The Website Launch Playbook: 8 Steps To Going Live Without Tanking Your Business

Most website launches are celebrated as finish lines. They should be treated as the riskiest moment in a brand’s life. I had a client I worked with for several years who recently relaunched on their own, without bringing me in for the work. The new site is absolutely beautiful, but within weeks, they’ve lost more than half their organic traffic.

And the new business that went with it.

The problem was everything they skipped before, during, and after going live. The redirect mapping was incomplete, the backlink equity was never accounted for, and no one baselined performance, so they could not even diagnose what broke. By the time the drop showed up in their reporting, the damage had compounded for weeks, and clawing back that authority is far harder than preserving it would have been.

A modern launch is not a deployment. It’s a controlled migration of accumulated value, including rankings, link equity, conversion paths, and user trust, from one environment to another. Here is how to do it without throwing that value away.

Step 1: Start From a Finished Design, Not a Live Deadline

The clock should not start until the new site is genuinely done. That means the design is approved, content is in place, templates are built out, and the staging environment is a faithful representation of what will go live. Launching against a calendar date rather than a readiness state leads teams to redirect to pages that don’t yet exist and discover broken conversion paths in production. Treat design accomplished as the prerequisite, then begin the migration work described below.

Step 2: Crawl and Inventory the Existing Site Completely

Before you touch anything, you need a complete map of what currently exists and what it’s worth. Run a full crawl of the live site with a tool like Screaming Frog and export every indexable URL. This is your source of truth for redirect mapping. But a URL inventory alone is the mistake that cost my former client half their traffic. A URL is not just an address; it’s a container for accumulated authority.

This is where a backlink audit becomes non-negotiable. 1 or your platform of choice, and identify which specific URLs are carrying the weighted, authority-driving links. A homepage with a thousand referring domains and a buried resource page that earned links from high-authority publications are not interchangeable.

When you map old URLs to new ones, those link-bearing pages must redirect to the most topically relevant destination with a 301, not get swept into a blanket redirect to the homepage. Redirecting everything to the root is one of the fastest ways to evaporate domain authority, because Google consolidates far less link equity through an irrelevant catch-all redirect than through a precise one-to-one match.

So the inventory has two layers:

1. Every URL for completeness.
2. The subset of URLs that hold meaningful backlink weight for prioritization.

The second list is where you spend your careful attention.

Step 3: Baseline Everything While the Old Site Is Still Live

You cannot measure a decline without a starting point. Before launch, capture a comprehensive baseline across every platform so that post-launch comparisons are grounded in real numbers rather than memory.

• In GA4, record your baseline traffic by channel, sessions, engaged sessions, conversions, and key events, ideally exporting the trailing 12 months to build in a year-over-year context.
• In Google Search Console, export your performance data, including clicks, impressions, average position, and the queries and pages driving them, along with your current index coverage and any existing issues.
• Capture Core Web Vitals for both the field data in GSC and lab data through PageSpeed Insights or Lighthouse, so you know your LCP, INP, and CLS before any template changes.
• Pull your SEO ranking data for your priority keywords from your rank tracker.
• Document your conversion metrics, form completion rates, and the engagement metrics that matter to the business.

Save all of this as dated artifacts. These snapshots are what let you, weeks later, prove whether a traffic change is a launch problem or normal seasonality.

Step 4: Test Conversion Paths With Real Data and Real Transactions

Staging tests with dummy data routinely pass while production quietly fails, because the things that break in production are exactly the things that test data papers over. Payment processors behave differently in live mode than in sandbox. Form integrations to your CRM, email platform, and analytics fire differently when real systems are connected. Address validation, tax calculation, and shipping logic only reveal their edge cases against real inputs.

Before and immediately after launch, walk every form and every conversion path end to end with real data, including real credit cards on real transactions you can later refund, real email addresses you control, and real phone numbers.

Confirm that each submission actually lands where it should: that the lead appears in the CRM, the order processes and the confirmation sends, the analytics event fires and is attributed correctly, and the thank-you or confirmation step renders.

A form that submits without error but never delivers the lead is worse than a form that’s visibly broken, because no one notices until the pipeline runs dry.

Step 5: Go Live as a Controlled Event

When you flip the switch, do so deliberately, with your eyes on the instruments. Push your redirect rules and confirm they resolve with proper 301 status codes, not 302s and not chains.

Verify the robots.txt is not accidentally disallowing the site, that the staging noindex tag has been removed, and that canonical tags point where they should. Submit the new XML sitemap to GSC and request indexing on priority pages.

Monitor server logs and analytics in real time for the first few hours to catch anything obviously wrong, such as a 404 spike or a tracking tag that didn’t deploy.

Step 6: Immediately Run an Exhaustive Post-Launch Audit

Once the site is live, repeat the full crawl against the production environment. This post-launch crawl report is where you catch the gap between intention and reality: redirects that didn’t fire, pages that 404 when they should redirect, redirect chains and loops, orphaned pages, missing or duplicated metadata, broken internal links, and any page that slipped into a noindex state.

Cross-reference this crawl against your pre-launch URL inventory to confirm that every old URL, and especially every backlink-bearing URL from your audit, resolves correctly to its intended destination. This is the single most important verification step, and it’s the one my former client skipped.

Then re-run every report you baselined. Pull fresh GA4, GSC, Core Web Vitals, ranking, conversion, and engagement data so you have a matching post-launch snapshot to compare against your pre-launch numbers.

Step 7: Annotate Across Every Platform

Drop a launch annotation in GA4 marking the exact date and time, and note it anywhere you’ll later be reading data, including your rank tracker, your reporting dashboards, and any internal documentation.

Months from now, when someone questions a trend line, that single marker is the difference between an informed conversation and a guessing game. Note major follow-on changes the same way, so every inflection point in the data has a documented cause.

Step 8: Compare on a Schedule, and Read Year Over Year

A launch is not validated in a day. Establish a regular cadence of comparing post-launch performance against your baseline, weekly for the first several weeks, then tapering as the picture stabilizes. Watch organic traffic, rankings on priority terms, index coverage, Core Web Vitals, and conversion rates against the numbers you captured before launch.

Crucially, read the data year over year, not just week over week. A traffic dip in a given week might be seasonal rather than a launch wound, and only the YoY view tells you which. If organic sessions are down twenty percent against last week but flat against the same week last year, you may be looking at normal seasonality.

If they’re down against both, you have a problem to chase, and your annotated, baselined data gives you the thread to pull, whether that’s a failed redirect, a high-authority backlink pointing to a 404, or a template change that gutted your Core Web Vitals.

The Lesson

The new design is the easy part, and it’s the part everyone focuses on. The value lives in the invisible infrastructure: the precise redirect map weighted by backlink authority, the baselined data that makes problems diagnosable, the conversion paths proven with real money, and the disciplined comparison schedule that catches a decline while it’s still recoverable.

My former client lost half their traffic, not because they built a bad site, but because they treated the launch as a finish line rather than a migration. The work that prevents that outcome isn’t glamorous, but it’s the difference between a relaunch that grows the business and one that quietly dismantles years of earned authority.