Why does a website cost $15,000 when Wix/Squarespace is $20/month

A $20/month Wix or Squarespace plan is a hosted editor seat. The $15,000 in the question is a planning example for the work that seat does not do: strategy, information architecture, authored design, a CMS your people can run, answer-engine markup, and ownership you can leave with. Comparing them as the same SKU is how buyers feel cheated and how studios look expensive for selling labor. This spoke sits under Websites That Feel Like Films. The cousin post on why custom sites cost five figures maps cost drivers and three-year ownership. This one answers the sticker objection.

The short answer

• $15,000 is a planning example, not a quote. It is buyer shorthand for “serious custom.” It is not a Spurlock Studios rate card, not a frozen market average, and not a promise of your invoice.
• $20/month is a platform seat. Hosting, SSL, the editor, templates, and the vendor’s roadmap. Verify today’s dollars on the vendor’s checkout — they move.
• The category error is unit of sale. One number is rent. The other is a project. You cannot divide project labor by twelve and call the remainder greed.
• Five jobs sit outside the subscription: design, IA, CMS modeling, AEO, ownership. Skip any of them and you still paid for a URL.
• Decide on the job, not the sticker. If you need a login this week, buy the plan. If you need a brand system that converts and cites, budget labor.

Thing people compare	What it actually is	Honest next question
“$20/month”	Hosted builder subscription	What work am I still doing myself?
“$15,000 website”	Planning example for a project	Which workstreams are in the box?
“They’re the same site”	Same category of URL, different product	What do I own on day one?
“I’ll pay forever either way”	Rent vs ownership + optional keep-alive	Whose card, whose accounts, whose exit?

The rest of this post unbundles those two numbers so you can argue with a designer — or with yourself — without mixing rent and labor.

Why $15,000 vs $20/month is a category error

You are not looking at two prices for one object. You are looking at a subscription and a project. A Squarespace or Wix plan is billed per month (or year) because the vendor is selling runtime: servers, the editor product, templates, apps, support at the plan’s level. A five-figure website number is billed once (or in milestones) because a studio is selling hours: research, composition, build, QA, handoff.

If you force them onto one line, the expensive one always looks like theft. If you keep the units honest, the cheap one often looks unfinished.

Unit	$20/month class	$15,000 planning example
What you buy	Platform seat	Scoped labor
When you pay	Every billing cycle	Up front / milestones
Who does strategy	You, or nobody	Included when the quote is real
Who designs	You + a template family	A designer, on purpose
Who owns the exit	The vendor’s export path	A handoff checklist in writing
What breaks first	Template ceiling, then a migration	Scope fights, if discovery was skipped

Procedure for catching the category error in a sales call:

1. Ask “is this number a plan, a setup fee, or a project?”
2. If it is a plan, write down what you still have to do (copy, photos, IA, SEO, AEO).
3. If it is a project, write down the workstreams — not the page cap.
4. Circle anything that exists in both columns. That overlap is hosting and an editor, not a brand system.
5. Re-rank the options on the leftover jobs, not on the cover number.

$20/month can be the right buy. It is the wrong comparison. The minute you treat a hosted login as a substitute for authored work, every custom quote looks padded. The minute you treat a custom quote as “hosting with a markup,” you will hire the cheapest theme flip and pay for a second site later.

What a $20/month Wix or Squarespace plan actually buys

The buyer’s “about $20 a month” line is useful as an order of magnitude for a basic hosted plan. It is not a forever quote. Wix’s own explainer has put paid plans at $17 to $159 per month on yearly billing (Wix: How much is a Wix website?). Squarespace’s public tiers live on squarespace.com/pricing; Wix’s current SKUs live on wix.com/plans. Taxes, domains, apps, commerce fees, and monthly-vs-annual billing sit on top. I will not freeze a checkout price here as if it were a law.

Both vendors are clear about the product: hosting is included with the plan. Squarespace’s pricing FAQ states hosting is included. Wix lists “reliable web hosting” on every paid plan. That is the point of the SKU. You are renting a running site, not buying a designer.

