Most law firms have GA4 installed and almost nobody uses it. The property was set up by whoever built the site, the default reports opened once, and the numbers never matched anything intake could confirm, so the tab quietly stopped getting opened. That is a rational response to a badly configured tool. GA4 out of the box is a general purpose product built for ecommerce and app publishers, and it buries the handful of views a professional services firm actually needs under dozens it does not.
The fix is subtraction, not more dashboards. A law firm needs six reports, four or five well defined key events, and a habit of reconciling against the intake log once a month. Everything else in the interface can be ignored permanently without losing a single decision. Here is how to configure that once and what to stop looking at.
Before Any Report, Fix What Counts as a Conversion
No report is worth reading until the events underneath it are correct. This is where the majority of law firm GA4 properties fail, and it is why the numbers never reconcile with intake.
The default GA4 install fires a set of automatic events including page_view, scroll, click on outbound links, and file_download. GA4 also lets any of these be marked as a key event, which is the renamed version of what used to be called a conversion. The temptation is to mark several of them so the dashboard shows healthy numbers. Resist it. If scroll depth or a phone number click on the header is marked as a key event, your conversion count inflates by a factor of five or more and every downstream ratio becomes fiction.
For a firm running a website plus phones, this is the working set.
- form_submit_contact fired on the confirmation state of the main contact or case evaluation form, not on button click. Button click counts abandoned validation errors as wins.
- form_submit_practice fired on any practice area page form, kept separate so you can see which practice areas convert without slicing.
- call_click_mobile fired on tel link taps from mobile and tablet only. On desktop a tel click means almost nothing.
- chat_lead fired when a chat widget captures a name and a matter description, not when the widget opens.
- qualified_call imported from your call tracking platform, defined by duration threshold, typically 60 or 90 seconds.
Mark those as key events and nothing else. Two notes on mechanics. First, send each one with parameters that survive analysis, at minimum page_location, form_id, and a practice_area value you set from the page template. GA4 will not report on a custom parameter until you register it as a custom dimension in the admin, and registration is not retroactive, so do it the same day you start sending the parameter. Second, respect the naming and structure rules in Google’s GA4 events reference, because event names are case sensitive, capped at 40 characters, and cannot be renamed after the fact without breaking historical continuity.
One more piece of hygiene that costs ten minutes and saves a quarter of bad data. Under data streams, turn off the automatic form interaction events. They fire on generic form elements the tag manager cannot distinguish, and they will double count against your deliberate form events. Also add your own office IP ranges and your agency’s to internal traffic filters, then set the filter to active rather than testing. A twelve person firm browsing its own site can account for a meaningful share of sessions on a low volume property.
Report One, Traffic Acquisition by Session Default Channel Group
This is the only top of funnel report worth a standing look. Set the primary dimension to session default channel group, add key events as a secondary metric column, and compare month over month rather than week over week, because legal volume is too spiky at weekly granularity to read.
What you are looking for is not the total. It is the ratio of key events to sessions inside each channel. Organic search sitting at two percent while direct sits at nine percent is not a sign that direct is a better channel. It usually means brand searches and returning prospects are landing in direct, which is a different intent entirely. Read channels as populations with different readiness, not as a leaderboard.
Watch for unassigned traffic. If more than about five percent of sessions land there, something is wrong with your tagging, usually missing UTM parameters on paid or email campaigns, or a consent banner blocking the initial page view.
Report Two, Landing Page and Screen
Landing page is where content decisions actually get made. Sort by sessions, add key events, and then do the thing most people skip, which is to add a comparison for organic search only so paid traffic does not contaminate the view.
The pattern that matters for a law firm is a practice area page pulling real organic sessions with a key event rate well below your site average. That is a conversion problem, not a traffic problem, and no amount of additional content fixes it. A page bringing 400 organic sessions a month at a 0.8 percent conversion rate is leaking more potential cases than a page bringing 60 sessions at 4 percent, and the fix is layout, proof, and form placement rather than another thousand words. That work is its own discipline, covered in our guide to conversion rate optimization for law firm websites.
Report Three, Pages and Screens Filtered to Practice Area Templates
Different question from landing pages. This one tells you what people read after they arrive, which reveals the practice areas prospects actually care about versus the ones the firm assumes it is known for. Filter the report to your practice area URL path and sort by views.
Firms are routinely surprised here. A department the partners consider secondary often outdraws the flagship by three to one, because search demand does not care about the firm’s internal org chart. That is a hiring and intake staffing signal as much as a marketing one.
