WrenchChat

1. One-liner

Voice-to-WhatsApp AI that turns mechanic jargon into customer-friendly estimates with one-tap approval for UK independent garages.

2. Trend signal — why now?

The UK independent garage market is going through a quiet digital transformation — but the tools available are either too heavy or too generic.

• BookMyGarage processed £134M in retail work in 2025, up 39% YoY, with 92% going to independent garages. 27% of all bookings now happen outside working hours. Customers expect digital-first interaction — garages that can’t deliver lose work.
• WhatsApp Business API matured in the UK market through 2025. Carsu (€29–60/mo, Southern European focus) proved the WhatsApp-first garage model works. UK workshops under 40 now expect text-based communication as default.
• AI voice transcription hit commodity pricing in late 2025. OpenAI Whisper, Deepgram, and AssemblyAI dropped below $0.006/minute. A mechanic voice-noting a diagnosis costs fractions of a penny to transcribe. LLMs can translate “O/S/F lower arm bush worn — advisory, nearside CV boot split” into “The front suspension bushing on the driver’s side is showing wear and the left CV boot has a split — both are advisories for now but will need attention in the next few months” for under $0.01.
• Digital vehicle health checks increase repair approval by 15–25% over verbal or paper-based advisories, per UK GMS vendor data. Yet most 1–3 bay garages still communicate verbally or via handwritten notes — the tools that offer digital VHC are locked inside £150–300+/mo management suites.
• Electronic Receptionist launched at £97/mo for UK garage call answering, validating that garages will pay for AI communication tools. But it only handles inbound calls and booking — it doesn’t touch the estimate-to-approval workflow where the real revenue leaks.

Provenance:

• Signal 1: BookMyGarage £134M in 2025 retail work, 39% YoY growth, 27% after-hours bookings — https://aftermarketinsight.spglobal.com/news/5084/uks-bookmygarage-sees-39-increase-in-bookings-with-record-134-million-retail-work-in-2025 — 2026-01
• Signal 2: AI voice transcription (Whisper/Deepgram) dropped below $0.006/min; LLM inference for jargon translation under $0.01/call — https://www.wickedfile.com/blogs/how-can-auto-repair-shops-use-ai-in-2026 — 2026-03
• Signal 3: Digital VHC reports boost customer approval by 15–25% but are locked inside £150+/mo GMS suites inaccessible to solo mechanics — https://esremedia.co.uk/blog/garage-management-software-uk-comparison — 2026-02 Category: Geographic arbitrage + Workflow automation

3. The opportunity

There are 30,000+ independent garages in the UK. The majority are 1–3 person operations — the mechanic IS the receptionist, the service advisor, and the bookkeeper. They’re losing revenue two ways:

1. Missed calls. When you’re under a car with an angle grinder running, the phone goes to voicemail. 80% of callers don’t leave one and never call back. At an average job value of £200–400, even 3 missed calls/week is £30K–60K/year walking out the door.

2. Low estimate approval. When the mechanic does reach the customer, they explain work in jargon: “cam belt tensioner’s gone, water pump’s weeping, recommend doing both while we’re in there — looking at about £450 parts and labour.” Customer hears £450 and an alien language, says “I’ll think about it,” and never calls back. Garages with digital VHC reports (photos + plain-English explanations) see 15–25% higher approval rates.

The existing solutions are full garage management suites — GarageHive (£200+/mo with Microsoft licences), TechMan (complex UI, “35 clicks to finish an estimate”), Dragon2000. These are designed for larger operations. A solo mechanic in Basingstoke doesn’t need a full ERP — they need a way to talk into their phone, have an AI draft a WhatsApp message, and have the customer tap “Go ahead.”

Electronic Receptionist (£97/mo) proved the market will pay for AI comms, but it stops at call answering. Carsu (€29–60/mo) showed WhatsApp-first works, but it’s built for Southern Europe and is a full GMS, not a lightweight comms layer.

WrenchChat fills the gap: a voice-first, WhatsApp-native AI communications layer that requires zero migration from existing tools. Works standalone or alongside any GMS.

