Reparo — ERIF quoter for EU repair shops

1. One-liner

Turns a fault photo and model plate into a binding, compliant EU repair quote form in under a minute.

2. Trend signal — why now?

The EU Right to Repair Directive (2024/1799) starts applying in every member state on 31 July 2026 — nine days from the date on this proposal. From that moment consumers gain the right to request a repair, and repairers can hand out a standardised European Repair Information Form (ERIF). If a repairer issues one, its terms — cost, expected duration, availability of a temporary replacement — are binding for 30 days, and consumers can lay two ERIFs side by side and pick the better offer. The whole point of the form is to make repair offers comparable, which turns a slow or sloppy quote into a lost job.

The buyer here — the independent repair shop — is a small business that has never had to produce a formal, binding, comparable quote on demand. Today they eyeball the fault, guess the part, guess the labour, and text back a number. That workflow does not survive a regime where the customer is comparing your form against the shop down the street and can hold you to the price for a month.

Provenance:

• Signal 1 (demand): EU Right to Repair Directive applies 31 Jul 2026; ERIF terms binding 30 days; form exists so consumers can compare offers — europa.eu / consilium.europa.eu — 2026-03/07
• Signal 2 (feasibility): Multimodal vision+LLM (e.g. Gemini Pro Vision) already reads a tattoo reference or appliance photo → structured estimate; used in shipping consumer tattoo/appliance price tools in 2026 — poliinternational.com / inkjin.com — 2026-07
• Signal 3 (economic): Independent repair sector is a named beneficiary of the directive; repair-shop SaaS (RepairDesk 2,700+ shops, RepairShopr) already sells $60–130/mo per shop, but none generate an ERIF — capterra.com / softwareadvice.com — 2026 Category: Regulatory arbitrage

3. The opportunity

The directive creates a brand-new document that did not exist on 30 July 2026 and is mandatory-to-be-good-at by 31 July 2026: a binding, standardised, comparable repair quote. Existing repair-shop software (RepairDesk, RepairShopr, RepairFlow) is ticketing + POS + invoicing built for phone-and-computer shops. It does not:

• read a customer’s fault photo and model plate and identify the exact model,
• look up the mandated spare-part price and a defensible standard labour time,
• assemble a compliant ERIF in the customer’s language, ready to send.

The incumbents run the shop after the job is booked. Nobody owns the 60 seconds before the job is booked — the quote — which is exactly the moment the new law weaponises. A focused AI-first tool can own that moment and do it 10× faster than a technician squinting at a parts PDF.

4. Target market

• Primary customer: Independent appliance, electronics and small-domestic repair shops in the EU — 1 to 8 staff, owner-operator or a couple of techs. The directive explicitly names “independent and community repairers” as ERIF users. Germany, France, Netherlands, Poland, Italy first (largest independent-repair populations, strongest enforcement culture).
• Why they buy: From 31 July their customers can request and compare quotes. A quote that’s slow, wrong, or non-comparable loses the job to the shop next door — and a binding quote that under-priced the part is money out of the owner’s pocket for 30 days. They need to produce a good ERIF fast, and they’ve never had to before.
• Rough TAM reasoning: Precise counts are hard, but the independent repair trade across the EU is tens of thousands of commercial shops (Germany alone runs ~1,800 repair clubs/cafés on top of its commercial base; existing repair SaaS already serves 2,700+ shops globally). A few thousand paying EU shops at €39–99/mo is a comfortable sub-$5M-ARR pond.
• Why now for them: The obligation and the comparison pressure both switch on the same day, 31 July 2026. There is a hard, calendar-driven urgency — the rarest thing in SMB software.

5. Product sketch (MVP)

• Snap a photo of the appliance + the model/rating plate → auto-identify make and model.
• Type or dictate the fault (“drum won’t spin, humming noise”) in any EU language.
• Tool returns a draft ERIF: likely part(s), part price, standard labour time, total, expected duration, temporary-replacement note.
• Shop owner reviews, tweaks the labour rate/margin, and one-taps to send a clean PDF/link ERIF to the customer.
• ERIF is stamped with the 30-day validity clock and stored, so the shop has a defensible record of what it quoted and when.
• Rate-card learning: the shop’s own accepted quotes feed back so estimates match their pricing, not a generic average.
• Language + currency localisation per member state, with the standardised EU field layout baked in.

