Quick Answer
An SEO and web design agency builds the site and runs the search work on it under one contract. Buying both from one team is sensible while nothing ranks yet, because the coordination saving is real and the downside is small. Once search brings customers, the rebuild is the risk (rendering, URLs, redirects and the missing baseline), so keep an independent reviewer for the migration window and hold your own measurement.
Search for this category and page one hands you city directories and local studios. Nobody there answers the question you actually have, which is who should own the rebuild.
The bundle is sold as a saving, and the saving is usually genuine. What it also removes is your control group: when the team that ships the site is the team reporting on the search result, a flat quarter has no independent read.
The expensive failures here are not design failures. They are four things a rebuild quietly does to a site that search only notices weeks later.
This covers what those four are, how bundled and split teams differ on risk, and the five questions that expose a rebuild plan before you sign.
What Breaks When a Site Is Rebuilt Without Search
Four things break when a site is rebuilt without search in the room: how pages render, what the URLs are, whether redirects exist on launch day, and whether anyone kept a baseline. A design failure is visible the morning after launch. These are not, which is why they get blamed on demand three months later.
None of them is exotic. An SEO and web design agency that has shipped a migration before raises all four unprompted, and that is the cheapest signal you get on a first call.
1. Rendering Hides Your Content From Crawlers That Skip JavaScript
A modern rebuild often ships as a single-page application, where the first HTML response is an empty shell and the words arrive only after JavaScript runs. Googlebot renders that shell, so the page usually survives Google. Not every crawler does, and the engines behind AI answers often skip JavaScript, so what they receive is the empty shell.
Google's JavaScript SEO basics is direct about the fix: “server-side or pre-rendering is still a great idea because it makes your website faster for users and crawlers, and not all bots can run JavaScript.” Server-side rendering or a prerender step puts the text back into the file everyone receives.
The same bar applies to AI answers. Google's documentation on AI features says a page has to be indexed and eligible to be shown with a snippet before it can appear as a supporting link in AI Overviews. A page whose text never reaches the HTML struggles at the first step.
Check it the way a crawler would. View source on a rebuilt page, search the raw HTML for a sentence you can see on screen, and if it is missing, the page is blank to anything that does not run scripts. We go further into that split in why Googlebot renders React and AI crawlers do not.
2. New URLs Arrive Without a Map From the Old Ones
A rebuild renames things. A services path becomes a solutions path, the blog moves under resources, and case studies gain a year in the URL. Each of those is a brand new page as far as an index is concerned.
Google's guidance on how to move a site asks for the mapping before the move, not after it: list the old URLs from your sitemap, your server logs and your analytics, then decide where each one lands. It also names the lazy version directly. “Don't redirect many old URLs to one irrelevant single URL destination, such as the home page of the new site,” it says, because that behaves as a soft 404.
The check is mechanical and takes an hour. Pull the top fifty URLs by impressions from Search Console, run them against the staging site, and count how many resolve to a page about the same subject.
3. Redirects Get Written After Launch, or Not at All
The redirect map is the one deliverable nobody sees in a design review, so it is the one that slips. A rebuild can look finished, ship on time, and drop its old landing pages into errors that nobody sees in a browser.
Use permanent redirects and keep them. Google's redirect documentation separates the two cases plainly: a permanent redirect shows the new target in search results, while a temporary one keeps showing the source page. HTTP 301 set server-side is the one you want, and Google's site move guidance says to keep redirects as long as possible, generally at least a year, so that signals transfer.
Ask for the map as a file, dated before launch, with a named owner. Our website redesign checklist covers what else gets locked down in that week.
4. Pages That Earned Traffic Get Cut for Design Reasons
Rebuilds delete pages. A long explainer that ranks does not fit the new template, three service pages collapse into one, and the questions section comes out because it makes the page feel heavy.
Every one of those is a ranking decision taken on visual grounds. The fix is one list before design starts: which URLs bring impressions, which bring leads, and which are genuinely allowed to disappear.
Merging is fine when it is deliberate. Google's site move guidance is explicit that content consolidated from several pages can redirect to the new single page — what it warns against is scattering unrelated old URLs onto the homepage.
5. Nobody Kept a Baseline From Before the Rebuild
Without a snapshot from the week before launch, every conversation after launch is opinion against opinion. Traffic fell, the build team says the market softened, and you suspect the migration.
