Contractors Hub

What this actually is

Not a contractor directory. A renovation decision service, with a vetted bench behind it.

The instinct is to build a listings page: trades down the left, contractors down the right, a form at the bottom. That version is easy to build, hard to defend, and worth very little — because HomeStars, Houzz, TrustedPros and Google Local already occupy it, with review volume you will never match.

The version worth building answers the questions that come before "who should I call":

• Should I do this work at all? What does it return in Oakville, Burlington, Milton — and what doesn't.
• What does it actually cost? Real ranges, from someone who has priced the work, not a national average.
• What's the correct scope? Most bad renovation outcomes are scoping failures, not contractor failures.
• What's hiding in my house? Asbestos, vermiculite, knob-and-tube, Kitec, poly-B, radon, moisture. Age of home predicts most of it.
• How do I not get taken? Reading a quote, deposit norms, holdbacks, WSIB and insurance certificates, permits in whose name.

Then, and only then: "Here are two people I would use, and I'll make the introduction."

That sequence is the product. The contractor bench is the payoff, not the pitch.

Why you can build it and others can't

Every agent site in Halton renders the same things: listings, calculators, market stats, neighbourhood pages. There is close to zero differentiation between them. Twenty-five years of residential and commercial construction is the one input none of them have and none of them can license.

The site already gestures at this in two places — /construction-advantage states the credential, and /reno-estimate offers the on-site walkthrough. Both are good pages. Neither is a system. The Hub is what turns a claim on an About page into a machine that produces leads while you sleep.

Three jobs it has to do

The competitive read

Be clear-eyed about who already owns this space, because it dictates what you should and shouldn't build.

You cannot out-directory the directories, and you should not try. What none of them offer is a named professional who has walked thousands of sites putting their own reputation behind a short list, and publishing the cost reality that the pay-to-play platforms have a commercial reason not to publish.

Fifteen contractors you would genuinely hire beats four hundred you wouldn't. Scarcity is the feature.

Publish the knowledge. Gate the names.

This is the recommendation that most changes what gets built, so I want to argue it properly. You described listing contractors by trade. I would publish everything except the contractor names.

What goes public — no form, no email, fully indexed

• A page per trade: what the work involves, what it costs in Halton, what licence or certification is legally required, what a good quote contains, what to inspect on completion, red flags.
• All education and hazard content.
• All calculators and tools.
• Suppliers, showrooms and supply stores by name — no liability in naming a tile showroom, and it's genuinely useful.

What stays behind an introduction request

• The contractor names themselves — delivered by you, per project, with context, after you understand the scope.

Four reasons

1. Liability. A published list reads as an endorsement. When a listed contractor damages a floor, the homeowner's first call is to you. A per-project introduction with written scope-limiting language is a far smaller target.
2. You keep the relationship. A public list is scraped, screenshotted and passed around. Nobody ever calls you again. An introduction puts you in the middle of the transaction — which is exactly where a future listing comes from.
3. It's the lead engine. The gate is the conversion event. A public list converts nothing.
4. It's honest. You cannot warrant that a contractor who was excellent in March is available and excellent in November. "Tell me the project and I'll tell you who I'd use right now" is a claim you can actually stand behind.

Compliance and liability

Settle this before a line of code is written. It is the part that can cost you a registration, not just a lead.

I'm flagging the issues; confirm the specifics with your broker of record and, for the money question, a lawyer. Nothing here is legal advice.

1. Referral fees are the landmine

Under TRESA and the RECO code of ethics, remuneration or a benefit received in connection with a trade generally has to be disclosed in writing to your client. Taking an undisclosed cut from a contractor working on a home where you also represent the seller is the kind of thing that ends careers. Even fully disclosed, it undermines the entire proposition — the value of your recommendation is that it isn't bought.

2. Brokerage approval

Royal LePage Pinnacle needs to sign off on the Hub as a marketing initiative — branding, disclaimers, lead handling, and whether operating a contractor referral service constitutes outside business activity requiring separate declaration. Get it in writing before launch, not after a complaint.

3. Scope-of-practice language

The wording already on /construction-advantage is the right model and should be repeated across the Hub: observations are practical and preliminary; they are not a home inspection, engineering opinion, environmental assessment, code review, or contractor quotation.

4. "Vetted, not warranted"

Every introduction and every trade page carries a standing line: contractors are independent businesses; the homeowner contracts directly with them; verification was performed on a stated date and is not a guarantee of performance, and does not make you a party to the contract.

5. Insurance

Confirm with your E&O carrier whether contractor referral activity is covered. If not, either restructure it or accept the exposure knowingly. Don't discover this after an incident.

The vetting standard

This is the actual product. Publish the standard itself as a page — it demonstrates rigour, educates homeowners on what to demand from any contractor, and makes the bench credible without naming anyone.

Site architecture

Build it as a section of yovan-gabric.ca, not a separate domain. A subfolder inherits the domain's existing authority, keeps one analytics picture, and puts the Hub's traffic to work for the listing pages. A separate domain starts from zero and splits your effort.

How it links to the rest of the site

• Every trade page links to the relevant cost section, the relevant hazard page, and the introduction form.
• Every hazard page links to the trade certified to deal with it and to the disclosure implications at sale — which routes to /sell.
• The Insights page gets a Hub section, so renovation content sits beside the Royal LePage market articles rather than in a silo.
• Community pages gain a housing-stock note — Glen Abbey's build era predicts poly-B and aluminum wiring questions; Bronte's predicts something else. This is where the construction knowledge and the neighbourhood pages compound.
• Header nav gains one item. Resist adding five.

