The 6 phases that make up the build
A real build is not one long stretch of coding. It is a sequence of phases, and each one has its own pace and its own way of stalling. Understanding the phases tells you where your timeline actually lives, and where you have control over it.
| Phase | What happens | Typical duration |
|---|---|---|
| Discovery | Trade, service list, service areas, competitors, current site audit if one exists | 2 to 4 days |
| Content and structure | Page map, copy for every page, photos gathered, proof points confirmed | 1 to 2 weeks |
| Design and build | Hand-coded pages built to the page map, mobile-first, on your Cloudflare account | 1.5 to 3 weeks |
| Integrations | Booking/CRM wiring, call tracking, forms, payments, review widgets, maps | 3 days to 1.5 weeks (parallel with build) |
| QA and speed pass | Cross-device testing, load-time check (target under 2 seconds), broken-link sweep, schema validation | 2 to 4 days |
| Launch and DNS cutover | Domain pointed, redirects mapped from old site, GSC and analytics verified live | 1 day, plus 24 to 48 hours propagation |
Stack those up and the honest range is 4 to 8 weeks. The phases that swing the timeline the most are content and integrations, and both of those live with you more than with the builder. A trade with a short, well-defined service list and photos already on hand moves through content fast. A trade with a dozen services across three counties and no organized photo library takes longer, not because the code is harder, but because there is more to decide and gather.
Integrations are the other swing factor. A phone-only lead flow with a simple contact form wires up in a few days. A site that needs to push leads into ServiceTitan, trigger call tracking numbers per campaign, and pull live reviews from three platforms takes longer to wire and test, because each integration is a separate connection that has to work correctly the first time a homeowner submits a form.
Every one of these phases happens on your own Cloudflare account, hand-coded, no page builder underneath. That is a build-quality decision, not a timeline shortcut, but it does mean there is no CMS setup or plugin configuration step eating days in the middle of the schedule.
What stretches a timeline past 8 weeks
Most contractor builds that run long do not run long because of the code. They run long because of a handful of predictable snags, and every one of them is avoidable if you know to watch for it going in.
- Slow content turnaround. A page map with 40 service and service-area pages needs copy, and copy needs facts: exact service list, licensing details, service radius, before/after photos. If those answers trickle in over three weeks instead of arriving in one batch, the build waits on you, not the other way around.
- Photo gaps. Contractors who have been in business for years often do not have an organized photo library. Waiting on a photographer, or digging through old phone backups for job-site shots, is one of the most common single delays in a build.
- Integration access delays. Wiring a site to ServiceTitan, Jobber, or Housecall Pro requires API credentials or admin access on your side. If getting that access takes a week of back-and-forth with your own software vendor, the integration phase stalls waiting on a login, not on any engineering work.
- Mid-build scope changes. Adding a new service line, a new city, or a rebrand halfway through a 40-page build resets parts of the page map and the copy already written. Scope decided up front in discovery stays on schedule. Scope decided in week 5 does not.
- Review and approval gaps. A build with 40+ pages needs someone on your end reading and approving copy in batches. If that review sits in an inbox for a week between rounds, the calendar slips a week, even though no one did anything wrong.
- DNS and domain access. Launch day needs access to your domain registrar or DNS settings. If that login is with an old web person who is not responding, cutover waits on that, not on the finished site.
None of these are reasons to distrust a build. They are the actual mechanics of what makes a project take 6 weeks instead of 10, and naming them up front is how a realistic timeline gets set and kept. The single best thing a contractor can do to protect a fast timeline is have the service list, photos, and integration logins ready before signing, not after.
Custom-coded vs a WordPress build: does the platform change the timeline?
Sometimes. A templated WordPress site using an off-the-shelf theme can look finished faster, because the page structure is already built and you are dropping in text and stock photography. That speed is real, and it is also the whole tradeoff: a templated site launches quicker and then carries a slower, heavier page underneath it for as long as it runs, along with a maintenance and plugin-security tax that a hand-coded site does not have.
A custom-coded build takes a bit longer up front because every page is built to the page map instead of poured into a pre-made template. That extra time in the build phase buys a site that loads in under 2 seconds without tuning, has no plugin surface to patch or get hacked through, and reads cleanly for both Google and AI search engines from day one. The honest comparison of platform and long-term cost lives in its own guide; the point here is narrower: expect a custom build to run a bit longer in the build phase, in exchange for not carrying platform overhead for years afterward.
A full WordPress-to-static migration, moving an existing WordPress site over to a hand-coded rebuild while preserving its rankings, adds its own steps: mapping every existing URL to a matching new one, auditing what is currently ranking so nothing gets lost, and a more careful redirect and cutover process on launch day. That is a different project shape than a from-scratch build, and it typically adds a week or two for the mapping and redirect work, done right, so the site does not lose the search equity it already earned.
