There is a specific request that the website industry is bad at answering. It sounds like this:
I’m a small business. I want to redesign my website. I don’t want Wix or Squarespace. I can’t afford an agency. I don’t want to think about it. And I want to be found when people ask ChatGPT.
Ask that question anywhere and you will usually get a two-part answer: pick a nicer design tool, then hire a freelancer to drive it. That answer is not wrong, exactly. It is just a different request than the one you made. You asked for a redesigned website. You were handed a design tool, a hiring decision, a scope document, and a maintenance problem.
This guide takes the five constraints seriously, one at a time, and works out what each one actually rules out.
The five constraints, and what each eliminates
1. “Redesign,” not “build.” You have a site. It has URLs, traffic, and history. This is a migration, and migrations are where redesigns quietly destroy years of search presence. This constraint eliminates anything that treats your existing site as a blank page.
2. “Not Wix or Squarespace.” Read this literally. Most people saying it do not mean the templates are ugly. They mean they are tired of being their own webmaster. If that is what you mean, then every other builder is the same job with different buttons, and a prettier editor does not solve it. We wrote a whole comparison of the options for owners who want it handled around exactly this distinction.
3. “Can’t afford an agency.” Agency retainers are real money for real work. If that is out of reach, the question becomes which parts of what an agency does you actually need, and who does them instead. Our breakdown of 2026 website management costs has the current numbers for agencies, freelancers, and everything between.
4. “Don’t want to think about it.” This is the constraint most recommendations quietly drop, because it is the hardest one. It does not mean “easy to use.” It means you do not want to be the operator at all, not on launch day and not in eight months.
5. “Found by ChatGPT.” This one is doable, widely misunderstood, and full of people selling things that do not work.
Notice that constraints 2 and 4 are doing most of the work. Together they rule out every tool-shaped answer, because a tool needs an operator, and you have just said you do not want to be one.
Before the lanes: the option you are currently taking
The most common outcome of researching a website redesign is not choosing a platform. It is closing the tabs and keeping the site you have.
That is worth naming, because it is a real choice competing with every option below, and it is not a stupid one. Every replacement sounds either expensive or annoying. The current site is neither. It is already paid for and it does not require a decision this week.
The problem is that the cost of keeping it never arrives as a bill, so it never gets prioritized. It shows up as things that do not happen. The customer who checked your hours, found last year’s, and called someone else. The AI assistant describing services you stopped offering, or a location you moved out of. The spam links injected into a site nobody has looked at in two years, quietly pointing your credibility somewhere else. You do not get an invoice for any of that. You get slightly fewer calls, and no way to connect it to a cause.
The honest reason people stall is not the monthly price. It is the decision cost: committing budget to a thing you cannot see yet, on a promise. That is a reasonable thing to resist.
It is also the specific thing worth solving before you compare anything else, and it is why Surmado rebuilds your site first and shows it to you before you pay. You are not deciding whether the promise is good. You are looking at the finished result and deciding whether you want it.
The four lanes
Every real option falls into one of four lanes. The only question that separates them is: after this, who does the work?
| Lane | Examples | Who operates it after launch | Typical 2026 cost |
|---|---|---|---|
| Build it yourself | Wix, Squarespace, Framer, Webflow | You | $200 to $800/yr in tools, plus your hours |
| Bundled AI toolkit | Durable, 10Web, B12 | You, with better tools | $22 to $99/mo |
| Hire a human | Freelancer, agency | Them at launch, you afterward | $500 to $1,500+ one-time, or $1,500+/mo retainer |
| Managed web presence | Surmado Sites | The service, continuously | $99/mo billed annually |
Only the fourth lane satisfies constraint 4. That is not a sales claim, it is a definition: the other three hand you the keys.
Lane 1: A better design tool (Framer, Webflow)
This is the most common recommendation, and it is a genuinely good one for the right person.
Framer in particular deserves its reputation. The design ceiling is far above Wix or Squarespace, and its technical foundations are strong: it generates a sitemap and robots file automatically, gives you real metadata control, handles redirects, and supports structured data. If someone tells you Framer is bad for search, they are selling you something.
The costs to know before you commit: Basic is $10 a month billed annually and is aimed at portfolios. A business site with content tools and forms is Pro at $30 a month billed annually. Since May 2026 your first editor seat is included and additional full editors are $20 a month each.
But look at what this lane asks of you. You are now the designer, the migration engineer, and the person who remembers to add structured data to the service page you create next spring. If that sounds fine, this is a great lane and you should take it. We compared the two approaches in detail in Surmado Sites vs Framer.
Lane 2: A freelancer driving that tool
This is the second half of the standard recommendation, and it fixes the design problem while leaving the harder one open.
A fixed-scope freelancer engagement for a template customization and migration commonly runs $500 to $1,500. You get a distinctive site on launch day, which is real value.
What it does not get you is day two hundred. Fixed scope ends at launch, by definition. New service pages, changed prices, seasonal copy, a new FAQ entry, structured data on anything added later: none of that is in a build contract unless you put it there. So the honest year-two math is that you are back to being the operator, or you are paying hourly, or you are on a retainer that starts to look like the agency you said you could not afford.
If you take this lane anyway, at minimum put these in writing: existing URLs preserved or redirected, metadata carried over, a valid sitemap at launch, structured data for your organization and services, and full ownership transferred to your account. Our post on questions to ask before hiring a website company covers the rest.
Lane 3: A bundled AI toolkit (Durable and similar)
These get recommended as the hands-off, affordable option, and the pricing is genuinely attractive. Durable’s Launch plan is $25 a month, or $22 billed annually, and includes a site, a CRM, invoicing, and AI marketing tools.