4. Target market

• Primary customer: Owner-operators and lead technicians at 1–3 bay independent garages across the UK. Typically sole traders or limited companies with 1–3 staff, no dedicated service advisor or receptionist, annual turnover £150K–£500K.
• Why they buy: They’re tired of losing jobs to missed calls and unapproved estimates. They know they should text customers instead of calling, but writing customer-friendly messages while covered in brake dust isn’t happening. Every unanswered phone and every “I’ll think about it” is money they worked to generate but couldn’t collect.
• Rough TAM reasoning: 30,000+ independent garages in the UK (SMMT data). If 10% adopt at £79/mo, that’s ~£2.8M ARR. At 20% penetration with a mix of basic (£79) and pro (£149) tiers, that’s £5M+.
• Why now for them: BookMyGarage’s 39% booking growth in 2025 means more customers expect digital communication. Garages that can’t respond quickly via text lose to the one down the road that can. The after-hours booking stat (27%) means customers are shopping when garages are closed — and the garage that sends a WhatsApp confirmation at 9pm wins over the one that calls back at 9am tomorrow.

5. Product sketch (MVP)

• Voice-to-estimate: Mechanic opens app, taps record, says what’s wrong with the car in mechanic-speak. AI transcribes, identifies parts and labour items, generates a customer-friendly WhatsApp message with plain-English explanation and estimated price.
• One-tap approval: Customer receives WhatsApp with diagnosis, price breakdown, and an “Approve Work” button. Tap it, work is authorised. No phone tag.
• Photo attach: Mechanic snaps a photo of the worn part. AI includes it in the customer message with a caption explaining what the customer is looking at.
• MOT advisory follow-up: Import MOT advisories from DVLA’s free API. When advisory items are approaching due, auto-send a WhatsApp reminder to the customer suggesting they book in.
• After-hours auto-reply: When the garage is closed, incoming WhatsApp messages get an AI auto-reply confirming receipt and offering online booking for the next available slot.
• Review nudge: After job completion, auto-send a Google review link via WhatsApp. Independent garages live and die by Google reviews — this is table stakes.
• Simple dashboard: Web dashboard showing pending approvals, sent estimates, approval rate, and revenue recovered from after-hours messages. No GMS — just the communication layer.

6. AI angle — what’s load-bearing

Without AI, this product doesn’t exist. The core value prop depends on three AI capabilities:

1. Voice-to-text transcription of mechanic speech. Mechanics talk fast, use abbreviations, and have regional accents. Whisper/Deepgram handle this well enough for 95%+ accuracy.
2. Jargon-to-plain-English translation. “O/S/F lower arm bush worn” → “The front suspension bushing on the driver’s side is showing wear.” This is the magic. No mechanic is going to type that out. An LLM does it in under a second.
3. Price estimation from voice. The AI matches mentioned parts to a pricing database (initially manual/configurable, eventually auto-populated from parts catalogues like TecAlliance or LKQ Euro Car Parts).

Remove the AI and you have… a WhatsApp Business account with some templates. The AI is what makes a solo mechanic’s 15-second voice note become a professional, customer-friendly estimate.

7. Localization angle (if any)

This is a UK-first play, exploiting specific local conditions:

• MOT system: The UK’s mandatory annual vehicle inspection creates a natural annual touchpoint. DVLA’s free MOT history API provides advisory data that feeds automated follow-up. No other market has this exact mechanic.
• WhatsApp dominance: 75%+ of UK adults use WhatsApp. It’s the default messaging app, unlike the US where SMS still dominates.
• Pricing norms: UK independent garages charge £50–90/hr labour. A £79/mo tool pays for itself if it converts one extra estimate per month.
• Parts catalogue integration: TecAlliance (TecDoc) and LKQ Euro Car Parts dominate UK aftermarket parts distribution. Both have APIs.
• BookMyGarage / WhoCanFixMyCar ecosystem: Online booking platforms already drive traffic to independent garages. WrenchChat handles what happens after the booking — the diagnosis, estimate, approval, and follow-up.

Expansion to Germany (20,000+ Werkstätten), France, and Southern Europe (where Carsu has validated the model) is natural once UK is established.

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

• Pricing: £79/mo Starter (voice-to-estimate, WhatsApp messaging, 100 messages/mo included). £149/mo Pro (unlimited messages, MOT advisory follow-up, Google review nudge, photo reports, priority support).
• ACV: £1,200–1,800/year depending on tier.
• Rough math to $1M ARR (~£800K): 550 garages × £1,450 avg ACV = £800K ARR. 550 garages is 1.8% of the 30,000 UK independent garage market. Achievable in 12–18 months.
• Rough math to $5M ARR (~£4M): 2,500 garages × £1,600 avg ACV = £4M. 8.3% market penetration. Requires adding Germany or France as a second market, or moving up to multi-site independents.
• Expansion path: Per-message overage charges (WhatsApp API costs are pass-through + margin). Parts catalogue integration upsell (auto-price estimates from live wholesale data). Multi-location dashboard for garage groups. Referral revenue from BookMyGarage integration.
• Unit economics: WhatsApp Business API costs ~£0.04–0.08/message in UK. At 100 messages/mo on Starter, COGS is ~£4–8. AI inference (Whisper + LLM) at ~£0.02/estimate, 200 estimates/mo = £4. Total COGS ~£10–15/mo per customer. Gross margin 80%+.

