PortalTab

1. One-liner

Catches every OEM portal fee your techs buy mid-job and puts it on the repair order before it invoices.

2. Trend signal — why now?

Three things moved at once, and they moved in the same direction.

The subscription burden got named out loud. Independent shops now juggle a portal per manufacturer. Tony Mercury of Autoshop Solutions, quoted in Fleet Maintenance this summer: “Right now, all-makes shops are juggling a bunch of OEM subscriptions just to cover different vehicles, and it gets expensive and inefficient.” Ben Johnson of Repairify, same piece: “It’s unrealistic to expect an aftermarket shop to buy 22 tools to cover all the manufacturers.” The trade press stopped treating this as a grumble and started treating it as a line item. Fleet Maintenance’s framing is the tell: the fight is “no longer only about whether the door to vehicle data is open — for many shops, the argument is whether the cost of getting the key has become too high.”

The purchase moment fragmented into micro-transactions. OEM data subscriptions aren’t annual anymore. Per DG Technologies, fees “vary by OEM and have several payment options: per-vehicle, 24 hours, 72 hours, weekly, monthly, or annually.” A tech hits a job, discovers they need a 24-hour Ford or GM window, puts $50 on a card, does the flash, moves on. Add AutoAuth’s gateway layer on top — $50/year per shop plus $60/year per vehicle brand for Autel users — and NASTF security transactions at $100–150 per scan. These are small, frequent, card-funded, and invisible to the shop’s books until the statement lands.

Federal law is about to make this the permanent shape of the industry, not a temporary one. On 10 June 2026 the House Energy and Commerce Committee advanced the Motor Vehicle Modernization Act (H.R. 7389), codifying the 2014 MOU into federal law. Read what that actually means for a shop owner: access is guaranteed, but paid access is guaranteed. The bill locks in the pay-the-OEM-portal model rather than dismantling it. Rep. Dunn’s complaint — that the marked-up version “does not fully reflect the original REPAIR Act and fails to protect… independent repair shops” — is the founder’s buy signal. Nobody is coming to make these fees go away. They’re now structural.

So: fees are frequent, fragmented, growing, card-funded, and about to be written into statute. And the shop is eating them.

Provenance:

• Signal 1 (demand): Shop-owner and vendor complaints that all-makes shops juggle dozens of OEM subscriptions, each with its own pricing model and interface; “unrealistic to expect an aftermarket shop to buy 22 tools” — https://www.fleetmaintenance.com/equipment/government-and-regulations/article/55384909/right-to-repair-the-cost-of-access — July 2026
• Signal 2 (feasibility): OEM subscriptions now sold per-vehicle / 24h / 72h / weekly, i.e. as frequent micro-purchases that a card feed can capture; AutoAuth at $50/shop/yr + $60/brand/yr; NASTF scans at $100–150 each — https://www.dgtech.com/j2534-reprogramming-faq/ and https://www.fleetmaintenance.com/equipment/government-and-regulations/article/55384908/right-to-repair-independent-shops-access-to-repair-data — July 2026
• Signal 3 (economic): H.R. 7389 advanced out of House Energy & Commerce on 10 June 2026, codifying the MOU and locking in paid-portal access as the permanent industry structure — https://www.sema.org/news-media/enews/2026/22/us-house-committee-advances-legislation-enact-right-repair-provisions — June 2026 Category: Workflow automation

3. The opportunity

The gap is not “shops can’t access OEM data.” That fight is being fought in Congress and the shops are more or less winning it. The gap is that shops pay for the access and then fail to charge for it.

Here’s the actual sequence. A 2019 Ram comes in with a module fault. The tech gets 75–80% of the way through on the aftermarket scan tool — Shari Pheasant of A Master Mechanic in Nevada describes exactly this — then hits the security gateway. To finish, someone buys a short-window OEM subscription. Maybe $40, maybe $175, depending on the brand and the window. It goes on the shop credit card or the service manager’s personal card. The flash completes. The car goes out.

Now: was that $175 on the invoice?

Overwhelmingly, no. Parts get marked up because parts flow through the parts system and the parts system talks to the repair order. Sublet gets billed because sublet has a vendor and an AP entry and Tekmetric has a whole sublet workflow for it. But a 24-hour Ford subscription bought on a card at 2pm is nothing — it’s a line on a credit card statement thirty days later, reconciled by a bookkeeper who has no idea which car it belonged to. It never touches the repair order. It is pure margin leak, and it leaks on exactly the jobs that are already the hardest and least profitable.