One detail matters a great deal for constraint 5, and it is usually reported incorrectly: the AI visibility features are not on the entry plan. Daily AI visibility rankings and competitor analysis are on Grow, which is $99 a month, or $85 billed annually. If you are choosing a bundled toolkit specifically because you want AI visibility, the relevant price is the $85 to $99 tier, not the $22 one.
That reframes the comparison considerably, and we worked through it honestly in Surmado vs Durable. The other thing to know is that a toolkit is still a toolkit. It will show you that you are absent from AI answers for a service you offer. Closing that gap is your task, using the tools it hands you.
Lane 4: Managed web presence
This is the lane built for the request as stated. Someone else migrates the site, hosts it, keeps it current, and does the ongoing work that keeps you findable, and you never open an editor.
This is what Surmado Sites is. Surmado rebuilds your existing site free, a human checks it against a 42-point QA list, most rebuilds are ready in about 24 hours, and you approve the preview before paying anything. After that you request changes in plain English. Pro is $99 a month billed annually, or $150 month to month, with hosting, the domain, ongoing changes, monitoring, and research included. We broke down exactly what $100 a month buys if you want the itemized version.
The honest limitation: if you have strong design opinions and want to art-direct every pixel yourself, you will be happier in lane 1 with a good freelancer. Managed is for owners whose goal is a site that works and stays current, not a design project.
The part every redesign guide skips: migration risk
If you already have traffic, the redesign is not the dangerous part. The migration is.
A redesign that drops your URLs, loses your redirects, or ships new pages missing the metadata the old ones had can erase years of accumulated search presence in a single weekend. And it fails quietly. Nothing errors. Your traffic just goes down two weeks later and nobody connects it to the launch.
Every lane above can be done correctly. The difference is whose job it is. In lanes 1 and 2 it is yours or your freelancer’s, and it is frequently out of scope. In lane 4 it is the core of the service, which is why Surmado’s rebuilds take about 24 hours rather than 30 seconds: information architecture, URLs, redirects, metadata, content, assets, forms, and business context are preserved, and a human verifies the result before you see it.
Whatever you choose, treat “existing URLs preserved or redirected” as a non-negotiable line item.
Cosmetic redesign vs. actual rebuild
There is a second thing worth separating, because “redesign” gets used for both and they are not the same purchase.
A cosmetic redesign changes what visitors see. New layout, new typography, new photos. If the underlying site was slow, broken, or compromised, it is still slow, broken, or compromised, wearing a nicer outfit.
A rebuild changes both. The parts worth keeping stay: your brand, your photos, your services, your content, your domain. The parts quietly costing you get removed: broken links, dead code, hidden spam links, security holes, and anything making customers wait.
That last category matters more than owners expect. Unmaintained sites are a standing target, and a compromised site can be injected with spam links pointing somewhere else entirely without anything looking wrong on the surface. We found 32 hidden spam links on a veteran-owned moving company’s site that nobody knew were there.
So when you price a redesign, ask which one you are buying. A template customization is usually the cosmetic version.
Getting found by ChatGPT, honestly
This is where the most nonsense gets sold, so here is the boring truth.
There is no listing to buy. There is no submission form, no directory, no paid placement that puts you into ChatGPT’s answers. Anyone selling one is running a scam, and we wrote up how that scam works.
Any public website can appear. OpenAI’s own guidance is that any public site can surface in ChatGPT search. The practical requirement is that you are not blocking the crawler that matters. OAI-SearchBot is the one used for search discovery, and it is separate from GPTBot, which is about training. You can allow one and block the other. Check your robots file for both.
The technical floor is real but low. A valid sitemap, crawlable pages, working redirects, unique titles and descriptions, relevant structured data, and semantic HTML. Framer does most of this by default. So does Surmado. So, honestly, does Squarespace. This is table stakes, not an advantage.
The actual work is content and currency. AI assistants describe your business from what they can read about you. That means pages stating plainly what you do, who you serve, where you operate, what you charge or the typical range, and the questions customers actually ask. That is Answer Engine Optimization, and it is ongoing work rather than a launch-day checkbox.
Nobody can promise you a ranking. Not Durable, not an agency, not us. No one controls what a model says about your business. What you can control is whether the source material is accurate, complete, structured, and current, and whether anyone is checking what those assistants currently say so you know what to fix. If a vendor’s marketing promises a rank, discount it. If it promises to keep your source material good and to show you what AI currently says, that is a real thing to buy.
If you want the deeper version of this, we maintain a full guide on how to get found by ChatGPT.
Picking your lane
Answer one question honestly: in eight months, who do you want doing this work?
- You, and you’ll enjoy it. Framer, ideally with a freelancer for the initial build and migration. Best design ceiling of any option.
- You, but cheaply. A bundled AI toolkit. Budget the $85 to $99 tier if AI visibility is why you are here.
- Someone else, occasionally. A freelancer on a small retainer. Get the scope in writing.
- Someone else, permanently. Managed web presence. This is the only lane that satisfies “I don’t want to think about it” as stated.
There is no universally correct answer here, and any guide that gives you one without asking about your time is selling rather than advising. But if your original request was the one at the top of this post, with all five constraints held at once, the fourth lane is the only one that does not quietly hand part of the job back to you.
If that is where you land, Surmado Sites will rebuild your existing site free so you can see the finished result before deciding. A human checks it, most rebuilds are ready in about 24 hours, and there is a 7-day money-back guarantee.