not call us

If you are a single location doing modest online volume on Stack A and your fitment is good enough, you do not have a software problem, you have a marketing problem, and we are the wrong studio. If you are mid-migration onto a jobber system your distributor subsidizes, finish it. If your real issue is that returns from three commercial accounts are unreconciled, that is a bookkeeping engagement before it is a build. And if what you need is an off-the-shelf catalog feed with 200,000 SKUs and a national brand behind it, buy that — we are not going to out-catalog the Auto Care Association.

How we work

Fixed price, agreed before we start. You talk to the people writing the code — there are two of us and no account layer. You own every line, in your repository, on your infrastructure, from the first commit. We ship in weeks, not quarters, and we start with the one screen that is costing you the most, which in this trade is almost always the fitment search or the core ledger. If a week of work would tell us both whether the rest is worth doing, that is what the $3,500 Prototype Sprint is for, and it comes off the price of the build.

Questions we get from parts store owners

How much are the tire fees in Maryland in 2026?

Since 1 January 2026 a new tire sold at retail in Maryland carries two separate fees: the tire recycling fee, which rose from $0.80 to $1.00 per tire, and a new tire fee of $5.00 per tire. Both were changed by Chapter 604 of the Acts of 2025, the Budget Reconciliation and Financing Act. A set of four tires now carries $24.00 in fees where it carried $3.20 a year ago, which is 7.5 times as much. The fees are imposed on the seller rather than the buyer, so there is no exemption for a nonprofit, a government body or a diplomat, and a tax exemption certificate does not remove them.

Are Maryland tire fees subject to sales tax?

Only if you fail to itemize them. The Comptroller's Technical Bulletin No. 61 says that if the tire recycling fee and the new tire fee are separately stated on a retail sale they are excluded from the taxable price, and that if they are rolled into a lump sum charge and not separately stated they become part of the taxable price of the tire. At 6% on $6.00 that is $0.36 per tire, or $1.44 on a set of four, decided entirely by how your invoice template renders the line.

Is a core charge taxable in Maryland?

Yes, with one narrow exception. Business Tax Tip #8 states that the taxable price is not reduced by core charges or by deposits to insure a trade-in, even if refunded. Tax-General §11–101 removes the core value from the taxable price for exactly four remanufactured truck parts and their components: air brake systems, engines, rear axle carriers and transmissions. Everything else, including a reman alternator, a brake caliper, an air conditioning compressor or a battery core, keeps its core charge inside the taxable price. House Bill 540 of 2016 would have extended the exclusion to any automobile, truck or marine component, part or battery, and it died in committee.

What records does Maryland require when a catalytic converter core is taken in?

Business Regulation §17–1009 requires a junk dealer or scrap metal processor to record the date and time, a description including type and grade, the amount paid, the seller's name and address, the seller's date of birth and driver's license number or state photo ID details, both signatures, and, since Chapter 484 of 2024, the vehicle identification number of the vehicle the converter came from. Records must be kept electronically for one year and a copy transmitted to the designated primary law enforcement unit by the end of the first business day after the transaction. Section 17–1009(d)(2) also forbids buying a catalytic converter from an individual unless that person identifies as a licensed automotive dismantler and recycler or scrap metal processor, or as an agent or employee of a licensed commercial enterprise.

What is the best software for an auto parts store?

There is no single best one, because the trade is three businesses under one sign: a retail counter, a commercial account book with delivery and returns, and an online catalog that has to answer a fitment question. Most stores end up running a counter system, a supplier ordering network such as PartsTech at $0, $45 or $85 a month, a storefront such as BigCommerce at $29 to $299 a month, and a fitment search layer such as Convermax at $250 to $850 a month. The right question is not which product wins but which of those four layers is worth replacing with something you own.

Why is VIN search so expensive on parts e-commerce platforms?

Because it is usually sold at the top of a tier ladder that is metered by traffic rather than by the feature. Convermax publishes Stage 1 at $250 a month for up to 50,000 monthly users, Stage 2 at $450 for 100,000, and Stage 3 at $850 for up to 1,000,000 users, and VIN search appears only on Stage 3. A Baltimore store with 40,000 sessions a month that wants VIN lookup has to buy a plan sized for a million, which is $7,200 a year more than Stage 1 for one field on a search form.

How many vehicle configurations does an auto parts catalog have to cover?

More than most catalog tools assume. The EPA fuel economy database lists 19,655 distinct year, make and model combinations for model years 2001 to 2026, and 29,679 distinct configurations once engine displacement, cylinder count, drive type, transmission and fuel are included, which is 1.51 configurations per base vehicle. A Ford F-150 alone accounts for 200 year and model rows and 372 configurations; a Honda Civic, 83 rows and 211 configurations. With the average vehicle on US roads at 12.8 years old, a Baltimore store cannot narrow that range much.

What does a custom online parts store cost to build?

Our packages are fixed and published: a one-week Prototype Sprint is $3,500, a custom online store starts at $6,000, and a custom app, internal tool or operations system starts at $12,000. For comparison, a modeled Baltimore stack of BigCommerce Scale plus Convermax Stage 3 plus PartsTech Complete runs $14,808 a year, of which $13,788 is the layer a custom store can actually replace, so a $6,000 build pays for itself in about five months against that stack and in about eighteen months against the cheapest one.