Incumbents don’t fix this because they’re all pointed elsewhere. ALLDATA and Mitchell 1 sell you the data subscription — they are the cost, not the cure. Tekmetric ($179–409/mo) and Shop-Ware have sublet workflows and shop-supply percentage fees, but both require a human to remember to create the line item; nothing watches the card. WickedFile is the closest thing in spirit — it hunts profit leaks in parts, cores, and vendor invoices by reconciling AP against repair orders — but it’s built around the parts vendor invoice, not around a tech buying a 72-hour portal window on a Visa. RepairLogic and CCC Repair Methods consolidate procedure lookup, mostly for collision, which is a different problem (finding the doc, not paying for it).

Nobody owns the moment where money leaves the shop for OEM access. That’s a small, ugly, unglamorous seam — which is exactly why it’s still open.

4. Target market

Primary customer: Owner or service manager of an independent, all-makes mechanical repair shop in the US. 2–8 bays, 3–12 employees, $600K–$3M annual revenue. Does its own diagnostics and module programming rather than sublettting everything to the dealer. Already runs Tekmetric, Shop-Ware, Mitchell 1, or Protractor. The tell: they own a J2534 pass-thru device and have an AutoAuth account.

Why they buy: Because they can feel the leak but can’t measure it. The shop owner knows they spent “a few grand” on OEM portals last year and knows almost none of it made it onto an invoice, but has no way to say which jobs, which cars, which techs, or how much. Net margins in this business run 10–30%; labor margin is where the money is, and every unbilled $175 portal fee comes straight off the bottom line on a job that already burned diagnostic hours. The Fleet Maintenance line lands because it’s true: shops are now asking “whether the cost of getting the key has become too high,” and they’re asking it without knowing the number.

Rough TAM reasoning: There are on the order of 160K independent repair establishments in the US. Cut to all-makes shops that actually do their own programming and carry a J2534 device and an AutoAuth registration — AutoAuth reports meaningful shop registration, and NASTF counts ~11,000 registered vehicle security professionals doing 300,000+ security transactions annually. That’s a defensible proxy for the serious end: call it 15K–25K shops that buy OEM access with any regularity. At $149/mo that’s a $27M–$45M ceiling. Small. Fine — the target here is $1–3M ARR, not a category-defining outcome.

Why now for them: The number of jobs requiring a gated OEM transaction is going up every model year, not down, and H.R. 7389 just told them it’s permanent. The leak is compounding.

5. Product sketch (MVP)

• Card feed watch. Connect the shop’s credit card (Plaid, or a Ramp/Brex-style virtual card issued per-tech). Every charge from a known OEM portal, AutoAuth, or NASTF is flagged within minutes of the swipe.
• Charge-to-RO matching. The moment a portal charge appears, match it to the open repair order it belongs to — using the timestamp, the tech who bought it, the vehicle’s make, and which ROs were live in the shop management system at that hour.
• The 30-second confirm. Push the tech a text: “$175 Ford Motorcraft, 2:14pm — is this the Kramer F-250 (RO 4471)?” One tap yes. Wrong guess, they pick from a short list. That’s the entire tech-facing surface area.
• Auto-line-item. Confirmed charge writes straight into the repair order in Tekmetric/Shop-Ware as a billable line with the shop’s configured markup, before the invoice closes. That’s the whole point — it has to land before the customer pays, not in a report afterward.
• Leak report. Weekly: dollars spent on OEM access, dollars recovered, dollars still unbilled, broken out by tech and by make. The number the owner has never seen.
• Subscription overlap alarm. Flags when a shop buys a 24-hour window for a brand it already holds an annual subscription on, or buys the same brand three times in a month (buy the month instead).
• Renewal watch. Tracks the annual portal subscriptions and AutoAuth brand fees so nothing auto-renews unnoticed on a make the shop stopped servicing.

6. AI angle — what’s load-bearing

The load-bearing AI is the matching, and it’s genuinely hard enough to matter.

A credit card feed gives you a merchant string and an amount — something like MOTORCRAFT SVC*WEB 800-XXX MI and $175.00. That’s it. There’s no vehicle, no RO, no tech. To turn that into a billable line you have to reason across: the merchant string (which OEM is this, actually — the descriptors are a mess and vary by processor), the timestamp, which techs were clocked in, which ROs were open, which of those ROs are on a vehicle of that make, and which of those had a diagnostic or programming job on the ticket. Then decide whether you’re confident enough to auto-write the line or you need to ask a human.

