Before you hire an analytics consultant, spend one hour trying to fail your own setup. Not because consultants are a scam, I am one, but because the hour changes the conversation. Either you find nothing alarming and buy confidence cheaply, or you find the alarming thing and walk into any engagement knowing what it is. Both outcomes beat paying someone to tell you what a network tab would have.
Here is the exact hour, five checks, in the order that catches the most with the least effort.
Check 1: who owns this thing? (10 minutes)
Open your analytics admin and your tag manager. List who has access, and at what level. Now answer three questions. Do you, the business, hold admin rights in your own accounts, or does an old agency? Is the tag manager container published by someone who still works with you? If the person who set this up disappeared today, could you change anything tomorrow?
A shocking share of businesses rent their own measurement. If your only admin is an ex-contractor's Gmail, stop the audit here and fix that first. Nothing else matters until you own the keys.
Check 2: does tracking cover the business? (10 minutes)
Write down, from memory, the five moments that make you money. A purchase, a booked call, a signup, a quote request, whatever they are. Then open your analytics and find the event for each. Not a pageview of a thank-you page that someone once said correlates with it. An event, named, counted.
Most setups I audit track dozens of things and miss at least one of the five that pay. The gap between "we have analytics" and "the moment we get paid is measured" is where most dashboard mistrust is born.
Check 3: watch the money page fire (15 minutes)
This is the check that fails most often. Open your site in a normal browser, open developer tools, network tab, and filter for collect or your analytics domain. Now do the most valuable thing a visitor can do short of paying: submit the lead form on a test entry, or run a test order if you can. Watch what actually leaves the browser.
You are looking for the event, once, with sensible parameters. Not zero times (broken), not twice (inflated), not with an empty value where revenue should be. The preview mode of your tag manager does not count as this check; preview is a rehearsal on your machine, and the difference is exactly the subject of why the GTM preview lies. The deeper version of this whole check lives in how to verify GA4 tracking.
Check 4: now decline everything (10 minutes)
Open a private window, load the site, and decline the cookie banner. Keep the network tab open. What still fires?
If analytics and ad tags fire anyway, your banner is decoration, and you have a compliance exposure nobody signed off on. If everything vanishes including things you believed were consent-exempt, you now know the size of your blind spot: the share of visitors who decline is the share your reports simply cannot see. Both findings change how you read every chart you own. The full version of this test is in how to test your consent banner.
Check 5: reconcile one number against reality (15 minutes)
Pick last month. Pull purchases (or leads) from your analytics, and the same count from the system that cannot be wrong: your shop backend, your CRM, your calendar. Divide one by the other and write the percentage down.
The point is not to reach 100%. You will not, for structural reasons: consent, blockers, measurement mechanics. The point is to know your number. A stable gap is a calibration constant. An unknown gap is a rumor, and a gap that moved since you last checked is an alarm. This one ratio, tracked monthly, is the cheapest analytics monitoring that exists. What "close enough" means and why perfection is not on the menu is covered in is 100% data accuracy possible.
Scoring yourself honestly
- Five passes: rare. Your problem is probably not measurement; go spend the energy on what the measurements say.
- One failure: fix it yourself if it is checks 1 or 2; both are admin work.
- Two or more failures, or any failure on checks 3 through 5: your reports are currently fiction to an unknown degree. Freeze big decisions that depend on the numbers, especially ad budget changes, until it is fixed.
Whether "fixed" means your developer, your agency, or an audit from someone like me is a budget question, not a shame question. What I will say plainly: do not buy dashboards, attribution tools, or more traffic on top of a setup that failed check 3. That money buys prettier fiction. The rest of the honest-measurement worldview lives in the analytics you can trust guide, and when the fix needs a second pair of hands, that is exactly what a Clarity Call is for.