Trade taxonomy

Ten categories, roughly ninety trades. Not all get a page — see the build order below the grid. Trades marked with a credential have a legal or certification requirement that becomes the spine of that page's content.

Suppliers and showrooms

These get named publicly. There is no meaningful liability in pointing someone at a tile showroom, and a curated Halton supplier list is a genuinely useful page that will earn links. It is also the only part of the Hub where a paid sponsorship model could sit cleanly later, clearly labelled and kept away from the contractor bench.

Each supplier entry carries: name, category, what they're genuinely good for, whether they sell to the public or trade-only, whether they deliver, and one honest note from you. That last field is the whole reason anyone reads the page.

Education and blog

Seven pillars. Every article resolves to a trade page, a tool, or the introduction form — never to a dead end.

Publishing model

• Cornerstone pages are permanent and updated, not dated blog posts. A cost index with a "last reviewed" date beats a post from 2026 that nobody trusts by 2028.
• One new piece a fortnight is enough if each is genuinely better than what ranks now. Twenty thin posts are worse than five deep ones.
• Every hazard and cost page carries a review date in the same spirit as the tax rules in the calculators — one source of truth, visibly maintained.
• Repurpose relentlessly: each cornerstone piece yields a short video walkthrough, a carousel, and an email. The construction footage is content nobody else in Halton has.

Tools, forms and the value ladder

The site's existing calculators are the right model: everything runs in the browser, nothing is transmitted, no sign-up wall. That's precisely what makes them linkable. Keep that principle and add a deliberate ladder — each rung asks for more, and gives more.

Form specifications

Homeowner introduction request. Name, email, phone, property address, city, project category (multi-select from the taxonomy), scope in their words, budget band, timing, own/buying/selling-within-2-years, how they found you, consent checkbox for follow-up. Honeypot field. Posts to the existing /api/lead endpoint with formName: "contractor-intro" — no new infrastructure needed.

Contractor application. Business name, contact, trades (multi-select), service areas, years in business, HST number, WSIB number, insurance carrier and policy expiry, licence numbers by type, three references with contact details, three recent projects, whether they provide written fixed-scope quotes, whether they pull their own permits, warranty terms, insurance certificate upload. This form is doing your intake work — make it long on purpose. Contractors who won't complete it have self-selected out.

Data model and technical plan

The Hub adds no new services, no new subscriptions and no new monthly cost. It is content and forms on infrastructure you already run: Next.js on Cloudflare, with Drizzle already installed against a schema that is currently empty by design.

Phase 1 — no database at all

Trades, hazards, suppliers and bench records live in typed data modules in the repo, exactly like precon-data.js on the Feras Riyal site. Editing one file and redeploying is faster than any admin panel, and it version-controls every change to a cost figure — which matters when a number is challenged.

Same house rule as the project cards: a missing field renders as absent, never as a guess. A trade page with no confirmed cost band says the cost band hasn't been reviewed. That's a credibility signal, not a gap.

Phase 3 — when a database earns its place

Only ever for your own leads, and only once the spreadsheet genuinely annoys you — realistically when you want to track which introduction converted to a listing across enough records that scrolling stops working. Cloudflare D1 is already wired through the existing Drizzle setup and getDb(), so this is a config change plus a schema when the day comes.

Note that .openai/hosting.json currently has d1: null, so no database is provisioned today and none is needed. A database added early is a maintenance burden with no payoff.

SEO and structured data

• Schema.org: HowTo on process content, FAQPage on trade and hazard pages, LocalBusiness on supplier entries, Article with a visible dateModified on cost pages.
• Internal linking is the whole game. Trade ↔ hazard ↔ cost ↔ community ↔ sell. Every page should be two clicks from the introduction form.
• Target the long tail: "asbestos popcorn ceiling Oakville", "cost to finish a basement Burlington", "do I need a permit to remove a wall Milton". Nobody local is competing for these properly.
• Do not generate hundreds of near-identical trade-by-city pages. That is the pattern that gets sites filtered out of search, and it's exactly what the pre-construction portals do to each other.

Phased rollout

Where each decision stands

Phase 1 is built. Two decisions were made by you; four were built under a stated assumption and are cheap to reverse; one is genuinely still open and is the one that matters most.

• 1 · Gated bench settled No contractor names published. Names given per project, after the scope is understood. The roster stays private and is never deployed. §04
• 7 · Informational site, personal introduction settled No contractor logins, now or later. The site publishes knowledge; the introduction is Yovan answering a client question. This is the framing for the brokerage conversation. §05
• 2 · Money model assumed Built on "take nothing" — stated in plain language on four pages. Changing it is a copy edit, but it weakens the core claim and needs brokerage and legal sign-off. §05
• 4 · Where leads land assumed Built on the existing /api/lead webhook. No new infrastructure. Your leads only — contractors are never stored. §11 · §12
• 5 · Section name and route assumed Built at /renovate. "Renovate" speaks to a homeowner; "Contractors Hub" sounds like it is for contractors. A rename is a directory move plus the nav label.
• 6 · Phase 1 scope assumed Built as twelve trade guides, five hazard guides, the cost page, the vetting standard, suppliers, hiring, and two forms. Forty-eight further specialties are listed but not written.
• 3 · Real cost bands under your name still open Every costBands array is empty and the cost page says so outright. The table renders automatically the moment you fill one array and set a review date. Until then the flagship page is a methodology page — good, but not the differentiator. §10

Metrics and risks

What to measure, from day one

Risks, ranked by what they'd actually cost you