Included in a typical paid builder plan	Not included, no matter the tier
Hosted runtime and SSL	Strategy for your offers
The editor / designer product	Original compositions
Template library and the vendor’s roadmap	Information architecture with a job per page
Support at the plan’s level	A CMS modeled to your editors’ real update jobs
Often a first-year domain voucher	Answer-shaped HTML and citeable schema
Apps you can add (usually extra)	Domain/hosting/CMS in your accounts as a handoff package

What the monthly fee covers, in vendor language:

1. A place for the site to live, with SSL.
2. The drag-and-drop (or equivalent) editor.
3. Templates, and thousands of other sites on the same family.
4. Platform SEO toggles, not an IA or a redirect map.
5. Support tickets about their product, not about your conversion path.

What it still costs you after you pay:

1. Your time to pick a template, write copy, and fight the layout model.
2. Stock or a shoot — the plan does not photograph your work.
3. Apps, fonts, email, analytics, a domain after the voucher year.
4. A designer retainer if you cannot make it look like you.
5. A migration project the day you outgrow it — see Leave Squarespace Without Torching the Rankings.

A plan that includes hosting is a good product. Calling it “a website” in the same breath as a custom build is how the $15,000 objection gets born.

What the $15,000 planning example is standing in for

Treat $15,000 as a shape, not an invoice. People say it because it is a round five-figure number they have heard in a pitch, a forum, or a cousin’s project. Studios quote from scope. Scope is pages-as-jobs, unique templates, motion, CMS collections, migrations, photography direction, stakeholder count, and handoff. Two “$15k websites” can contain different products. One number with no workstreams is not a bargain. It is an incomplete sentence.

I will not invent a market average for “what a website costs.” Trade blogs publish ranges so wide they are trivia until you name the job. This post uses $15,000 the way the question uses it: a planning example for serious custom work. Your quote may land under or over that example. The unbundling still holds.

Planning line (example, not a quote)	What “done” looks like	What a $20/month plan substitutes
Discovery / offer map	One primary action per key page	You guess, or you copy a competitor
Information architecture	Sitemap with jobs, not a junk drawer	Template nav, then you add pages until it breaks
Authored design	Compositions unique to this brand	A theme family your peers can buy this afternoon
Production build	The design actually ships on phones	Whatever the editor will allow
CMS modeling	Fields a manager can fill without breaking type	A generic blog or “portfolio” widget
AEO / SEO basics	Titles, FAQ answers in HTML, schema, Search Console in your account	Toggles and a blog module
QA / launch	Forms, redirects, analytics, a performance pass	“Looks fine on my laptop”
Ownership handoff	Domain, billing, CMS, DNS, analytics under you	Their platform account, their export path

Checklist — the $15,000 example is being used honestly when:

• Someone says it is a planning example, not “our price”
• Workstreams are named in English, not “website + SEO”
• Unique templates are counted, not “unlimited pages”
• Editors are named (who publishes next Tuesday)
• Ownership is a list of accounts, not “you own everything”
• Change-order rules exist, so the example cannot silently become $40,000 of taste rounds

If a studio hears “$15,000” and immediately says yes without asking about jobs, editors, or exit, they are selling a number. Walk. If they take the example apart into the table above, you are finally comparing products.

Hours, not hosting, are why the planning example lives in five figures. A five-to-ten-page brand site with original compositions, a usable CMS, and a clean launch is tens to well over a hundred hours once you count revisions and QA. At professional rates, that shape is five figures before anyone has argued about “unlimited pages.” I will not pretend those hours are a Spurlock estimate. They exist so you can see why a login and a project cannot share a sticker.

Hosting is not design, IA, CMS, AEO, or ownership

This is the whole objection, unbundled. Hosting is a utility. The other five are jobs. A Wix or Squarespace plan bundles hosting with an editor. It does not bundle the five jobs. Custom quotes look large because they are mostly the five jobs. If you only needed hosting, you would buy hosting.

Job	What it produces	What $20/month hosting does instead
Design	Brand-specific type, space, motion rules, proof in the hero	A template family. Different photos, same bones.
IA	A sitemap where each URL has a job	Pages get added until the menu is a junk drawer.
CMS	Collections and field limits a named editor can run	A generic editor. Easy on day one, chaotic on day sixty.
AEO	Citeable answers in HTML, schema.org types, entity consistency	Meta fields and a blog. Models still have nothing clean to quote.
Ownership	Your registrar, your billing, your CMS admin, your Search Console	Their workspace. Export later. Maybe.

