The First 90 Days After Launch: What to Measure

Launch day feels like a finish line. The new site is live, the old one is gone, and everyone who worked on it can finally exhale.

Then, usually within a week, someone asks the question: “So, is it working?”

Most teams answer by opening Google Analytics and looking at total traffic. If the number went up, there’s relief. If it went down, there’s a quiet worry that the redesign was a mistake. Either way, the conversation is built on the number that tells you the least.

I want to offer a better way to spend the first 90 days, one month at a time, with a clear question for each.

Why total traffic tells the wrong story

This year, one of my clients saw both outcomes at once, in the same report. Search impressions fell 37%. Active users rose almost 80%. Depending on which number you looked at first, the launch was either a failure or a triumph.

Neither was true.

When we looked closer, clicks from search had risen 10%, and click-through rate had nearly doubled. Their summer camps page, the page most tied to enrollment, moved from position 17 in Google to position 3. Fewer people were seeing the site in search, but far more of the right people were choosing it.

And that 80% jump in users? It came mostly from paid ads running at the same time. Engagement per visitor actually fell about 42%, because ad traffic tends to browse briefly and leave.

So the big headline numbers pointed in opposite directions, and both were misleading. The real story was in a handful of smaller, more specific signals.

That client runs programs, not a museum. But I see the same pattern on museum and nonprofit sites after almost every launch: the headline numbers create the mood, and the specific signals tell the truth. That’s what the first 90 days should be about.

Before you start: name your key pages

Every measurement below depends on one decision made in advance: which pages matter most.

For a museum, that might be visit planning, exhibitions, membership, and donate. For a nonprofit, it might be programs, volunteer, donate, and contact. For a school or health organization, admissions or appointment pages. Pick three to six. These are the pages tied to your mission and your revenue, and they’re the ones you’ll watch.

It also helps to confirm three things are set up on launch day:

• Google Search Console is verified for the new site, and your sitemap is submitted. Google calls a sitemap especially helpful right after a launch or site move
• Google Analytics 4 is tracking your key actions as key events: form submissions, donation completions, ticket or registration clicks, newsletter sign-ups
• A redirect list maps every important old URL to its new home
• Your outside platforms are connected. If tickets, donations, or memberships happen on another platform, tracking has to follow visitors there (more on this in Month 2)

If any of those are missing, fix them first. Everything else is guesswork without them.

Month 1: Can search engines find your key pages?

The first month isn’t about growth. It’s about visibility in the most literal sense: can Google find, read, and index the pages you just built?

What to check:

• The Pages report in Search Console. Google’s Page indexing report shows which pages are indexed and, for those that aren’t, the reason Google gives. Google notes that it can take a week or so to start crawling a new site, so don’t worry about gaps in the first few days. Look specifically for your key pages. A missing key page is the single most important thing to catch in month one.
• 404 errors and broken redirects. Old URLs that now lead nowhere quietly lose the visibility they spent years earning. Search Console and a simple crawl will surface them.
• Your forms. Submit every important form yourself, and confirm the message actually arrives where it should. A broken form after launch is more common than most people expect, and it is invisible in traffic reports.
• SEO titles and descriptions. Make sure each key page has its own, rather than a default or a duplicate.

I cover several of these in more detail in 5 Website Health Checks Every Organization Should Run, and they’re worth repeating after any launch.

What to ignore in month one: total traffic, and day-to-day ranking changes. Google is still re-learning your site. Its own guidance on site moves tells site owners to expect temporary ranking fluctuation, and reacting to that movement usually creates more problems than it solves.

Month 2: Are people reaching key pages and taking action?

Once your pages are indexed, the question shifts from “can they find us?” to “are the right people getting where they need to go?”

What to check:

• Visits to your key pages. Not total sessions, but sessions on the three to six pages you named. Are they steady or rising compared to before launch?
• Completed key events. Forms submitted, donations completed, registrations started. This is the closest thing your website has to a mission outcome
• The path to action. If people visit the donate page but rarely complete a gift, or reach the volunteer page but never submit the form, that gap tells you exactly where to look next
• Engagement on key pages. GA4’s engagement rate on those specific pages tells you whether visitors are reading or leaving

If your donations or tickets happen somewhere else

This is the most common measurement gap I find on museum and nonprofit sites, and it’s easy to miss.

