The two definitions of "custom" that produce two different timelines
Almost every timeline conversation in real estate web goes sideways for the same reason: the word "custom" is doing two different jobs. Until both parties name which one they mean, every estimate answers a different question.
The first definition is a custom-skin-on-template. One to three weeks. A templated platform — BoldTrail, Sierra Interactive, Luxury Presence, Real Geeks, Placester, Chime, kvCORE — with brand colors swapped in, the agent's logo dropped on, a font replaced, a hero image uploaded, a few headlines edited. The underlying templates, IDX widget, and page structure are what the platform ships. This is what most agents are sold, and most of the time it is what they need. There is nothing wrong with it as a category. It is simply a different product than the second.
The second definition is a bespoke custom build. Six to ten weeks for a single-agent luxury site, ten to sixteen weeks for a team or boutique brokerage. A site designed from the brand outward — typography hierarchy, motion language, listing card design, neighborhood guide template — with native IDX wired into your own templates rather than embedded in an iframe, a documented design system handed over at the end, and the content surface yours to manage. This is the product we describe in the sister piece on custom builds with IDX, and the rest of this article is about what its timeline actually looks like.
Why "fully custom luxury site in 2 weeks" is selling the first category
The math is straightforward. A real discovery phase — audience, market, positioning, brand audit, content audit, IDX scope — alone is one to two weeks. The design system phase, even compressed, is another two to three. IDX wiring against your own templates takes three to four weeks of active development on top of whatever the MLS board's approval cycle adds. There is no shortcut that compresses all of that into ten business days. When a vendor quotes two weeks and calls it custom, what they are quoting is a brand skin over a template they have already built a hundred times. That product is legitimate, but the conversation should name it accurately.
The 8 phases of a real custom luxury build
A bespoke build is not one job. It is eight, and they overlap. Treating them as discrete buckets is what makes the timeline legible.
1. Discovery (week 1–2)
Audience, market, positioning, brand audit, content audit, IDX scope. Which MLS board. Which feed tier. RESO Web API or a vendor feed. Where the content surface needs to be — neighborhood guides, market reports, a press archive, a sold gallery. The single biggest predictor of project delay is short-changing this phase. Studios that try to "save time" by jumping straight to design end up rebuilding the design when the discovery answers surface in week six.
2. Design system (week 2–4)
Typography hierarchy, color, image treatment, motion language, listing card design, listing detail design, neighborhood guide template, market report template. The output is a documented design system — in Figma or an equivalent — not a single homepage mockup. The difference matters at handover. A homepage mockup tells a developer what one page looks like. A design system tells a developer what every page looks like, including the listing page Google sees the most of. This is the artifact that makes the site extensible after launch, and it is one of the criteria in our breakdown of what separates a serious luxury site in 2026.
3. Build — frontend (week 3–6)
The design system implemented as production templates. Components, layouts, animations, responsive behavior. Performance budget held under two seconds LCP with listings rendered. Built with the assumption that the agent's team will edit copy, swap photography, and publish new neighborhood guides without involving the studio every time.
4. IDX wiring (week 4–7, often the gating phase)
RESO Web API or vendor-direct feed integrated against your own templates. Listing search, listing detail, map experience, lead capture — all rendered by your code, not by an iframe sitting inside a rectangle on your page. This is where most custom builds slip. Vendor approval cycles. MLS board approval cycles. Data dictionary mismatches between what the feed publishes and what the templates expect. Feed throughput limits. Image CDN setup so the listing gallery does not block first paint. The technical work is tractable. The paperwork around it is the variable. This is also the phase we go deep on in the IDX integration guide.
5. Content (week 5–7)
Neighborhood guides, market reports, listing pitch content, the about-the-agent or team page, testimonials, transaction histories where appropriate. Real content, not lorem ipsum. A studio can structure, edit, and shape. The source material — the agent's voice, the local knowledge, the why-this-neighborhood — has to come from the agent. This is why content runs in parallel with the build rather than after it.
6. CRM and lead routing (week 6–8)
Lead capture forms landing cleanly in whichever CRM the agent already runs. Mapped to source. Mapped to the specific listing where the inquiry came from. Notification rules so a luxury inquiry is not waiting in a queue with thirty rental leads. Re-engagement triggers for saved searches and alert signups. The integration is configured to the agent's existing stack, not the studio's preferred vendor.
7. SEO setup (week 6–8)
Technical SEO — sitemap, schema, canonical tags, structured data on listing pages. On-page — titles, descriptions, neighborhood guide structure, internal link architecture. Redirect map from the old site so existing equity is preserved, not orphaned. Search Console verification, analytics, conversion goals. The SEO work is not a launch-week task. It is built in from week one and finalized as the templates settle.
