Google Analytics 4 (GA4)

We agree with you what counts as a conversion, then set up Google Tag Manager and GA4, plus server-side tracking through Stape for sites that spend on ads. Before we hand over, we check the data is actually arriving.

Without it you're guessing which ads, pages and emails bring customers. Ad platforms also bid on what you tell them, so missing or wrong conversions mean they find the wrong people.

We look at your ads, website and CRM together, then make the biggest fix first. That might be a test, a creative refresh, a landing page change or a tracking repair.

A fixed-fee retainer is the same monthly cost for an agreed scope. A performance-based one adds a share of results to a base fee. Fixed is the usual starting point, and we agree the fee model before we start.

It cuts the conversions lost to ad blockers and browser limits, but it doesn't remove the loss, and we don't promise a percentage. With server-side tracking, events go to a server container that sends them on to GA4, Google Ads and Meta, and each conversion carries one event ID so the platforms count it once. On our own site we load the tags from our own domain through Stape's Custom Loader, and the gain is resilience to ad blockers.

A marketing funnel is the route someone takes from first seeing you to becoming a customer: an ad, a landing page, an enquiry, a follow-up, a sale. It matters because you can only fix where people drop out if you can see each step.

An essential funnel covers the basics: a foundational website, essential tracking and one way to capture leads. Advanced adds multi-platform ads, automated nurturing and deeper analytics. Enterprise adds segmentation, a customer data warehouse and deep CRM integration.

We start from the action that counts, usually a form, a booking or a purchase. Your site pushes a named event to the data layer when it happens, a trigger in Tag Manager listens for that event, and the tags it fires send the conversion to GA4, Google Ads and Meta with the same event ID. We test it in a set order, and we don't treat Preview alone as proof.

GA4 links to Google Ads and BigQuery from its Admin screen, and Tag Manager sends it the events. On sites that spend on ads a server container feeds it, and the raw events go to BigQuery, where we can join them to your CRM. Looker Studio reads it for the reports.

It's part of our technical tracking service.

Mistakes we check for:

• Conversions that don't match the business. Tag what counts as a customer, not what's easy to tag.
• A click ID overwritten. A misconfigured server-side power-up can write the word "stape" where Google's click ID should be. Preview still looks fine, so we check the next day's data.
• One property, several sites. Filter reports by hostname, or the numbers blend.
• No ID on the form. Without a first-party ID in the form, a lead can't be tied to its visit.
• An export that quietly stopped. BigQuery gives no error, so we check the newest table's date.
• Personal data in GA4. Where data is sensitive, only an ID goes in, never names, emails or phone numbers.

We treat predicted and modelled figures as estimates, and check the report against the raw BigQuery export, which has no modelling. GA4's predictive metrics only switch on when a property has enough data (purchase and churn predictions each need at least 1,000 returning users who triggered the condition and 1,000 who didn't), so many small sites never get there.

GA4 also fills some gaps with modelled data when visitors refuse cookies, and that has its own thresholds.

GA4 supports Google's Consent Mode, so what it collects can follow the visitor's cookie choice. When someone refuses, GA4 can model some of what it missed if the property has enough traffic, but modelled numbers are estimates and refused consent still means less data.

We connect your consent tool to your tags and test it, and where data is sensitive only an ID and a click ID go into GA4, never names, emails or phone numbers. GA4 doesn't make a site compliant by itself.

We start with what the business counts as a customer: a booked call, a paid order, a qualified lead. Then we build that as an event in Tag Manager, name it plainly, test it in preview and mark it as a key event in GA4 (Google's current name for a conversion).

The next day we check the data against the raw BigQuery export, and send the same definition to the ad platforms. It's part of our technical tracking service.

Universal Analytics counted sessions and page views. GA4 counts events, so a form submit or a purchase is recorded the same way as a page view. Standard Universal Analytics properties stopped processing data on 1 July 2023, and Google removed access to its data in July 2024, so there's nothing to upgrade from now: you need GA4 in place.

Event-level data in a standard GA4 property is kept for 2 or 14 months, so we turn on the BigQuery export to keep a raw copy. It's part of our technical tracking service.

They share data once someone links them: Google Ads and GA4, Search Console and GA4, Tag Manager for the tags, Looker Studio for the reports. We use them together by sending the same definition of a customer to GA4 and Google Ads, and by building reports from GA4, Google Ads and BigQuery.

Review it every quarter, and rewrite it properly once a year or when something big changes in your business or market. A review is a short check against your goals, not a new plan.

Once your ads, tracking, website and reports are live, we look after them: we check the numbers against the raw data, fix what slips, and pick the biggest improvement to make next. Each report ends with a decision to push, fix or stop.

We track organic traffic, where your pages show for the searches that matter, engagement, and the leads and conversions that started with a search. You get regular reports on what's working, and we focus on real business results, not vanity numbers.

We track how quickly leads get a reply, how many move from one stage to the next, and email opens and clicks. We also carry the ad or page each lead came from into the CRM, so you can see which sources bring customers.

Webflow for sites where it suits, Google Ads and Meta for paid, Google Tag Manager, Stape and GA4 for tracking, Looker Studio for reports, and a CRM or email tool that fits your business, including our own CRM built on Twenty. n8n joins them up. The mix depends on what you sell.

We find the stage losing the most people and fix that first. It might be the ad, the landing page, the form or the follow-up emails, and we test one change at a time so we know what worked.

We map your customer journey first, then build each stage: the ads, landing pages, forms, tracking and follow-up emails. We check that tracking and forms work before launch, then keep adjusting once it's live.

If enquiries come in but too few turn into customers, or you can't say where people drop out, funnel management is worth a look. We look at the whole path, from the ad to the follow-up, and fix the weakest step first.

We measure against leads and sales in your own records, not just what the ad platform reports. Cost per lead, cost per customer and return on ad spend come from there.

Want to run your own ROAS or cost per customer against a margin? Try the spend check.