The snapshot costs an afternoon. Export ninety days of Search Console queries and pages, export the same range from analytics, save the top fifty URLs with their positions, and keep all three outside any vendor's account.
Then set the review date in advance, six to eight weeks after launch, and write down now what a bad result would look like. A number agreed before launch is evidence; the same number argued afterwards is a negotiation.
How Bundled and Split Teams Differ on Risk
Bundled and split teams differ on one axis that matters, which is who can tell you the rebuild worked. Cost, speed and meeting count move less than the pitch on either side suggests.
| Factor | One Bundled Team | Split or Reviewed |
|---|---|---|
| Who grades the rebuild | The team that built it | Someone with nothing to defend |
| Coordination cost | One contract, one meeting | Two calendars to keep aligned |
| Migration risk | Carried by the builder | Checked before launch day |
| Time to launch | Faster by a week or two | Slower by the review window |
| Fits when | Pre-launch, nothing ranking yet | Search already brings customers |
Read the rows against what you already have. With nothing ranking, the bundle wins on every practical row and the grading question costs you nothing at all.
With traffic to protect, the third column is usually the honest answer, and it is narrower than it sounds. One builder, one reviewer for the migration window only, and no second retainer running forever. Better Marketing works this way on the build itself, because the site is the first stop a buyer makes and search belongs in the room while it is being designed.
Note what a bundle usually does not include. Most SEO and web design agency contracts cover the build and the on-page work, then stop short of content marketing for tech companies, which is a different retainer with a different team.
That is a related question, and it has its own answer in whether one agency should run SEO and content.
Five Questions That Expose a Rebuild Plan
Five questions separate an SEO and web design agency that has shipped migrations from one that has only shipped designs. Ask them on the first call and listen for specifics rather than reassurance.
- Who owns the redirect map, and can I see it dated before launch? You want a named person and a file, not an assurance that it gets handled.
- Will the new pages carry their text in the HTML before JavaScript runs? The answer should be a rendering method, not a yes.
- Which current pages are you planning to remove? If nobody has looked at impressions per URL, nobody has looked.
- What does the first week after launch look like? Crawl errors, index coverage and a rollback trigger, or it is not a plan.
- What would prove to you that the rebuild underperformed? Anyone who can answer this has been held to a number before.
A vague answer is not disqualifying by itself. It tells you where the second opinion belongs, and question one predicts the rest more reliably than the others.
That second opinion is smaller than a retainer. An AI SEO service scoped to a migration window is a few weeks of work: check the rendering, check the map, check the HTTP 301 responses on launch day, then read the recovery.
What Stays Yours Whoever Builds the Site
Three things stay in your name regardless of who builds the site, and all three are free.
- The analytics property, created under your own account rather than the agency's.
- The Search Console property, verified by you, with the agency added as a user you can remove.
- The pre-launch export: ninety days of queries, pages and positions, stored outside any vendor's tooling.
An SEO and web design agency that prefers to hold these on your behalf is rarely being difficult on purpose. It is a habit carried over from local-services work, and it quietly makes the vendor the only source of truth about its own results.
The same rule covers what comes after the build. Content marketing for tech companies runs on this data, and a B2B demand generation agency you hire next year will ask for the same exports in its first week.
When Splitting the Work Costs More Than It Saves
Splitting costs more than it saves when there is nothing yet to protect. A pre-launch site with no rankings, a handful of pages and no organic traffic has no signal to lose and no baseline worth defending.
In that state a second opinion buys you a slower launch and a longer invoice. The coordination tax is real: two vendors disagreeing about a heading costs a week, and that week is worth more than the ranking risk you are insuring against.
The same holds when your traffic is not search traffic. If every customer arrives through a newsletter or a founder's LinkedIn posts, a migration review protects something you do not have, and the money does more with a B2B demand generation agency.
Two conditions flip it back. Any page currently earning impressions, or a plan to earn them inside the year, and an independent read on the SEO and web design agency's migration work is worth its cost again.
Decide by What the Site Already Earns
An SEO and web design agency can own both jobs, and what settles it is what the site earns today. Nothing ranking yet means bundle it and move faster. Search already bringing customers means keep one builder and add an independent read across the migration window.
Either way, hold your own analytics, your own Search Console and a baseline export from the week before launch. Better Marketing runs the AI SEO service half of that work, the part about being found on Google and inside AI answers, and it starts with whether the page can be read at all.