8. Launch and post-launch (week 8–10)
Pre-launch QA, launch, post-launch iteration on real traffic. The first two weeks after launch surface the issues no internal QA pass catches — the listing that has thirty-eight images instead of the typical twelve, the rare neighborhood that breaks the map clustering, the email client that renders the inquiry receipt at half width. Plan for the iteration window. Do not pretend launch day is the end of the project.
Where delays actually come from
Not the build. The build is the most predictable phase. Delays on a custom luxury project almost always come from three places.
The first is content. The agent has not written the neighborhood guides, the studio cannot write them in the agent's voice without source material, and the launch slips by the number of weeks it takes the agent to finish writing. The fix is to start content work the day discovery ends, not the week before launch.
The second is IDX approval cycles. MLS boards have their own paperwork timelines. Some boards approve within days. Some take six weeks. The vendor licensing for a fuller feed tier adds another cycle on top. None of this is in the studio's direct control. The fix is to start the board paperwork on day one of discovery, not when the IDX wiring phase begins.
The third is decisions deferred. Logo direction, color palette, photography selection, copy direction — when these are not settled by the end of discovery, every downstream phase blocks waiting on them. A two-day decision held for three weeks compounds into a six-week delay. The fix is to treat brand decisions as a gating deliverable of discovery, not a parallel workstream that resolves whenever.
Realistic timelines for the three common scopes
Across the engagements we have run, three scope shapes account for most luxury builds. The realistic ranges look like this.
Single-agent luxury site, content mostly ready, single MLS feed. Six to eight weeks. The agent has neighborhood content drafted, photography selected, brand direction settled. Discovery moves fast because the answers are already in the room. IDX is one board. This is the cleanest version of the engagement.
Single-agent luxury site, content needs writing, single MLS feed. Eight to ten weeks. Same technical scope, but the content has to be developed from interviews and source material. Content runs in parallel from week two, and launch is gated by content finalization rather than by code. This is the most common shape.
Team or boutique brokerage, multi-agent profiles, multi-MLS feeds. Ten to sixteen weeks. Multiple agent profile templates, varying permission structures, brokerage-level brand alongside individual agent voice, multiple board approvals running in parallel, content multiplied by team size. The build is not linearly longer — it is structurally different — and the timeline reflects the additional coordination.
How to compress timeline without cutting quality
The shortcuts that actually work are unglamorous. Do the content work in parallel with the design system phase, not after it. Make brand decisions before discovery ends so the design system has somewhere to anchor. Get the MLS board paperwork started the day discovery begins so the approval clock runs while the studio designs. Identify the photography you will use before week four — not the week before launch. Hand over CRM access and the existing site's analytics on day one so the SEO work and the redirect map can be roughed out early.
None of this changes the build clock. The build clock is the build clock. What it changes is the gate clock — the time the project spends waiting for inputs that are not the studio's to produce. On most custom builds the gate clock is longer than the build clock, and that is where weeks are actually recovered.
Why "fast" is usually the wrong frame
The website is going to represent the agent for the next three to five years. It is the first impression for every referral, every cold prospect, every listing pitch, every press inquiry. Six extra weeks at the front to get the design system and IDX right pays back across the life of the site many times over. A two-week templated launch followed by a six-month rebuild eighteen months in is not the fast version. It is the slow version, paid for twice. The argument we make in the piece on why luxury agents need custom websites in the first place is the same argument that applies to timeline: optimizing for speed of launch alone almost always lengthens the total clock, because it shortens the life of the asset.
How AgentCentric runs the timeline
AgentCentric is a luxury website studio. The typical engagement is six to ten weeks for a single-agent build. Discovery runs in week one. The design system is week two through four. IDX wiring is the gating phase from week four onward, with board paperwork started on day one to keep the approval cycle in parallel. Content runs from week two through seven. CRM, SEO, and analytics work overlap from week six. Launch lands between week eight and ten depending on scope, content readiness, and the speed of the MLS board's approval cycle.
What we hand over is a production site, a documented design system, a CRM wired to the actual leads coming in, a redirect map preserving search equity, and a team that can run the content surface without coming back to us for every update. To see the output, walk the portfolio, review what is included, or work through the FAQ.
Six to ten weeks for a site that should run for the next five years is the right shape of the trade. Anyone quoting two weeks is quoting a different product. The right answer for a luxury site sits in the middle, and it is mostly a function of how ready the agent is on day one.