Design is not a skin. Authored composition is how a site stops looking like the other seven in the niche. The craft standard lives in the pillar: film-grade pages, not brochure themes. Hosting cannot do that. A template can look expensive in the gallery and still be interchangeable on a phone.

IA is not “we’ll add pages.” If Home, About, and Contact are doing three jobs each, visitors bounce. Builders make adding a page cheap. They make deciding the job expensive — because nobody was paid to decide.

CMS is not “there’s an editor.” The question is whether your manager can change a date, a merch image, or a press quote without breaking crop, type, or the CTA. A modeled CMS is labor. A generic one is a login.

AEO is not a plugin. Answer engines cite passages they can lift. That means question-shaped headings, direct-answer leads, FAQ HTML, and types from schema.org that match the page (Organization, FAQPage, MusicGroup, LocalBusiness — pick the real one). Google’s structured data intro is still the floor: markup has to match visible content. A $20/month plan can host that HTML. It will not author it.

Ownership is not a vibe. If the studio’s card is on the Wix or Squarespace account, you are renting a hostage. If Search Console is on a freelancer Gmail, you will rebuild proof when they vanish. Hosting included in a plan is fine. Hosting that you cannot take with you is the expensive part.

• Write the five jobs on a whiteboard
• Put a check only where a human is paid to do the job
• Hosting checks itself. That is not a win.
• Empty rows are the real gap between $20/month and the $15,000 example
• Fill empty rows with DIY, a contractor, or a project. Do not pretend the plan filled them.

Utility ≠ system. Pay for the utility monthly. Pay for the system once, on purpose, or admit you are the system.

Why DIY on Wix is not a $15,000 discount

The other version of the objection is: “I could just do it myself on Wix.” You could. That does not make the planning example a markup on your evenings. DIY is a staffing choice. You are assigning design, IA, CMS, AEO, and ownership to yourself. The $20/month plan still only buys the seat. If you do the five jobs well, you earned the site. If you skip them, you did not beat custom. You deferred it.

I will not invent an hourly rate and multiply it into a fake “true DIY cost.” Your time is not a studio invoice. It is also not free. Treat it as capacity: nights you will actually spend, and nights you will not.

Job	What DIY actually is	What people pretend it is
Design	Learning the editor and making original compositions	Dragging a template and changing colors
IA	Deciding page jobs before you add pages	Adding every page the template demo had
CMS	Building the weekly update path you will keep	“I’ll just edit the page when I need to”
AEO	Writing answers, marking them up, verifying Search Console	Checking a platform SEO toggle
Ownership	Registrar, billing, 2FA, exports, backups	Assuming the plan is the company
QA	Phone, form, slow network, a friend who is not you	Looking at it on the laptop you built it on

Procedure if you are going to DIY and still be honest about the objection:

1. Write the five jobs. Put your name next to each one you will actually do.
2. Calendar the hours. If you cannot point at evenings this month, you are not staffing the job.
3. Ship one money page to a phone before you add a seventh page.
4. Create Search Console and analytics on day one, under your account.
5. Set a review date 30 days out: editor test, CTA test, peer-sameness test.
6. If three jobs are still empty at that date, you do not have a cheap website. You have a stall. Either keep DIY with a tighter scope, hire for the empty rows, or stop calling the $15,000 example a ripoff.

• You own billing and the domain before you design anything
• You can name the one action the homepage is for
• You have real photos, or you have admitted the hero is a placeholder
• You will not hide missing IA behind more apps
• You know the export / leave path while the site is still small

DIY can be the right call for validation. It is the wrong argument against labor. A studio is not charging you $15,000 because they are too proud to open Wix. They are charging because someone has to do the jobs the plan does not do. If that someone is you, great. Put it on a calendar. If that someone is nobody, the $20/month site will look like it.

How this mismatch shows up in conversion

The objection is usually about money. The damage is usually about behavior on the page. A hosted template and an authored site can both collect a form. They do not spend the first seconds the same way. I will not invent a conversion-lift percentage. Watch what people actually do.