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

1. Scrape BookMyGarage and WhoCanFixMyCar listings. Both platforms list independent garages with reviews, location, and contact details. Filter for 1–3 bay operations in English regions with high Google review activity (signals digital savviness). Send a personalised WhatsApp message (ironic, yes) showing what a WrenchChat estimate would look like for their type of work. Target: 2,000 garages, expect 5% response rate, 20% conversion = 20 customers.
2. Partner with 2–3 parts distributors. LKQ Euro Car Parts, GSF Car Parts, and Andrew Page have reps who visit independents weekly. Offer a rev-share on referrals (£10/mo kickback per active customer). Their reps already have trust and walk-in access. Target: 30 customers from distributor referrals in first 6 months.
3. Autotechnician magazine and PMM sponsorship. These are the two trade publications every UK independent reads. Sponsor a “garage efficiency” column with real case studies. Cost: £500–1,000/month. Target: 20 inbound leads/month.
4. IGA (Independent Garage Association) partnership. 3,500+ member garages, annual conference, monthly newsletter. Offer IGA members a free 30-day trial. Target: 200 trials → 30 paying customers.
5. YouTube demo channel. Short videos showing a real mechanic voice-noting a diagnosis → customer receiving a polished WhatsApp estimate in 10 seconds. UK garage owners watch YouTube for tool reviews. Target: 5,000 views/video → 10 customers/month from organic.

10. Build complexity — justification

Medium. The core tech stack is off-the-shelf: Whisper API for voice transcription, GPT-4o-mini or Claude Haiku for jargon translation, WhatsApp Business API via Twilio or 360dialog, DVLA MOT API for vehicle data. Custom work needed: the parts/pricing database (initially a spreadsheet, later TecAlliance API integration), the approval workflow (simple stateful messaging flow), and the mobile-first web app for the mechanic-facing side. A two-person team (full-stack dev + product/sales) ships a credible v1 in 10–12 weeks.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	No regulated advice. Standard B2B SaaS. WhatsApp Business API ToS compliant for service businesses. GDPR-compliant data handling required — standard.
Ethical — no harm / dark patterns	✅	Helps mechanics communicate more clearly with customers. No dark patterns. Customers can opt out of WhatsApp at any time.
Market exists (evidence above)	✅	30,000+ UK independents. £134M flowed through BookMyGarage in 2025. Electronic Receptionist proved £97/mo price point.
1–5 person team can build this	✅	Off-the-shelf AI APIs, standard web stack, WhatsApp Business API. Two people, 10–12 weeks.
Launchable with <$50K / ₹40L	✅	£15–20K for 3 months of two-person build + WhatsApp API setup costs (~£500) + initial marketing budget (£2–3K). Well under £50K total.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	17/20	Mechanics lose jobs daily to missed calls and jargon-heavy estimates. Monetary loss is direct and measurable (£30–60K/year per garage from missed calls alone).
Demand evidence	15	12/15	BookMyGarage’s £134M and 39% growth proves digital demand. Electronic Receptionist at £97/mo proves willingness to pay for AI comms. Digital VHC data shows 15–25% approval lift. No direct competitor in the lightweight voice-to-WhatsApp-estimate niche — demand is inferred, not proven with pre-sales.
Build feasibility	15	13/15	Commodity APIs throughout: Whisper, GPT-4o-mini, WhatsApp Business API, DVLA MOT API. Standard web stack. Main integration challenge is parts pricing — solvable with manual entry initially.
Distribution clarity	15	12/15	Named channels: BookMyGarage/WhoCanFixMyCar scrape, parts distributor reps, IGA membership, trade press. Conversion math is speculative but channels are concrete and cheap.
Revenue mechanics	15	13/15	£79–149/mo is well within independent garage wallet (they pay £97/mo for Electronic Receptionist, £60–300/mo for GMS). Gross margin 80%+ at these volumes. Path to £800K ARR needs only 550 garages (1.8% penetration).
Time to first revenue	10	8/10	10–12 week build, then 30-day free trials converting to paid. First revenue at ~week 16–18. Could be faster with a concierge MVP (human-in-the-loop for first 10 customers).
Defensibility	10	5/10	Low structural moat early on. Any GMS could bolt this on. Moat builds through: UK parts pricing database, garage-specific AI fine-tuning, workflow lock-in (customers approve via your links), and distribution partnerships with parts distributors.
Total	100	80/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy — needs a builder who can integrate voice AI, LLMs, and WhatsApp Business API. Sales to garages is relationship-driven but not enterprise-complex — most sales happen via demo + free trial.