6. AI angle — what’s load-bearing

Remove the AI and there is no product. The core is a multimodal chain: vision identifies the model from a plate photo, an LLM maps the free-text fault to the probable failing component, and a pricing step assembles a defensible ERIF from a spare-part catalogue plus a standard-repair-time estimate, in the shop’s own rate logic and the customer’s language. A dumb form-filler would still leave the hard 90% — “what’s broken, what part, how long, what’s it cost” — on the technician. The AI is the 90%.

7. Localization angle (if any)

Localization is the moat, not a nice-to-have. This is an EU-first product by construction: the ERIF field layout is EU-defined, the pricing must be in local currency, the fault-to-part reasoning must speak German, French, Polish, Italian, Dutch, and the temporary-replacement and 30-day-validity rules must match each member state’s transposition. A generic US repair-POS tool cannot serve this without rebuilding around the directive — which is exactly the wedge.

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

• Pricing: €49/mo per shop (solo), €99/mo (multi-tech), usage add-on above N quotes/mo. Benchmarked against RepairDesk (~€99) and RepairShopr (€55–120) — but Reparo sits upstream of them, so it can be an add-on rather than a rip-and-replace.
• ACV: ~€700–900/shop/year blended.
• Rough math to $1M ARR: ~1,300 shops × €65/mo × 12 ≈ $1.0M (at ~€/$ parity).
• Rough math to $5M ARR: ~5,500 shops, or fewer shops plus a per-quote overage and a white-label deal with one national repair-shop association or a spare-parts distributor bundling Reparo to its dealer base.
• Expansion path: per-quote overage → chain/franchise seats → spare-parts affiliate margin (the tool already knows the part; buying it through the tool is one tap) → adjacent categories as the directive adds product scope.

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

• Ride the deadline panic. For the 60 days around 31 July 2026, run a “Are you ERIF-ready?” free single-shop audit — upload one repair, get a sample compliant ERIF back. It’s a demo and a scare in one.
• Scrape and DM the directories. Google Business + national repair-shop directories (e.g. Germany’s ZVEH/handwerk listings, France’s Annuaire de la réparation, RepairMonitor networks) list independents by trade and city. Cold-email 2,000 with a pre-made ERIF for a common model in their language; expect 3–5% to book a demo.
• Association + distributor partnerships. National repair-craft associations and spare-parts distributors are actively publishing “get ready for the directive” content. One co-branded webinar or newsletter placement puts the tool in front of hundreds of anxious shop owners at once.
• Right-to-repair community channels. The Repair Café / repair-movement blogs and shop-owner forums (RepairPlugin, FixFirst readerships) are already asking “what does this mean for my shop” — answer it with the tool.

3–5 named channels, each with realistic conversion math. First 100 is a two-month sprint riding a legal deadline.

10. Build complexity — justification

Medium. Off-the-shelf: multimodal model, LLM, standard web stack, PDF/e-sign, WhatsApp/email delivery. The custom grind is (a) a spare-part price + standard-labour-time reference good enough to make a binding quote defensible, and (b) the ERIF field layout and 30-day/replacement rules per member state. That’s domain work with a repair-trade advisor and access to a parts catalogue, not a research breakthrough. Estimate 10–14 weeks to a Germany-first v1 for a technical founder paired with a repair-shop domain expert.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Tool helps comply with the directive; no approval needed to sell it.
Ethical — no harm / dark patterns	✅	Increases quote transparency for consumers; aligned with the law’s intent.
Market exists (evidence above)	✅	Directive live 31 Jul 2026; repair-SaaS already sells to thousands of shops.
1–5 person team can build this	✅	Multimodal + web stack; domain data is the only heavy lift.
Launchable with <$50K / ₹40L	✅	No hardware, no capex; parts-data access is the main cost.

