AI Repair Intake Marketplace

Opening

Most home repair marketplaces start from a weak assumption: the homeowner already knows who to call.

Plumber. Electrician. Appliance repair. Handyman.

But real life usually does not work that cleanly. When something breaks at home, the homeowner often goes through a messy loop: asking friends, posting in a local Facebook group, Googling symptoms, watching YouTube videos, calling a few numbers they found online, choosing a contractor with decent reviews and hoping it is the right person.

That is not a clean buying journey. It is trial and error.

The problem is that even contractors do not always know the root cause from the first call. A good electrician, plumber, or appliance technician may still need to inspect, test, rule out causes, replace parts, or order components from a manufacturer. Home repair does not always have a clear answer at the first touchpoint.

So the real problem is not just “find a pro.” The better problem is: how can we help homeowners understand the repair issue clearly enough before deciding whether to fix it themselves, call a contractor, or post a job?

That is the thesis behind Wrenchy: the repair request is the unit of value.

A clearer repair request can help all three sides — homeowners waste less time guessing, contractors receive better-quality leads, and the marketplace has a stronger basis for matching, quoting, and monetization.

Market observation

Home repair is a large market, but it is messy.

Repair costs can be high. In some cases, calling a contractor can cost almost as much as replacing a cheaper appliance or fixture. Because of that, many homeowners naturally ask: can I fix this myself?

But they often do not have enough information to answer the next set of questions. Is the issue dangerous? Is it safe to attempt a DIY repair? Which trade should I call? Which part is compatible? How long should this take? What can go wrong if I try to fix it? What is a reasonable contractor cost?

Research takes time. A common pattern: something looks like a simple cleaning issue at first. After several failed attempts, the homeowner realizes the issue may be mechanical, electrical, model-specific, or part-related. Then they run into another problem — the YouTube video is for the wrong model, the part does not match, or the product design changed across versions.

That is the gap many marketplaces do not solve well. They help people find someone to do the work. They do not always help the user understand the problem before looking for someone.

Existing marketplace pattern

Platforms like Jiffy improve access. The homeowner can find a service provider faster, follow a clearer booking flow, and feel more trust than they might from a random Google search. But a managed marketplace has a cost — the more the platform controls pricing, scheduling, payment, refunds, expectations, and payouts, the heavier the operational burden becomes.

HomeStars leans more toward reviews, advertising, and lead generation. Contractors get visibility, homeowners get social proof, but the upstream problem remains: the homeowner still needs to understand the issue well enough to decide what to do next.

Wrenchy tests a different angle: if the repair request is clarified first, can the marketplace create a better lead? See the embedded market intelligence report below for the full competitive landscape and Canadian regulatory context.

Product thesis

Wrenchy does not start with contractor search. It starts with problem clarification.

The flow is intake-first: a homeowner uploads a photo or describes symptoms, the AI-assisted intake asks follow-up questions, and the system creates a structured repair request showing likely causes, urgency, DIY risk, estimated effort, and suggested trade. If a contractor is needed, they receive better context upfront.

The important point: AI is not treated as a magic repair expert. It is more useful as a structured intake assistant — gathering information, asking the right follow-up questions, reducing ambiguity, and turning a vague issue into a clearer repair request.

MVP built to test the thesis

Wrenchy was built as a pilot-ready MVP, not just a static concept. The MVP was not built to prove that the market is already won. It was built to test a narrower and more useful question:

The product surfaces below show what was built across the homeowner, contractor, and platform layers.

What the MVP looks like

The following screens are from the working MVP — built with React, TypeScript, Supabase, and Vercel. Each artifact is rendered from real codebase components, not mockups.

Why the DIY path matters

A key design decision in Wrenchy is that not every issue should be forced into “book a contractor.”

Many homeowners genuinely want to know whether they can fix this themselves. If the issue appears simple and low-risk, Wrenchy can guide the homeowner through likely causes, basic safety checks, tools needed, possible replacement parts, estimated time, estimated DIY cost, and when to stop and call a professional.

This matters because homeowners do not need repair services every day or every week. If the product is only a booking marketplace, many users will go back to Google the next time something breaks. For retention, Wrenchy cannot only be a marketplace — it needs to become a structured repair knowledge layer that helps users make better decisions even when they do not immediately book a contractor.

