90-Day Optimisation Window: When Websites Get Better

The "site is live" email lands on a Friday. There's a moment of celebration, the agency closes the project, your team moves to the next priority, and the site stands exactly as it was on launch day. Six months later, the numbers haven't moved much. Traffic is coming in. Leads aren't coming out.

That gap — between a site that gets visitors and a site that converts them — is what happens when website post launch optimisation is treated as optional. The site at launch is the least optimised version it will ever be. Real visitor data starts flowing the moment you go live, and without a structured process to act on that data, you're leaving the most valuable learning period of a website's life unused.

The 90 days after launch are where a website earns its budget. Not in the build, not at the design sign-off meeting, and not on the day you hit publish. In the weeks of iteration that follow, grounded in how real visitors actually behave on the live site.

Why Launch Day Is the Starting Line, Not the Finish

Every design decision made before launch was based on a hypothesis. You hypothesised that visitors would find your navigation intuitive, that your primary call to action was compelling enough, that your contact form wouldn't cause drop-off, that your services page held attention past the first scroll.

Some of those hypotheses will be correct. Many won't. You won't know which until real visitors arrive and behave differently to what was expected. Usability reviews and stakeholder sign-offs are useful, but they're not a substitute for real user data at scale.

When there's no post-launch process in place, failed hypotheses stay in production indefinitely. The form that's too long stays long. The pricing page with a buried call to action stays buried. The mobile navigation element that's hard to tap on a small screen continues frustrating visitors on the devices that now account for more than half of most sites' traffic. Every day without a fix is another day of traffic that didn't convert.

The cost is quiet but it compounds. Traffic investment continues — SEO, paid channels, content — while the site's conversion rate stays anchored to its launch-day level. The gap between actual and potential performance grows wider the longer it goes unaddressed.

What Website Post Launch Optimisation Actually Looks Like

The 90-Day Optimisation Window, as defined within the Built to Last™ 2.0 framework for web, is a structured engagement built around four core activities. It sits within P06 — The Right Team — because optimisation requires the right people acting on data with the authority to make changes. A process without accountable people isn't a process; it's a good intention.

Analytics review. Weekly review of real visitor behaviour using GA4 data. Which pages attract the most traffic? Which have the highest exit rate? Where do users drop from conversion funnels before completing a goal? The analytics review is where raw data becomes actionable insight — but only if GA4 was configured correctly before launch. A default GA4 install that tracks sessions and pageviews tells you people arrived. It doesn't tell you what they did, what stopped them, or where they went when they left. Conversion event tracking, scroll depth, form interactions, and goal completions need to be configured deliberately, before go-live.

Search Console monitoring. Google Search Console surfaces crawl errors that prevent pages from being indexed, click-through rates by search query, and ranking movement in the weeks after launch. Ranking volatility in the first 90 days is normal — search engines are re-assessing the site's authority and relevance against its live URL. Errors caught early are recoverable. A crawl error that sits undetected for three months can represent significant lost indexing coverage and SEO investment wasted on pages that were never found.

Conversion optimisation. Using analytics data to identify specific drop-off points, then forming a testable hypothesis about the cause. A long contact form might cause drop-off before submission. A vague call to action might produce hesitation. A dense services page might lose visitors who can't quickly identify whether the offer is relevant to them. The job of conversion optimisation at this stage isn't to redesign — it's to make the smallest possible change that solves a specific, measured problem, then observe what changes in the data.

A/B testing where traffic supports it. Sites with sufficient volume can run structured split tests on high-traffic pages, comparing two versions of an element directly to see which performs better. Sites below the threshold for statistical significance use qualitative tools instead — session recordings, scroll depth heatmaps, click maps — to understand user behaviour without requiring large sample sizes. The tool changes with the traffic level. The discipline of hypothesis-driven iteration stays constant regardless of scale.

Who needs to be in the room

Three roles make post-launch optimisation function: someone who reads and interprets the data, someone who makes changes to the site, and someone who understands what the business needs from those changes. Without the business perspective, optimisation becomes a technical exercise. Without an analyst, it becomes guesswork. Without a developer or CMS editor, it becomes a wish list that never ships.