Key assumptions to validate (3–5)

1. Assumption: Solo mechanics will adopt a new tool on top of (or instead of) their existing workflow. How to test: Offer 20 garages a free 2-week trial with white-glove onboarding. Measure daily usage after day 7 — if >60% are still sending estimates through WrenchChat, adoption friction is manageable.
2. Assumption: Voice-to-estimate accuracy is good enough without custom training. How to test: Record 50 real mechanic voice notes from 5 different garages, run through the pipeline, have a qualified mechanic rate accuracy. Target: 90%+ acceptable without manual correction.
3. Assumption: Customer approval rates increase measurably when switching from phone calls to WrenchChat WhatsApp estimates. How to test: A/B test with 10 garages — half their estimates via traditional method, half via WrenchChat. Measure approval rate delta over 4 weeks.
4. Assumption: Parts distributor reps will actively refer the tool. How to test: Approach 3 local LKQ reps with a referral incentive. Track whether they mention it in garage visits. If <2 referrals per rep per month, this channel is dead.

Risk flags

1. Platform dependency: Heavy reliance on WhatsApp Business API. Meta could change pricing, rate limits, or ToS. Mitigation: also support SMS fallback via Twilio, and email for customers who prefer it. But WhatsApp is the wedge.
2. GMS incumbents bolting it on: GarageHive or TechMan could add voice-to-estimate as a feature. Mitigation: they’d add it as a feature inside their complex platform. WrenchChat’s advantage is simplicity and zero-migration — it works for garages that DON’T want a full GMS.
3. Pricing pressure: At £79/mo, there’s limited room to undercut. If a free or freemium competitor appears, the value prop weakens. Mitigation: lock in with parts pricing data and workflow state (customer approval history, MOT follow-up chains) that create switching costs.
4. Regional accent accuracy: UK has significant dialect variation. “Wishbone’s knackered” in a Glaswegian accent is a different transcription challenge than the same phrase in RP English. Mitigation: use Deepgram’s UK English model; fine-tune with early customer voice data.

14. Structured verdict

Score:                  80/100
Verdict:                STRONG GO
Confidence:             Medium
Best-fit builder:       Technical founder with UK automotive aftermarket awareness (or willingness to immerse)
Time to revenue:        16–18 weeks (12-week build + 4-week free trial → paid conversion)
Capital to launch:      £15–20K ($19–25K)
Top 3 assumptions to validate first:
  1. Solo mechanics will voice-note estimates daily (test: 20-garage free trial, measure day-7 retention)
  2. Voice-to-estimate accuracy ≥90% without custom model training (test: 50 real voice notes scored by a mechanic)
  3. WhatsApp estimate approval rate exceeds phone-call approval rate by ≥10pp (test: 4-week A/B with 10 garages)
Kill criteria:
  - Abandon if <30% of trial garages send ≥3 estimates/week after week 2
  - Abandon if voice-to-estimate accuracy is <80% after testing with 5 regional accents
  - Abandon if 50 cold outreach attempts to garages yield <5% interest rate

15. Next step — 1-week validation sprint

• Day 1: Build a throwaway prototype: WhatsApp Business sandbox + Whisper API + Claude Haiku. Record 10 sample mechanic voice notes (recruit one local garage owner for an hour, buy them lunch). Run through the pipeline. Evaluate output quality.
• Day 2: Take the 10 AI-generated estimates to 5 non-mechanic friends. Ask: “Would you understand what’s wrong with your car from this message? Would you approve this work?” Score comprehension and trust.
• Day 3–4: Cold-contact 30 independent garages within 30 miles via WhatsApp (they’ll appreciate the irony). Message: “I’m building a tool that turns your voice notes into customer-friendly WhatsApp estimates. Would you try it free for 2 weeks?” Track response rate, interest level, and objections.
• Day 5: Decide go/no-go. Go if: ≥5 garages want to try it AND prototype output quality is rated ≥7/10 by non-mechanic testers. No-go if: <3 garages respond OR prototype output quality requires heavy manual correction (>30% of estimates need editing).