Site rebuilds cost rankings because the agency skipped 4 things you can't skip
The most common rebuild horror story: a business spends $30,000-$80,000 on a beautiful new site, launches it, and watches organic traffic drop 60% in three weeks. Their old agency disappears. The new agency says "give it time, Google needs to re-index." Six months later it's still down. They've now lost $200,000-$500,000 in leads and revenue, and the site they paid for is the asset that did it.
Almost every one of those stories has the same root cause. The agency that built the new site didn't do any of these:
- Pull the Search Console baseline first. They started designing without knowing which URLs were actually ranking, what queries those URLs ranked for, and what the impression footprint was.
- Build a URL migration map. They launched the new site without a 1:1 plan for every ranking URL on the old site. Old URLs hit 404. Google deindexes them. Equity vanishes.
- Implement redirects at the server level. They used JavaScript redirects, meta-refresh tags, or "soft" redirects that Google ignores. Equity does not transfer through these.
- Audit the redirect chain depth. They created 3-hop and 4-hop redirect chains (old URL → intermediate → another → final). Google leaks equity at every hop; chains over 2-3 hops lose ranking value entirely.
What follows is the actual procedural sequence we run for every rebuild — and the one we just used on our own site, migrating 9 years of content and 1,683 redirect rules from ultrawebmarketing.com to ultraweblabs.com without losing the position-8 ranking we hold for "web design boca raton" (publicly verifiable in our GSC). The work is procedural. The agencies that do it well don't have a secret; they have a checklist they actually run.
Red flags — your agency is about to crater your SEO
If you're evaluating an agency for a rebuild and you hear any of these, walk away. Each one is a known indicator of an agency that will leave you worse off than the site they're replacing.
-
"We'll redirect URLs as we find them"
Translation: we don't have a migration map and we don't intend to build one. You will lose URLs.
-
No mention of pulling Search Console baseline
You can't preserve what you haven't measured. Skipping the baseline means flying blind.
-
Redirects done via "the SEO plugin"
Often means JS-based or meta-refresh redirects. Verify they're HTTP 301s with curl before accepting.
-
No pre-launch staging for SEO testing
If you can't curl-test the redirects on staging before DNS cutover, you're testing in production. That's where rankings die.
-
No post-launch monitoring scope
If the engagement ends at DNS cutover, no one is watching the 60-day signals that tell you whether it worked.
-
"You'll see better rankings on the new site"
No legitimate agency promises this. Rankings depend on content + signals + competition; a new site is a fresh start, not an automatic upgrade.
-
They've never asked about your current rankings
If they don't know what you rank for today, they can't preserve it tomorrow.
-
They want to change the URL schema mid-build
Every change orphans the redirects. Schema gets locked before build starts. Mid-build changes signal disorganized process.
Questions owners ask before signing for a rebuild
-
How long should a SEO-preserving rebuild take?
For a 20-50 page site with existing rankings worth preserving, plan for 8-16 weeks. Discovery + baseline pull: 2 weeks. URL schema + migration map: 1-2 weeks. Design + build: 4-8 weeks. Pre-launch gate + redirect testing: 1-2 weeks. The variable is content — if the client owes content for the new pages, every week of late content delays launch. Don't accept "8 weeks total" for any project with substantial content scope; it's a quote designed to win the bid, not to ship the site.
-
Can I rebuild and keep my domain?
Yes, almost always preferable. Keeping the domain preserves all link equity automatically; you only need to handle URL-level redirects within the domain. Changing domains adds 12-18 months of rebuilding domain authority, which most businesses can't afford. Only change domains when there's a real reason (rebrand, brand confusion, legal). Even when we changed from ultrawebmarketing.com to ultraweblabs.com on our own rebrand, we set up domain-level 301s and continue to maintain them — the old domain still works and forwards to the new one.
-
How much organic traffic should I expect to lose temporarily?
Even with a well-executed migration, expect 10-20% temporary impression decline in weeks 1-2 as Google re-crawls and re-evaluates. Recovery to baseline typically happens by week 4-6. Improvement above baseline (from the new schema, faster CWV, etc.) typically appears by month 3-4. If you're still 30%+ below baseline at week 8, something is wrong — either redirects aren't firing, schema is broken, or content is substantially weaker than the old pages.
-
What if my existing site is on a builder I can't migrate from cleanly?
Wix, Squarespace, Webflow, and proprietary builders have varying export support. Wix is the hardest — there's no real export. Squarespace has limited content export. Webflow has decent export. The migration in these cases involves rebuilding content from scraping/screenshots and treating the new URLs as a fresh structure with a more aggressive redirect strategy. Plan for an additional 2-4 weeks of migration work and a 30-90 day recovery window.
-
Should I delete old content during a rebuild?
Yes, aggressively. Most 5-year-old sites have 40-70% of their content as thin/outdated/cannibalizing. Deleting it and 301-redirecting to the survivors strengthens topical authority on the survivors. On our own rebrand we deleted 66 cannibalizing posts and the surviving pillar pages absorbed roughly 200K monthly impressions of split authority. The work is mechanical: cluster by head term, identify the survivor, delete the rest, 301 into the survivor.
-
What about my backlinks during the rebuild?
Backlinks point to old URLs. If those URLs 301 correctly to the new URLs, the backlinks transfer equity. If those URLs 404, the backlinks are wasted and equity is lost. This is the single biggest reason redirect mapping is non-negotiable. Pull a backlink report (Ahrefs, Semrush, Majestic) before the rebuild; cross-reference with your migration map; ensure every URL that has external backlinks is explicitly preserved or redirected.
-
Should I rebuild incrementally or all at once?
For most sites under 100 pages, all-at-once is cleaner — one DNS cutover, one set of redirects, one re-indexing event. For larger sites (500+ pages), incremental section-by-section can work if you have the discipline to maintain dual URL schemas during transition. Most agencies aren't set up for incremental and try to bill it as "phased delivery" — that's usually a delivery management problem, not a strategic choice. Default to all-at-once for sites under ~100 pages.