Neither path is automatically faster in every case. A simple 6-page brochure site can go up fast on either platform. The gap widens as the site grows: the more service and service-area pages a contractor needs to compete in a real market (94+ cluster pages is typical for a competitive metro), the more the templated approach either breaks down or quietly slows down the finished site, while a hand-coded build stays fast page after page because there is nothing extra shipping with each one.
How trade and service area change the schedule
The trade itself does not usually change the build timeline much: the code for a roofer's site and an electrician's site is the same kind of work. What changes the schedule is how many services, how many locations, and how much documentation the trade typically needs to show.
- Single-trade, single-location contractors (a plumber or electrician working one metro) usually have the shortest content phase: a defined service list, a known service radius, and a straightforward proof set (license number, years in business, maybe photos). These builds tend to land at the 4 to 5 week end of the range.
- Multi-service contractors (a remodeler covering kitchens, baths, additions, and exteriors, or an HVAC company doing install, repair, and maintenance contracts) need a page map for every service, and each one needs its own copy and proof points. That adds real time to the content phase, typically pushing the build to 6 to 7 weeks.
- Multi-location or multi-county contractors (a roofer covering three counties, a landscaper serving a whole metro's suburbs) need service-area pages built out for each city or county, on top of the service pages. This is usually the single biggest driver of a longer build, because a real service-area footprint can mean dozens of pages, each needing local specifics, not copy-pasted boilerplate with a city name swapped in.
- Trades with heavier compliance or documentation needs (licensed trades like electrical and plumbing, or insurance-facing trades like roofing after storm damage) sometimes add a short review step to confirm licensing language and claims are stated accurately. This is usually measured in days, not weeks, but it is worth planning for if your trade carries that kind of scrutiny.
The practical takeaway: if you know going in that you are a multi-service, multi-county operation, budget for the 7 to 8 week end of the range and have your service area list and per-service proof points ready before kickoff. If you are a single-location, single-service shop, you can reasonably expect the shorter end, and showing up to discovery with your service list and photos already organized can trim it further.
What happens after launch: the first 30 days
Launch day is a milestone, not the finish line, and a realistic timeline should include what the first month after going live actually looks like. The site does not need more building at this point, but a few things need to be watched and confirmed.
The first 48 to 72 hours after DNS cutover are for verification: confirming the domain resolves everywhere, that old URLs redirect correctly if this was a rebuild, that forms are actually delivering leads to your inbox or CRM, and that Google Search Console and analytics are tracking the new site instead of a blank slate. This is quiet, unglamorous work, and it is also the work that catches a broken form before it costs you a lead.
Search visibility itself takes longer than the build does, and this is the part that trips up contractors who expect the phone to start ringing the week the site goes live. A brand-new site, or a rebuilt one, needs time for Google to crawl it, index the new pages, and start ranking them against competitors who have been sitting in those positions for months or years. For competitive terms in a real market, that runway is typically 4 to 9 months, not 4 to 9 weeks. The build timeline and the ranking timeline are two different clocks, and conflating them is the single most common source of frustration after a new site goes live.
That gap is exactly why a website build and an SEO or local-search push are usually planned together rather than sequentially: waiting until the site is finished to start earning rankings just adds the build timeline onto the front of the ranking timeline. A site built with clean structure and full service and service-area coverage from day one is already positioned to start climbing the moment it launches, instead of losing weeks to a second phase of technical cleanup before ranking work can even begin.
How to keep your build on the fast end of the range
Most of the levers that speed up a build sit on the contractor's side of the table, not the builder's. A few habits, done before and during the project, are the difference between the 4 to 5 week timeline and the 8-plus week one.
- Bring your service list and service area finalized to kickoff. If the discovery call ends with "we'll figure out exactly which services to list," that decision becomes a delay later. Walk in with the list decided.
- Gather photos before the project starts, not during it. Job-site photos, crew photos, before/afters, whatever you have. Sending them in one batch in week one beats trickling them in over a month.
- Line up integration access early. If the site needs to connect to ServiceTitan, Jobber, Housecall Pro, a payment processor, or a review platform, get the admin logins or API access sorted before the integration phase starts, not when it starts.
- Assign one decision-maker for copy approval. A build with three people weighing in on every page, on different schedules, is slower than one person reviewing a batch and signing off. Designate who has final say before review rounds begin.
- Hold scope steady once the page map is set. Adding new services or cities mid-build is fine to plan for, but doing it after copy is already written resets that work. If you know a new service line is coming, say so in discovery.
- Have your domain and DNS access ready before launch week. If your registrar login lives with a past web person who is hard to reach, sort that out in week one, not launch week.
None of this is about rushing a build. A rushed build skips the parts (proof points, service-area specificity, integration testing) that make the site actually work once it is live. This is about removing the waiting that has nothing to do with the work itself, so the calendar reflects the build, not the back-and-forth around it.