Report Four, A Custom Exploration for Path to Conversion
Build one path exploration in the Explore section, starting from your key event and walking backward. Legal buying is rarely single session. The realistic pattern is a blog or informational page first, a practice area page days later, then an attorney bio, then the form. If you only look at last click, the informational content looks worthless and gets cut, which is how firms accidentally dismantle the top of their own funnel.
Be honest about the limits. GA4’s default data retention is two months unless you change it to fourteen in the admin, and almost nobody changes it, so most properties cannot look back far enough to see legal buying cycles. Change that setting today. Even at fourteen months, cross device journeys break, and a prospect who researches on a phone and submits on a laptop looks like two people. Path analysis tells you which content participates in conversions. It does not tell you which content deserves credit, and pretending otherwise leads to bad budget calls. The credit question is a different exercise entirely, which we work through in attributing law firm marketing to signed cases.
Report Five, Key Events Over Time with Practice Area as a Dimension
A simple free form exploration. Rows are your practice_area custom dimension, columns are the key event names, and the date range covers a rolling quarter. This one report replaces about six conversations a year, because it converts a vague sense that things are going well into a table a managing partner can read in under a minute.
Keep the event types separate rather than summing them. Form submissions and calls behave nothing alike. Calls skew heavily toward mobile and toward urgent matter types, and form fills skew toward research heavy matters and after hours submission. A blended number hides both.
Report Six, User Acquisition Restricted to New Users
Used sparingly, once a quarter, for one purpose, which is checking whether the marketing spend is bringing genuinely new people or recycling the same audience. If new user counts are flat while sessions climb, you are paying to re engage people who already know the firm. That can be fine for a brand campaign and is a problem for a growth campaign, and the distinction is worth confirming before renewing a contract.
What to Ignore Permanently
These consume attention and return nothing for a law firm.
- Engagement rate and average engagement time as goals. They are diagnostic context at best. A prospect who lands, reads a phone number, and calls in 30 seconds is your best outcome and your worst engagement metric.
- Realtime. Useful for verifying a tag fired during setup, useless afterward, and addictive in a way that wastes real hours.
- Demographics and interests. Thin coverage on low volume properties, modeled rather than observed, and not actionable for a firm that serves whoever has the matter.
- Every ecommerce and monetization report. Nothing populates them.
- Bounce rate. It returned as an inverted engagement rate and carries none of the meaning people remember from the old tool.
- Google Signals demographic reporting expectations. Thresholding will suppress rows on small properties and you will conclude data is missing when it is being withheld.
Reconciling GA4 With What Intake Actually Logged
This is the step that decides whether anyone in the firm trusts the analytics, and it is a monthly thirty minute job.
Pull three numbers for the same calendar month. GA4 key events by type. Your call tracking platform’s qualified call count. Your case management or intake system’s count of new inquiries logged. They will not match. The goal is not matching, it is knowing the size and direction of the gap so you can apply it consistently.
Expect GA4 to undercount. Typical causes are consent banners suppressing tags for a share of visitors, ad blockers and iOS privacy features, and calls placed from a number the visitor saved months ago with no session attached. Expect intake to overcount web attribution, because referral calls and repeat clients get logged under whatever source the receptionist guessed. On a healthy setup, GA4 form events should land within roughly ten percent of the forms in your case management system, since both are server confirmed. Calls are the messy side and can diverge by 30 percent or more without anything being broken.
Close the call gap properly by using dynamic number insertion so the displayed number swaps by traffic source, and by importing the call events back into GA4 rather than reading them only in the call platform. That, plus a response time standard on the intake side, is the difference between analytics that describe traffic and analytics that describe cases. The mechanics are laid out in our piece on call tracking and speed to lead for law firms.
One caution on dynamic number insertion. Keep the number in your Google Business Profile and your core citations fixed as the real office line and swap only the on site display number, or you will create a NAP inconsistency that costs more in local visibility than the tracking is worth.
Set It Up Once, Then Stop Fiddling
The configuration described here takes a competent implementer somewhere between four and eight hours, most of it in tag setup and testing rather than in GA4 itself. After that the maintenance is a monthly reconciliation and a quarterly check that nothing broke during a site update, which is the single most common way a firm loses months of conversion data without noticing. Test your key events in DebugView after any theme change, plugin update, or form migration. No exceptions.
None of this makes the phone ring. It makes the firm able to tell which of the things it is already paying for makes the phone ring, which is a prerequisite for spending the next dollar sensibly. Rubiks configures GA4 for law firms as part of every engagement, events, custom dimensions, call import, and the monthly reconciliation against intake. Talk to Rubiks and we will audit what your property is currently counting and tell you which numbers you can trust before you make another budget decision.