A website analysis is worth paying for when it tells you what to do next, and not worth paying for when it hands you a long report and no decisions. Before you commit money to a redesign or to ads, you need to know five things: whether your numbers can be trusted, where visitors come from, where they drop off, how fast your pages load, and whether search can find you. Trustworthy tracking and where inquiries come from are the ground covered by analytics and reporting work, and you can check most of all five yourself in an afternoon. This post lists what a good analysis covers, so you can judge an analytics consultant's work before you buy it or run the first pass yourself.
Start With Whether the Tracking Can Be Trusted
Every other finding depends on your tracking, so a real analysis tests it first. Three checks matter most: key events, consent, and duplicate tags.
Key events
In Google Analytics 4, a key event is an event that measures an action important to the success of your business, and any event you collect can be marked as one. The question is whether the right actions are marked: form submissions, calls, bookings, purchases. If your contact form is not a key event, it is missing from the Key events column that reports such as User acquisition and Landing pages show, so you cannot compare channels by the inquiries they produce, and every later conclusion about channels is guesswork.
Consent
Google says consent mode lets you pass your visitors' consent choice to Google, and that when visitors deny consent, Google Analytics can fill some data gaps with modeling, depending on how consent mode is implemented. Google also says consent mode does not provide a banner. It interacts with the banner or consent platform you already use. So ask whether your banner is actually connected to your tags, because a broken connection can make your reports undercount visitors who did consent, or record visitors who did not. Whether your setup is legally sufficient depends on where your visitors live, and that is a question for your counsel, not for an analytics report.
Duplicate tags
Google warns that if the same Google tag is configured two or more times on the same page, it may cause duplicate data. That can happen when a tag is pasted into the site's code and also added through a tag manager. The test is simple: connect Google's Tag Assistant to your site, accept your own cookie banner, submit your contact form once, then open DebugView (in GA4 Admin, under Data display) and check that the form event appears once. DebugView shows the events collected from a device in real time, but it shows nothing for a visitor who has not consented to analytics cookies, which is why you accept the banner first.
Find Where Visitors Come From and Where They Drop
Acquisition data tells you which channels bring visitors who go on to take an action, and funnel data tells you which step they stop at. Read both by channel and by page, because a single site-wide conversion rate hides the page and audience differences that decide what to fix.
GA4's funnel exploration report is built for the second job: Google describes it as a way to visualize the steps users take to complete a task and see how well they succeed or fail at each step. Define the steps first, such as landing page, service page, contact page, form submitted. Then look for the one step where most people leave. Measuring drop-off at every step is how CRO consulting starts, with a funnel audit, and the usual causes at the weakest step are covered in why a website gets traffic but no leads.
Measure Page Speed With Core Web Vitals
Core Web Vitals are Google's three measures of loading, responsiveness and visual stability: Largest Contentful Paint (LCP), Interaction to Next Paint (INP) and Cumulative Layout Shift (CLS). Google's web.dev guidance sets the "good" thresholds at an LCP within 2.5 seconds, an INP of 200 milliseconds or less, and a CLS of 0.1 or less, measured at the 75th percentile of page loads and segmented across mobile and desktop.
A useful speed finding names the page, the metric that fails and the cause. "The site is slow" is not a finding. "The service page fails LCP on mobile because of an oversized hero image" is one.
Check Your Search Visibility Before You Change Anything
Search Console shows how Google sees your site: which queries and pages appear in results, how often, and at what average position. Google defines impressions as how many times your site appeared in Search results, and clicks as how many times a user clicked through, so the gap between the two is your first visibility finding.
A page with many impressions and few clicks has a different problem from a page with none. If the first already sits near the top of the results, it needs a better title or a better match to the query. The second may not be indexed or may not be about what people search for. Before any redesign, write down which pages currently earn impressions, because those are the pages a rebuild can damage. The hidden SEO cost of a website redesign explains how that loss happens. If you serve a local market, read local queries as their own group. GrossiWeb is based in Atlanta, and the same check applies in any city.
Turn the Findings Into a Short List of Decisions
An analysis is finished when every finding has a decision attached and the list is short enough to act on. Three to five decisions beat a fifty-item audit, because each one can be given an owner and a date.
Order the decisions by what depends on what. Tracking comes first, and only then does the redesign-or-ads question have an answer:
- If tracking is broken, fix it before spending anything, or you will not know what the spend produced.
- If visitors arrive but stop at one step, fix that step before paying for more visitors.
- If pages fail Core Web Vitals, fix the named causes before deciding the design is the problem.
- If few pages earn impressions, the issue is visibility, and ads will only cover the gap while you pay for them.
What to Ask Before You Buy an Analysis
Buy an analysis only from someone who can answer four questions plainly. The answers separate a diagnosis from a sales document.
- Will you test the tracking with a real form submission and a real call, not just read the dashboard?
- Will the data live in my own GA4 and Search Console accounts, so I keep it if we part ways?
- Will the result be a ranked list of decisions, each tied to a number?
- Will you tell me which findings are measured and which are your judgment?
What to Do This Week
Spend one afternoon on the free version of this analysis. Use only the tools you already have access to.
- Open GA4 and list your key events. Confirm your main inquiry action is on the list.
- Submit your own contact form once with Tag Assistant connected and check in DebugView that it is counted once.
- Open Search Console and list the ten pages with the most impressions. Do not change them before you decide what a redesign must preserve.
- Open the Core Web Vitals report in Search Console, or run your top three pages through PageSpeed Insights, and compare the field data with the thresholds above.
- Write down three decisions. If you cannot, you have found the real gap, and it is where an outside analysis would earn its fee.