Many of your most important actions don’t happen on your website at all. Tickets are sold through Tessitura, Eventbrite, or another ticketing system. Donations go through Givebutter, Bloomerang, DonorPerfect, or a similar platform. Memberships often have their own portal. The visitor starts on your site, clicks a button, and finishes somewhere else.

If tracking isn’t set up to follow that handoff, Google Analytics sees a visitor click “Donate” and then disappear. Your website did its job, but the report says it didn’t. Worse, the completed gift may show up in analytics as if it came from the donation platform itself, which hides the fact that your website sent the donor there.

There are three practical ways to close this gap, depending on the platform:

• Cross-domain measurement in GA4, when the platform allows your analytics code to run on its pages
• Click tracking on your Donate, Buy Tickets, and Join buttons, so you at least measure every handoff
• Comparing reports side by side, matching handoff clicks from your site against completed transactions in the platform’s own reporting

Even the simplest version, counting button clicks, tells you far more than total traffic does. It’s the clearest evidence that your website is sending people toward your mission.

What to ignore in month two: sitewide averages like overall bounce rate or average time on site. Those numbers blend your most important pages with your least important ones, and the result describes no real visitor. Watch the pages that matter, not the average of everything.

Month 3: What brings people to you?

By the third month, you have enough data to see patterns in how people actually arrive.

What to check:

• Search queries in Search Console. The Performance report shows the words people typed before finding you. This is one of the most useful pieces of information a website produces, because it tells you how the public describes what you do, which is often different from how you describe it internally.
• Clicks, click-through rate, and position for key pages. Rising clicks and a better click-through rate usually mean your titles and descriptions now match what people are looking for, even if total impressions are lower. Google itself recommends focusing more on clicks and impressions than on position alone.
• AI referrals. Google Analytics now has an AI Assistant channel that groups visits from tools like ChatGPT, Gemini, and Copilot, so you no longer have to dig for them. Two caveats are worth knowing. Clicks from Google’s own AI Overviews are counted as Organic Search, not AI Assistant. And AI visits that arrive without referral information still show up as Direct, so the channel shows the minimum, not the full picture. It’s also worth checking your Referral report for perplexity.ai. These numbers are usually small at first, so the trend matters more than the total. A steady trickle means AI tools are starting to read and recommend your site, which is a good sign that its structure is clear.

A note on seasons: museums and nonprofits run on calendars that most website advice ignores. An exhibition opening, a gala, year-end giving, GivingTuesday, or the start of the school year can all shift your numbers sharply. Year-end alone is significant: according to the M+R Benchmarks study, summarized by Neon One, December accounts for roughly 40% of the average nonprofit’s annual online one-time giving. If you launched just before one of these, your first 90 days may look unusually strong or unusually quiet. Where you have last year’s data, compare against the same season, not only the previous month.

What to ignore in month three: third-party “authority” scores and single-week swings. Scores from SEO tools are estimates that Google doesn’t use. And one unusual week, a holiday or a press mention, says very little on its own. Compare month to month, and season to season where you can.

What to tell your board

By day 90, you can replace “traffic is up” or “traffic is down” with five clearer questions:

1. Can Google find and index our key pages?
2. Are people reaching the pages tied to our mission?
3. Are they completing an action?
4. What words bring them to us?
5. Are AI tools starting to send visitors?

These answers are more honest, and they’re also more useful. They point to specific decisions, like rewriting a page title, fixing a form, or improving a page people visit but rarely act on, instead of a general sense of good or bad news.

A smaller number of the right visitors, taking action, is what a website is built for.

The launch is the beginning of the record

The first 90 days set a baseline. Everything you measure after that is compared against it. That’s why this period matters so much: it’s when you learn what “normal” looks like for the new site, and when small problems are cheapest to fix.

It’s also the point where many organizations stop paying attention. The launch team moves on, nobody owns the reports, and the site slowly starts to drift, which I’ve written about in Why Every Institutional Website Requires Maintenance. A website is a living system, and it keeps changing whether or not anyone is watching it.

If you’d like someone to watch these signals with you every month, the Sunlight Monthly Retainer is built for exactly this: ongoing care, monthly measurement, and clear explanations of what the numbers mean and what to do about them. It’s how a launch keeps paying off long after launch day.