That’s a messy multi-signal inference over dirty text and sparse context. A rules engine gets you maybe 60% and mis-bills the rest, and mis-billing a customer is the one failure mode that gets you fired. An LLM over the merchant string plus the shop’s live RO context gets high confidence on the easy cases (one Ford open, one Ford charge) and — critically — knows when it isn’t sure and asks. The confidence gate is the product.

Strip the AI out and you get a spreadsheet the owner has to fill in by hand at month-end, which is precisely the thing they already don’t do. The reason the leak exists is that the manual version costs more attention than the money it recovers. AI is what pushes the recovery cost below the recovery.

Second-order: after ~90 days of the shop’s own confirmed matches, you’ve learned that shop’s patterns — which techs buy what, which makes come in, what a normal charge looks like. Auto-confirm rate climbs and the tech gets pinged less. That compounding is the only moat here, and it’s a thin one.

7. Localization angle (if any)

N/A — this is a US-only play, and deliberately so. The whole product is shaped by American structures that don’t transfer: the 2014 MOU, H.R. 7389, AutoAuth, NASTF, the FCA security gateway, and the specific US OEM portal pricing model. The EU has its own right-to-repair regime with different data-access rules (and a Euro-5/6 RMI framework that already mandates cheaper standardized access), so the same wound doesn’t bleed the same way. India and SEA independent shops don’t do gated OEM flashing at anything like this volume. Forcing a geography angle here would be pretending.

8. Business model — path to $1M–$5M ARR

• Pricing: $149/mo per shop location, flat. No per-seat — techs must not be a cost to add, or the shop won’t onboard them all and the matching data goes sparse.
• Why $149 sticks: It’s below Tekmetric’s own $179 entry tier, so it reads as an add-on rather than a second platform decision. And the pitch is arithmetic, not vibes: if the shop recovers one $175 portal charge a month, the tool is free. Most shops doing real programming work buy several a month. This is the rare SMB tool where you can show ROI on the demo call from the prospect’s own card statement.
• ACV: $1,788.
• Math to $1M ARR: 560 shops. Against a serious-end pool of 15K–25K, that’s roughly 2–4% penetration. Reachable, not trivial.
• Math to $5M ARR: 2,800 shops — that’s 11–19% of the realistic pool, which I don’t believe happens on the core product alone. Getting to $5M means expanding scope: multi-location groups at higher tiers, adding sublet-to-dealer cost capture (the same leak, different vendor), and eventually issuing the virtual cards yourself and taking interchange on the ~$3–8K/yr each shop routes through OEM portals. Interchange is the honest path to $5M; the SaaS fee alone probably tops out around $2–3M.
• Expansion path: Location count first (multi-shop groups are common and buy centrally), then card interchange, then the same capture logic pointed at every other unbilled shop expense — towing, mobile diag calls, ADAS calibration sublet.

9. Go-to-market wedge — first 100 customers

This market is small, listed, and reachable, which is the main reason it’s worth doing.

• The AutoAuth / NASTF list is the buyer list. A shop that registered with AutoAuth or has a NASTF VSP credential has self-identified as one that pays OEM access fees. That’s a qualified lead by definition. NASTF’s ~11,000 registered VSPs is the densest concentration of this customer that exists. Get the registry, match to shop names and addresses, work it.
• Open with the number, not the pitch. Cold outreach that says “we help you track subscriptions” gets deleted. Cold outreach that says “send me last month’s card statement and I’ll tell you exactly how much OEM portal spend never hit an invoice” gets replies, because the owner genuinely doesn’t know and genuinely wants to. Do the first 30 by hand, free, in a spreadsheet. That’s the audit-as-lead-magnet play and it’s the whole wedge — you deliver the diagnosis before you sell the cure. Convert at 20–30%, because the audit is the demo.
• Diagnostic trainer channel. The independent diag world runs on a small number of trusted trainers and communities — Diag.net, ScannerDanner, the AVI/CTI training circuit, regional ASA chapters. These people have direct, credible, non-advertising access to exactly the shops that do their own programming. Sponsor, show up, do the free audit for their audiences. One respected trainer’s endorsement moves more shops than $20K of ads in this business.
• Shop management system marketplaces. Tekmetric and Shop-Ware both have integration directories. Being the “recover your OEM subscription spend” app in the Tekmetric marketplace is durable, low-cost inbound from customers who are already the right profile. This is slower — expect months, not weeks — so it’s the third leg, not the first.
• Trade press has already written your ad copy. Fleet Maintenance, Ratchet+Wrench, Vehicle Service Pros are actively covering the cost-of-access story. Pitch them the data: “we audited 50 shops and found $X of unbilled OEM access per shop per year.” That’s a story they want and it’s a story only you have.