Structured contractor leads

When a job is posted, the contractor should not receive a vague message like “Sink broken. Please quote.” They should receive a structured lead with a clear problem description, photos, urgency, likely issue category, what the homeowner already tried, possible part requirement, estimated complexity, and location.

This helps the contractor decide faster — whether to accept, whether to bring a specific part or tool, whether to quote remotely, whether to ask follow-up questions first. The lead preview artifact above shows the contrast between a traditional lead and a Wrenchy structured lead.

Not by sending more leads. By sending clearer leads.

Monetization experiment

The MVP initially explored a marketplace payment model where homeowners pay through the platform and contractors receive payouts through the system. Technically, this is possible. But the business sequencing is not right for the earliest stage.

A payment-led marketplace model creates operational complexity before the platform has proven liquidity and trust. It requires contractor KYC, payout handling, refund policies, dispute management, chargebacks, reconciliation, legal wording, and customer support. That is too much too early.

So the payment marketplace model was intentionally postponed. Not because it could not be built. Because it should not be the center of the MVP yet.

Subscription-first, but not pretending it is final

A subscription-first model is a cleaner first test. But it should not be presented as proven. Wrenchy has not yet proven the best revenue model. There are several possible paths.

This matters because homeowners do not repair things every day. If the platform wants repeat usage, it needs knowledge value, not just booking value.

Repair knowledge network

The home repair market has countless products, models, parts, versions, and failure modes. Even within the same brand, a few years can change the manual, layout, compatible parts, or repair method.

This is why Google and YouTube are useful but incomplete. A video may be correct for a similar model but wrong for the user's actual appliance. A manual may be outdated. A replacement part may be unavailable. One wrong step can waste hours.

Wrenchy could evolve into a repair knowledge layer: user-submitted DIY cases, verified repair guides, model-specific troubleshooting, part compatibility, local part availability, manufacturer manuals, and contractor-verified instructions. In the future, manufacturers could also contribute updated manuals, verified repair videos, replacement guides, known issue databases, and warranty guidance.

This is hard. But if it works, Wrenchy is not just another contractor marketplace. It becomes infrastructure for repair knowledge.

What this MVP proves

The MVP proves several things:

• An end-to-end product surface can be built
• An intake-first flow is more thoughtful than a pure “find a pro” flow
• AI-assisted intake can turn vague homeowner input into a structured request
• Contractors can receive leads with better context
• Marketplace payment should be sequenced carefully
• The biggest marketplace risk is not technical — it is operational

The most important lesson: building the MVP made the business risks clearer than a market report alone could.

What it does not prove yet

This is not a customer success story. It does not yet prove customer traction, paid contractor demand, homeowner repeat usage, marketplace liquidity, AI accuracy across real-world repair cases, CAC/LTV, stable revenue model, dispute handling at scale, or contractor trust.

That is the right framing. Wrenchy is a market and product experiment, not a proven business.

Market research

The following report covers the Canadian home repair landscape: industry size, Right to Repair legislation, competitive positioning, target segments, and go-to-market strategy.

View full report →

Next validation steps

If Wrenchy continued, the validation should happen in this order: recruit 10 to 20 contractors in one small local area, start with 3 to 4 common trades, let 10 beta homeowners use the intake flow with real repair issues, measure lead response rate, quote conversion, contractor acceptance, willingness to pay after contractors receive real leads, whether homeowners return, and whether DIY content creates repeat usage.

The main question is not “can this be built?” The MVP already answers that. The real question is: can this marketplace be operated well enough for both contractors and homeowners to trust it?

Final view

Home repair marketplaces still have unsolved gaps. But the hardest part is not AI. The hardest part is operations.

To make this work, the platform needs to handle contractor supply, quality control, homeowner acquisition, complaints, failed jobs, expectation management, knowledge maintenance, manufacturer data, and long-term trust.

A home repair job is not like an Uber ride. The cost is higher. The risk is higher. Every job is more different. Trust is harder to build.

So Wrenchy should not be understood as “AI replacing contractor marketplaces.” The better framing is: AI can improve repair intake. But the marketplace still wins or loses through trust, supply quality, knowledge depth, and operational execution.

Technology can reduce ambiguity at the first step. The business still has to earn trust over time.