What gets documented

Every change should be recorded: what changed, why, what the data showed before, and what it showed after. Over 90 days, this log becomes the evidence base for every subsequent decision. At the end of the window, it becomes the brief for the next phase of the site's development. The discipline of documentation is what separates a 90-day optimisation window from three months of ad hoc tweaks with no institutional memory.

Running the Window: What to Prioritise and When

The prerequisite for everything that follows is a functioning data foundation. If GA4 isn't tracking the events that matter — form submissions, goal completions, checkout steps, scroll milestones — or if Search Console isn't verified and returning data, fix this first. You cannot improve what you cannot measure, and conclusions drawn from broken tracking aren't reliable.

A Tale of Two Sites

The value of the 90-day window is clearest in comparison.

Consider two comparable service businesses — similar industry, similar traffic levels at launch, similar design quality. The first launched, closed the agency engagement, and made no deliberate changes for the following twelve months. By month twelve, organic traffic had grown modestly through passive SEO accumulation, but conversion rate was unchanged from launch day. The site was generating visits. It was not generating business proportionate to the investment in those visits.

The second ran structured weekly analytics reviews from week one. In the composite pattern this scenario represents — the trajectory we see when active post-launch optimisation is in place from the start — the team identified three friction points in the first month: a contact form with too many required fields for the type of enquiry being requested, a services page where the primary call to action was positioned below the fold on mobile screens, and a page load delay caused by unoptimised images in the hero section. Each was addressed with a targeted fix. No redesign, no rebuild — three specific changes in response to specific data.

In the scenario this composite represents, conversion rates improved by roughly 18% over the 90-day window. Imagine a business investing in organic search and content to drive, say, 500 visitors a month — a lift of that order means meaningfully more enquiries without any additional traffic acquisition spend. The same visitors, the same brand, the same underlying offer — but a site that was learning from real behaviour rather than repeating the same launch-day assumptions indefinitely.

The difference between these two trajectories was not budget, design quality, or the capability of the teams involved. It was the presence of a structured process to act on what the data was already showing.

When This Is Non-Negotiable — and When You Can Simplify It

The full 90-day optimisation programme is non-negotiable for any site where the primary measure of success is a conversion metric: lead generation forms, eCommerce checkouts, SaaS trial sign-ups, booking or enquiry flows. If the site's performance is measured by what it converts, you need a process to improve conversion systematically — and the first 90 days are when the data to do that is freshest and most actionable.

It matters most when traffic volume is sufficient to generate meaningful data, when the conversion goal is clearly defined and tracked, and when there is internal or agency capacity to act on what the data reveals within the same week it's reviewed.

For pure brochure sites — where success means brand presence rather than an active conversion target — the minimum viable version is: verify Search Console, confirm GA4 is configured correctly, and review the data monthly. You don't need a full optimisation programme for a site with no active conversion goal. But you do need the data foundation regardless.

Where teams can reasonably defer the full programme: if analytics wasn't configured properly before launch, the first two weeks should repair the data setup rather than attempt optimisation. Drawing conclusions from misconfigured tracking is worse than drawing no conclusions at all.

Where to Start This Week

Run an analytics audit on your current site. Check that GA4 is tracking the events that matter — not just sessions, but form completions, conversion goals, and the specific actions that indicate a visitor became a lead. Verify that Google Search Console is connected and showing no critical crawl or indexing errors. Write down, in a single sentence, what a conversion means on this site.

If those three things are in order, schedule your first weekly analytics review. Bring the data, a hypothesis about what it's telling you, and someone with authority to make a change based on what you find. The 90-day window doesn't require a programme to begin — it requires a decision to use the data the site is already generating.

For businesses looking at web design and delivery that builds the post-launch accountability layer in from the start — analytics configured before go-live, Search Console verified, optimisation process agreed before launch day — our web design and delivery process sets out how we structure engagement from discovery through to the optimisation window.

Frequently Asked Questions