10. Build complexity — justification

Low. Everything is off the shelf. Card transaction feed is Plaid or a card issuer’s API. Tekmetric and Shop-Ware both expose APIs for repair orders and line items. The tech confirmation is a text message. The matching is an LLM call over a structured prompt — merchant string plus the shop’s open ROs — with a confidence threshold. There is no custom model, no hardware, no scraping of OEM portals (deliberately — that’s a legal and maintenance swamp, and we don’t need it; the card charge tells us what we need).

The real work isn’t technical, it’s the merchant-string dictionary: learning how each OEM’s portal, AutoAuth, and NASTF actually appear on a card statement across processors. That’s grunt work, it’s finite, and it gets better with every shop you onboard. Solo builder ships v1 in 6–8 weeks. First integration (Tekmetric) is the long pole.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Reading the shop’s own card feed with their consent. No OEM portal scraping, no ToS violation, no access circumvention.
Ethical — no harm / dark patterns	✅	Helps a small business bill a cost it legitimately incurred. Markup is the shop’s own configured policy, not ours. If anything it makes shop invoices more itemized and honest than the current “bury it in the labor rate” default.
Market exists (evidence above)	✅	Named shops and vendors complaining on the record; AutoAuth/NASTF pricing is public; H.R. 7389 makes it permanent.
1–5 person team can build this	✅	Solo builder, 6–8 weeks.
Launchable with <$50K / ₹40L	✅	Well under. Cost is the founder’s time plus API fees.

All five pass.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Real money, leaking weekly, on the shop’s least-profitable jobs. Not hair-on-fire — the shop survives without fixing it, which is exactly why it’s gone unfixed — but every owner recognizes it instantly and it recurs constantly. Docking it 4 because it’s a margin problem, not an existential one.
Demand evidence	15	11/15	Strong: named industry figures complaining on the record, public pricing showing the micro-transaction shape, federal legislation making it permanent. Weak: I could not find shop owners explicitly saying “I fail to bill this” — I’m inferring the leak from the mechanics of card-vs-RO, and that inference is assumption #1 below. Honest 11, not a 14.
Build feasibility	15	13/15	Off-the-shelf all the way. Only friction is SMS API integration and the merchant-string dictionary grind.
Distribution clarity	15	12/15	AutoAuth/NASTF registries are a near-perfect buyer list, and the free-audit lead magnet is unusually strong — you show the prospect their own leak. Docked because the registries may not be cleanly obtainable and the trainer channel takes relationship time.
Revenue mechanics	15	11/15	Pricing is benchmarked and the ROI story is arithmetic. But $1,788 ACV against a 15–25K shop pool means $1–3M is the realistic SaaS ceiling; $5M requires the interchange bet, which is unproven.
Time to first revenue	10	7/10	The free audit converts fast — you can hand-audit shops before the product exists and pre-sell. But shop owners move slowly and the Tekmetric integration gates real value. 8–10 weeks to first dollar.
Defensibility	10	3/10	This is the weak axis and I won’t pretend otherwise. The merchant-string dictionary and per-shop matching history compound a little, but Tekmetric could ship this as a feature in a quarter if they noticed. Execution-and-focus moat only. A profitable niche business, not a defensible one.
Total	100	73/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · sales-heavy

Technical because the matching engine and two SMS integrations are the product. Sales-heavy because this customer does not buy software off a landing page — the free audit has to be delivered by a human who can talk to a shop owner without sounding like a software person. A founder who’s never been in a shop will struggle to get taken seriously. Domain credibility here is bought with time in the bays, or with a trainer/advisor who lends you theirs.