Symptom on a $20/month-class build	What a visitor does	What authored work changes
Hero looks like the niche’s default theme	They assume you are interchangeable	Proof, type, and composition that could only be you
Nav is a junk drawer	They never reach the offer	IA with one job per key URL
CTA is “Learn more” on every page	They wander, then leave	Named next step: listen, book, quote, buy
Forms go to a personal inbox	Leads die in spam	Routed, spam-filtered, logged
Pages have no citeable answer	AI Overviews skip you; humans skim past	Direct-answer leads + FAQ HTML
LCP is a 12MB gallery	Phone users bounce before the pitch	A named LCP element and an image pipeline

Google’s Core Web Vitals still set a field floor most marketing sites should treat as real: LCP within 2.5s, INP at or under 200ms, CLS at or under 0.1, at the 75th percentile. A builder plan can pass those tests. A heavy template with autoplay and uncompressed heroes often does not. Passing Vitals is not “custom.” Failing them on a phone is a conversion tax the $20/month sticker never showed.

Checklist — conversion evidence you can collect without a fake ROI calculator:

• Record a 15-second phone pass of the homepage with sound off. Can a stranger name the offer?
• Count unique templates, not pages. If everything is one layout, you have a brochure, not a path.
• Click every primary CTA on a mid-range phone. Note what actually happens.
• Search your brand + category in Google and in one answer engine. Are you cited, or only listed?
• Ask one manager to update a real date or image. Time it. If they Slack the designer, you do not have a CMS.
• Compare three peer sites. If a stranger cannot tell you apart, the template ceiling is the cost.

Conversion is why people pay project money. Not because hosting is expensive. Because an interchangeable first impression is expensive. If your offer is still a moving target, that tax may be acceptable. If you are already losing deals to sites that look like films, the $20/month plan is not the savings. It is the ceiling.

What you should ask a designer about the $15,000 number

Ask them to unbundle the planning example. Do not ask “can you do it for less than Squarespace.” That question is the category error again. Ask which jobs are in the number, which are yours, and what you own when they are done.

Procedure on the call (write the answers down — a reel is not diligence):

1. “Is $15,000 a quote, a starting range, or a planning example?” If they treat it as a quote without a sitemap, stop.
2. “Which workstreams are in, and which are change orders?” Demand English: discovery, IA, design, build, CMS, AEO/SEO, QA, handoff.
3. “How many unique templates, not how many pages?”
4. “Who edits next Tuesday, in what tool, on what device?”
5. “Where does Search Console live, and who verifies it?”
6. “Whose card is on hosting, and can billing move tomorrow?”
7. “If I leave, what do I take — repo, CMS export, redirect map, accounts?”
8. “What does ‘SEO’ or ‘AEO’ mean in HTML, not in a package name?”

Answer you want	Answer that should end the call
“$15k is a shape until we count templates and editors”	“Yes, $15,000, unlimited pages”
Named workstreams + change-order rules	“Website + SEO included”
You are the account owner; they get invited	“We host it for you so it’s easier”
CMS demo: they watch you change a real field	“We’ll train you later”
Schema types named, FAQ in HTML	“We optimize for Google”
Handoff checklist dated	“You own everything” with no account list

• You heard “planning example” or you heard a scoped quote — not a magic number
• Hosting is a line, not the story
• Design, IA, CMS, AEO, ownership each have an owner (studio, you, or out of scope)
• You can name the editor
• You can name the exit
• You did not accept “it’s like Squarespace but nicer” as a scope

If they cannot unbundle $15,000, they cannot defend it. If they unbundle it and you still only need a login, buy Wix or Squarespace and stop interviewing studios. The expensive hire is the one who agrees the numbers are the same product.

How to unbundle a mixed bid so hosting stops winning

A lot of quotes are hybrids: “Squarespace setup for $X, then $20/month.” That is a real product. It is still not the $15,000 planning example unless the five jobs are in the setup fee. Mixed bids are where buyers get confused, because one line looks like custom labor and the other looks like the plan they already feared.

Force three columns. Hosting lives in one column. Labor lives in another. Ownership lives in the third. If a bid cannot fill the grid, it is a vibe.