All five pass.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	13/20	Real, calendar-driven, and it costs money to get wrong — but many shops will muddle through with a template and shrug at first.
Demand evidence	15	10/15	Strong regulatory signal + existing paid repair-SaaS, but no direct evidence shops will pay specifically for ERIF generation yet. It’s brand new.
Build feasibility	15	11/15	Stack is off-the-shelf; the binding-quote parts/labour data is the risk.
Distribution clarity	15	10/15	Named directories + association channel + a hard deadline; conversion still unproven.
Revenue mechanics	15	11/15	Pricing benchmarked, add-on positioning helps, ACV modest.
Time to first revenue	10	7/10	Deadline creates urgency; short trial-to-paid plausible in weeks.
Defensibility	10	4/10	Thin. Incumbent repair-SaaS could bolt this on; the shop’s own rate-card data is the only compounding edge.
Total	100	66/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required — needs multimodal build chops and a repair-trade advisor who knows parts, labour times, and how the directive lands in at least one member state.

Key assumptions to validate (3–5)

1. Assumption: Independent shops will pay a separate tool for quote generation rather than wait for RepairDesk/RepairShopr to add it. How to test: 20 shop interviews in Germany/NL; offer a paid pilot at €49/mo, count how many put a card down.
2. Assumption: A photo + fault text yields a part/labour estimate accurate enough for a binding 30-day quote. How to test: blind-test the estimate on 100 real historic repair tickets against what the shop actually charged; measure error band.
3. Assumption: The comparison pressure is real enough that a slow quote actually loses jobs. How to test: interview shops post-31-July on whether customers are bringing competing ERIFs.
4. Assumption: Parts-price + standard-labour-time data is acquirable at acceptable cost/coverage. How to test: scope one distributor feed or catalogue licence for a single appliance category.

Risk flags

1. Platform/incumbent dependency: RepairDesk (2,700+ shops) or RepairShopr can add ERIF generation as a feature and blunt the wedge within a year. The defence is speed + a better rate-card loop + distributor bundling, not IP.
2. Regulatory ambiguity: ERIF use is a right, not (for most shops) a strict obligation on day one, and member-state transpositions vary. The “must-have” urgency may soften if enforcement is light early. This is the biggest reason for a VALIDATE, not a GO.
3. Liability of a binding quote: if the AI under-prices a part and the shop is held to it for 30 days, the shop eats the loss — and blames the tool. Estimate error bands and clear “review before send” UX are non-negotiable.

14. Structured verdict

Score:                  66/100
Verdict:                VALIDATE
Confidence:             Medium
Best-fit builder:       Technical founder + EU repair-trade domain advisor
Time to revenue:        6–10 weeks (deadline-driven pilots)
Capital to launch:      €5–15K ($6–17K) — mostly parts-data access
Top 3 assumptions to validate first:
  1. Shops pay for standalone quote generation — 20 interviews + paid pilot with card-on-file
  2. Photo+fault → binding-grade estimate — blind test on 100 historic tickets, measure error band
  3. Comparison pressure loses jobs in practice — post-deadline shop interviews
Kill criteria:
  - Abandon if <15% of 40 interviewed shops will pay €49/mo after seeing a sample ERIF
  - Abandon if estimate error band exceeds ±25% on the historic-ticket test (too risky for a binding quote)
  - Abandon if a major repair-SaaS ships native ERIF generation before your v1

15. Next step — 1-week validation sprint

• Day 1–2: Build a thin demo — photo + fault text in → sample ERIF PDF out — for one appliance category (washing machines) in German. Hand-curate the parts/labour data; no real backend needed.
• Day 3–4: Get 15 real historic repair tickets from 2–3 friendly shops. Run the demo blind and compare its estimate to what they actually charged; record the error band.
• Day 5: Show the sample ERIF to 20 shop owners (cold-DM’d from a national directory) and ask for a €49/mo pre-order.
• Decide: Go only if the historic-ticket error band is within ±25% and ≥3 of 20 shops put money down. Both are falsifiable; either miss = no-go.