Key assumptions to validate (3–5)

1. Assumption: Shops genuinely fail to bill most OEM portal charges — the leak is real and material ($2K+/yr/shop), not something they already handle via a shop-supplies percentage fee. How to test: This is the assumption and it’s the one I could not source. Get 20 shop owners to hand over one month of card statements plus the matching ROs. Count what was billed. If most shops already bury it in a shop-supplies percentage or a diag rate that covers it, the product has no reason to exist. Do this before writing any code.
2. Assumption: A shop owner will pay $149/mo to recover money they’re currently losing invisibly. How to test: After the free audit, ask for a card on the spot. “I found $2,400 you didn’t bill last year. It’s $149/mo to stop that.” If fewer than 1 in 5 audited shops convert, the pain isn’t priced right — or isn’t pain.
3. Assumption: Techs will actually tap the confirmation text mid-job rather than ignore it. How to test: Run a 2-week manual pilot in 3 shops — a human texts the tech on every portal charge. Measure response rate and latency. Below 70% response and the auto-billing breaks down into another report nobody reads.
4. Assumption: Tekmetric’s API allows writing billable line items to an open RO. How to test: Read the docs, build the integration spike. Day one. If they don’t permit writes, the product degrades to a report and the value collapses.

Risk flags

1. Platform dependency (severe): The entire value chain sits on top of Tekmetric/Shop-Ware write access and a card data feed. Either one closing off, or Tekmetric deciding to build this natively, is fatal. Given defensibility is a 3, this is the thing that kills the company.
2. Assumption risk on the core premise: I’m inferring the unbilled-cost leak from the structure of how these charges are made — card, mid-job, no vendor invoice — not from shop owners saying so directly. It’s a sound inference and it’s how every other card-funded micro-expense behaves. But it is an inference, and if shops already absorb this via a blanket shop-supplies fee, the whole thesis is wrong. Hence the free-audit-first sequencing.
3. Regulatory risk (cuts both ways): If a future REPAIR Act version actually forces OEMs to make access cheap or free, the leak shrinks and so does the product. H.R. 7389’s watered-down form makes this less likely near-term — which is why the political news is a buy signal — but a stronger bill on the floor would be bad news.
4. Market ceiling: 15K–25K serious shops is a small pond. This is a $1–3M ARR business that has to work hard to become $5M. That’s fine if it’s what you want. It is not fine if you need it to be bigger.

14. Structured verdict

Score:                  73/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical solo founder with a shop-world advisor or trainer relationship
Time to revenue:        8–10 weeks
Capital to launch:      $5–8K
Top 3 assumptions to validate first:
  1. The leak is real and material — audit 20 shops' card statements against their ROs before writing code
  2. $149/mo converts off the audit — ask for the card immediately after showing the number; need 1 in 5
  3. Tekmetric API permits writing billable lines to an open RO — docs + integration spike on day one
Kill criteria:
  - Abandon if fewer than 12 of 20 audited shops show >$1,500/yr in unbilled OEM portal spend
  - Abandon if fewer than 4 of the first 20 audited shops convert to paid within 30 days of seeing their number
  - Abandon if Tekmetric ships native OEM-subscription cost capture before v1 ships

15. Next step — 1-week validation sprint

The sprint tests the one thing everything else rests on: is the money actually leaking?

• Day 1–2: Build the target list. Pull shop names from AutoAuth registrations, NASTF VSP listings, and Diag.net’s active membership. Get to 200 all-makes independents that do their own programming. Write the outreach: “I’m researching how much OEM portal spend never makes it onto a repair order. Send me one month of card statements and I’ll do the reconciliation free and send you the number.” Send 200.
• Day 3–4: Take every shop that says yes. Reconcile by hand — every OEM portal, AutoAuth, and NASTF charge against their repair orders for that month. Find out how many were billed. Sit on a call with 5 of them and watch their face when you show them the number. That reaction is the product-market fit signal, and you cannot get it from a survey.
• Day 5: Decide. Go if: ≥20 shops sent statements, and ≥60% of them show more than $125/mo in OEM access charges that never touched an invoice, and at least 4 owners say some version of “how do I stop that.” No-go if: shops turn out to already cover this via a shop-supplies percentage or a diagnostic rate that absorbs it — in which case the leak is a bookkeeping artifact, not lost margin, and there is no business here.

Falsifiable, cheap, and it produces the exact number the sales pitch needs. If the audit says the leak is real, you launch with 20 warm prospects who’ve already seen their own bleeding.