Column	Builder setup bid	Authored project (planning example)	What to reject
Hosting / plan	Named vendor plan, your billing	Host or CMS plan, your billing	“We include hosting” with no account
Design labor	Theme customization, how many templates	Original compositions, template count	“It will look custom” with a demo URL
IA	Sitemap in writing, or none	Sitemap with jobs	“Unlimited pages”
CMS	What you can edit without them	Collections + field limits + training	“The editor is easy”
AEO / SEO	Titles, FAQ HTML, Search Console owner	Same, plus schema types named	“SEO included” as a package name
Exit	Export path, redirect plan if you leave later	Handoff checklist at launch	“We’ll transfer if you ask”

Procedure with two PDFs and a Wix tab open:

1. Draw three columns: Hosting, Labor, Ownership.
2. Drop every line from each bid into a column. Orphans go in a fourth column labeled Unclear.
3. Send Unclear back as one email. Do not negotiate the cover number until the grid is full.
4. Price the empty labor rows yourself: DIY, a contractor, or out of scope. Empty is allowed. Hidden is not.
5. Compare filled grids. The cheaper cover number often moved labor into your nights or into “phase two.”
6. If both grids are full and you still only need a login, pick the builder setup. That is a coherent buy.

• Plan dollars are verified on the vendor site, not on a blog (including this one)
• Setup fee lists templates, not a page cap
• Search Console property is named as yours
• No studio-owned Wix / Squarespace / Webflow / Framer workspace as the long-term home
• Redirects appear as a line if any old URLs exist
• You can explain the bid to a partner in one minute without saying “it’s just more premium”

This is the anti-sticker move. Once hosting is a boring utility row, the $15,000 example has to defend labor or it should lose. Once labor is visible, $20/month has to admit what it does not do. That is an argument you can win without inventing a market average.

When is a custom site worth it instead of $20/month?

Custom is worth it when the five unpaid jobs are now the bottleneck, not when you are bored of the template gallery. A builder is worth it when you need a credible URL this month and the offer is still moving. I have shipped hundreds of production sites. The pattern that wastes money is buying project labor for a validation site — or staying on a template after the brand has already outgrown it.

Signal	Stay on Wix / Squarespace	Budget a custom project
Offer	Changes weekly	Stable enough to IA
Proof	You do not have real photos yet	You have work, press, or a catalog worth composing
Editors	You, on weekends	A manager who will publish without you
Peers	Also on templates	Already look like films
Conversion path	One page, one email	Multiple jobs: listen / tour / merch / quote / book
Ownership	You already own the plan	A freelancer’s login is the site
Migration	You just launched	Indexed URLs you still need
AEO	You have nothing citeable to say yet	Buyers ask recommendation questions and you are invisible

Decision list — custom is the right spend when three or more are true:

1. Prospects compare you to peers and you lose on trust before the call.
2. A manager cannot ship a real update without a designer.
3. Merch, dates, case studies, or service areas need structure the theme will not hold.
4. You are about to pay someone to fight the template for the second time.
5. Ownership sits on someone else’s card or Gmail.
6. You need pages that answer questions out loud — not a gallery with alt text.

Custom is the wrong spend when:

1. You refuse real photography and want “Apple vibes” from stock.
2. The offer will be a different company in six weeks.
3. You want five figures of motion you will disable on mobile.
4. You will not sit through discovery. Then you are buying a prettier guess.

Worth-it is a threshold, not a vibe. The $20/month plan is correct below the threshold. The $15,000 planning example is a way to talk about labor above it. Do not use either number as a personality test.

What usually fails first when teams treat these as the same product

Failure mode: a brand hears “website,” buys the cheap SKU, then tries to extract custom outcomes from a template. Or it hears “$15,000,” buys a renamed demo, and still has a login. Both fail the same way: the five jobs were never staffed. The invoice was not the product.

Failure	What it costs	What you do instead
Paying project money for theme setup	Five figures, still a template family	Unbundle. If design is “pick a demo,” you bought a costume.
Paying a designer monthly to fight the builder	Retainers that never become a system	Either accept the template ceiling or fund IA + authored templates once
“SEO included” with no URL map	Rankings vanish on the next rebuild	Inventory URLs before you touch DNS
CMS that is just the page editor	Every copy change is a studio ticket	Model collections around the weekly job
Schema that does not match the page	Manual actions, or silent ignore	Markup equals visible HTML, per Google’s structured-data rule
Studio-owned billing	Hostage on a missed invoice	Your card, your registrar, their invite
Launch without a phone QA pass	Forms fail, LCP fails, CTAs fail	Mid-range phone, real network, real form destination

The migration version of this failure is the one I see after the template ceiling hits. People leave Squarespace in a weekend, keep the domain, and skip the 1:1 redirect map. Google’s site-move guidance still wants that map and server-side 301 or 308 redirects. The $20/month years were not wasted. The wasted money is the salvage project because nobody treated exit as a workstream. If you are already leaving, do not improvise — use the migration spoke, not a DNS cutover party.

Procedure after a failure (do this before you write a second check):

1. List the five jobs. Mark each as done, DIY, or missing.
2. List accounts: registrar, hosting/plan, CMS, Search Console, analytics, repo/export.
3. Anything not in your email is a hostage. Fix that this week, even if you stay on the builder.
4. If the missing jobs are strategy and IA, do not buy motion. Buy discovery.
5. If the missing job is ownership only, you may not need a rebuild. You need a transfer.
6. If you are rebuilding, budget redirects as a line, not a favor.

Cheap and expensive fail for the same reason when they are the same incomplete product. Staff the jobs. Then pick the stack.

How do I measure whether the spend is working?

Do not measure “do I feel like $15,000 was fair.” Measure whether the jobs you paid for are doing work in the wild. A builder plan is working if you can publish and the URL is credible. A custom project is working if the paths you designed are used, the editor can run it, and you still own it.

Job you paid for	Working looks like	Not working looks like
Design / trust	Strangers can tell you from three peers in 10 seconds	“Which template is this?” from people who should hire you
IA	One primary action per key URL, in analytics	Homepages with five equal CTAs and no winner
CMS	Named editor ships a real update without Slack	Every comma is a ticket
AEO / SEO	Search Console in your account; pages get impressions; answers get cited	Studio Gmail; “we submitted it”; no property
Conversion path	Form, pre-save, book, or buy completes on a phone	Hero looks expensive; the button 404s
Ownership	You can fire the studio tomorrow and keep the site	You cannot change DNS without them
Performance	Field Vitals on a mid-range phone, not a designer laptop	Cinema on Wi-Fi, spinner on LTE

Procedure — a 30-day readout, not a vibe:

1. Pick three URLs that had a job (home + two money pages).
2. For each, write the intended action in one sentence.
3. In analytics, count completions of that action. If you have no analytics in your account, the project is not working yet — you cannot see.
4. In Search Console, confirm the property is yours and the new URLs are indexed.
5. Have the named editor perform one production update on a phone. Pass/fail.
6. Confirm billing and registrar email. Pass/fail.
7. Re-run the 15-second silent homepage test with someone who did not attend the kickoff.

• You can name the action each key page is for
• You can see that action in a property you control
• An editor who is not the designer published once
• Search Console is not a screenshot in a Slack thread
• Hosting is not the only line that “worked”

If the only metric that improved is “it looks nicer in the office,” you bought decoration. Decoration can be worth it for a brand surface. Call it that. Do not call it conversion.

What should I skip if I only have a week?

A week is not a custom website. A week is triage: pick the SKU, staff the jobs you can, and refuse work that pretends to be a rebuild. If you try to extract a $15,000-class system from seven days, you will ship a template with a new logo and call it custom.

Do this week	Skip this week
Buy or confirm the builder plan in your account	A “full rebrand site” with no sitemap
Write one job for Home, one for the money page, one for Contact	Twelve pages of lorem
Put real proof in the hero (your work, not stock)	Motion, video, and “cinematic” as a substitute for a sentence
Connect a domain you control	Letting a freelancer register it “for speed”
Create Search Console and analytics under your Google account	“We’ll add SEO later” as a package
Make the primary CTA work on a phone	A CMS with twenty optional fields
Screenshot the current URLs if you already have a site	A migration without an inventory

One-week procedure:

1. Decide: stay on the builder or stop interviewing until you have a real scope. Do not half-hire a studio for a weekend skin.
2. If staying: paid plan, your billing, custom domain, SSL (included), one offer, one CTA, real images.
3. If leaving later: do not cancel the old plan. Crawl or export URLs. You will need them.
4. Write the five jobs. Assign DIY or “not this week.” Empty is allowed. Lying is not.
5. Book a scoped conversation only if three worth-it signals from the table above are already true.

• No new registrar in someone else’s name
• No “unlimited revisions” in a seven-day window
• No AEO package without FAQ HTML you can read
• No hosting on a personal Netlify / Wix / Squarespace login you do not own
• No hero video until the title card is readable on LTE

A week can produce an honest builder site. It cannot produce the $15,000 planning example. Anyone who says otherwise is selling speed, not a system.

When is this not worth doing yet?

The custom spend is not worth it yet when you cannot name the job, the editor, or the proof. The builder spend is not worth it yet when you do not even need a domain — a Linktree and a Google Business Profile may be the real v1. Paying either number to avoid a decision is the waste.

Not yet if…	Do this instead
The offer changes every time you explain it	One-page builder, or no site, until the sentence stabilizes
You have no photos of the actual work	Shoot or gather proof. Do not buy design to hide the gap.
You will not attend discovery	Stay DIY. A studio will guess, and you will hate the guess.
You want the $15,000 example to include infinite taste rounds	That is not a project. That is a hostage situation with extra steps.
You cannot name who publishes next Tuesday	Do not buy a CMS. You will not use it.
Indexed URLs exist and you refuse a redirect map	Do not migrate. You are choosing to torch the equity.
You are shopping to punish a vendor’s $20/month price	You are still in the category error. Stop.

Checklist — wait if two or more are true:

• No stable offer sentence
• No real assets
• No named editor
• No willingness to own accounts
• No time for QA on a phone
• The only brief is “make it not look like Squarespace”

Not-yet is a valid state. A $20/month plan can cover it. A planning example of $15,000 cannot. Come back when the jobs exist. Then unbundle. Then hire.

I will not put a package price in this post. Quotes follow scope. If you want a scoped Website sprint, start from jobs, editors, and constraints — not from “can you beat Wix.” Explore the lane when you are ready to talk about a system, not a login.

FAQ

Why does a website cost $15,000 when Wix/Squarespace is $20/month?

Because those are not two prices for one product. $20/month is a hosted editor seat — hosting, SSL, templates, the vendor’s roadmap. $15,000 in the question is a planning example for labor: strategy, IA, authored design, CMS, AEO, QA, and ownership. Hosting is included in the plan. Hosting is not the five jobs. Verify current plan dollars on Wix and Squarespace; they move, and they still do not buy a brand system.

How do I measure whether a $15,000-class site is working?

Measure the jobs you paid for, not whether the sticker feels fair. Working means key pages have one intended action you can see in analytics you control, a named editor can publish without Slack, Search Console is in your account, and you could fire the studio tomorrow without losing DNS. If the only change is “it looks nicer in the office,” you bought decoration. Call it that.

What usually fails first when teams try this?

The five jobs never get staffed. Teams buy a template and expect custom outcomes, or they buy a five-figure demo skin and still rent a login. Next to fail: SEO with no URL map, a CMS nobody can use, and billing that sits on a freelancer’s card. Unbundle hosting vs design vs IA vs CMS vs AEO vs ownership before you write the second check.

How long does this take to show results?

A builder plan can be a credible URL in days if you already have copy and photos. A custom project shows up as a system at launch — paths, editor, ownership — not as a ranking miracle on day two. Search and answer-engine citations take longer and depend on whether answers actually live in HTML. If you need a 30-day readout, use the action / editor / Search Console / billing pass-fail list above, not a promised traffic spike.

What should I skip if I only have a week?

Skip the rebuild. Skip motion, a twenty-field CMS, and any “AEO package” with no FAQ HTML. In a week: your billing, your domain, one offer, one working CTA, real proof in the hero, Search Console in your Google account. A week can ship an honest builder site. It cannot ship the $15,000 planning example. Anyone offering both is selling a costume.

When is this not worth doing yet?

Skip custom spend if the offer still changes every time you explain it, you have no real photography, you will not sit through discovery, or you cannot name who publishes next Tuesday. Stay on a builder — or on no site — until the jobs exist. The $20/month plan is for that stage. The $15,000 example is not a way to buy your way out of an unstable brief.

CTA

If you are comparing a $20/month login to a $15,000 planning example, unbundle the jobs before you pick a sticker.

Explore /websites or book a Website sprint at /contact?intent=websites-sprint.