The number
$1,094 a month of ad spend leaking, found in a line-by-line account audit
Service · Conversion tracking
Conversion tracking that matches what you sold.
Conversion tracking is the number every ad dollar is judged by, and in the accounts on this page it was wrong. We set up GA4 and Google Tag Manager, cut Google Ads conversions to the real outcomes, and carry every click ID across the domain hop so the purchase is attributed where it belongs.
A written audit. No pitch unless you ask for one.
0 → 5
key conversion events live and linked to Google Ads, on a launch that began with an empty tag container
Venture we operate
receipt
SOURCE · Analytics property change log for a marketplace we operate
METHOD · Key events and the Google Ads link created in the analytics property, read from the property’s change log.
VERIFIED · 2026-09
CLASS · platform account
✓ every number on this site opens its receipt
CLAIM · key-events-zero
How we count
✓ Server-recorded
events the API writes, so they count with no browser open
events the API writes, so they count with no browser open
Click IDs carried
gclid, fbclid and the rest survive the hop to the store
gclid, fbclid and the rest survive the hop to the store
01 · What goes wrong
The dashboard is confident. The order history is not.
What businesses bring us with their tracking, in their words. Each one makes every number downstream wrong.
- Unmeasured
GA4 shows no key events, or the wrong ones.
The property was created, the tag was pasted, and nobody wrote a tracking plan. Page views flow in and the outcomes that matter never get named, so Google Ads has nothing real to optimize toward.
- Overcounted
Google Ads counts clicks and views as sales.
Several conversion actions marked primary, some of them engagement events, one a duplicate purchase tag. Every campaign looks profitable because the account is grading itself.
- Lost attribution
The click ID dies at the domain hop.
A landing page on one domain, a store on another, and the gclid dropped between them. The ad gets no credit for the purchase, and the next decision is made without it.
- Browser-only
Browsers drop the events.
Consent banners, ad blockers and short cookie lifetimes mean a growing share of conversions never leave the browser. The platform fills the gap with modeled numbers you cannot check.
- Offline
The real sale happens offline, and Google never hears about it.
A lead becomes a customer on a phone call or a sales floor weeks later. Without an offline conversion import, the campaign that produced it is judged on the form fill alone.
- Untagged
Every person tags links their own way.
Three spellings of the same source, campaigns with no medium, and print and radio with no link at all. The attribution report is a list of typos.
02 · What we do
A tracking plan, then the tags, then the proof that they match the orders.
Every piece is written down before it is built, and checked against the system of record after.
01GA4 + GTM
GA4 and Google Tag Manager set up from a written tracking plan
The outcomes that matter named as key events with their parameters, the container organized so the next person can read it, and consent handled in the tags rather than around them.
02Google Ads
Google Ads conversions cut to the real outcomes
One primary conversion action per real outcome, engagement and duplicate actions removed or demoted, values attached, and the Google Ads link to the analytics property confirmed.
Google Ads management03Server-side
Server-side tagging where browsers lose events
A server container on your own cloud, events recorded by your API when the outcome actually happens, and enhanced conversions so the match holds when the browser cannot.
04Click IDs
Click IDs that survive the domain hop
gclid, wbraid, gbraid, fbclid, msclkid and ttclid carried from the landing page into the store and onto the order, so the purchase is attributed to the click that caused it.
05Offline
Offline conversion imports from your CRM
Sales that close on a call or a sales floor sent back to Google Ads against the original click, so campaigns are judged on customers rather than form fills.
06UTMs
One documented UTM matrix for every channel
Pre-built tracked links for every channel, online and offline, with one spelling per source and medium, so the attribution report reads as one list instead of a dozen.
We also set up Meta Pixel and Conversions API events on the same plan, so both platforms are judged against the same store orders.
03 · Proof you can check
Tracking we rebuilt, and tracking we set up from an empty container.
A DTC brand’s measurement rebuilt after it counted video views as purchases, a dealership’s campaigns tagged across every channel, and the launch we run ourselves, labeled as ours. Clients whose sign-off is pending are described, not named. Every figure opens its receipt.
0 → 5
key conversion events live and linked to Google Ads, on a launch that began with an empty tag container
Venture we operate
receipt
SOURCE · Analytics property change log for a marketplace we operate
METHOD · Key events and the Google Ads link created in the analytics property, read from the property’s change log.
VERIFIED · 2026-09
CLASS · platform account
✓ every number on this site opens its receipt
CLAIM · key-events-zero
How we count4 → 1
actions Google Ads was counting as a sale, cut to the one real store purchase
Two were video-engagement actions and one was a duplicate purchase tag that could never fire.
receipt
SOURCE · Google Ads conversion settings for a DTC pet-food brand, before and after (account change log)
METHOD · Primary conversion actions in the account before and after the change, confirmed by re-reading the account’s conversion settings.
VERIFIED · 2026-09
CLASS · platform account
✓ every number on this site opens its receipt
CLAIM · conversion-actions
How we count12
tracking parameters carried across the landing-to-store hop
Every ad click ID (gclid, wbraid, gbraid, fbclid, msclkid, ttclid) plus UTMs, passed into store deep links so attribution survives the domain change.
receipt
SOURCE · Landing-system repository for a DTC pet-food brand
METHOD · Parameter passthrough list in the landing system’s link builder.
VERIFIED · 2026-09
CLASS · repository
✓ every number on this site opens its receipt
CLAIM · tracking-params
How we count11
channels in one documented UTM matrix, every link in English and Spanish
Google Search and Display, Facebook, Instagram, TikTok, Yahoo, Bing, email, QR codes, print and radio.
receipt
SOURCE · Campaign repository for a new-car dealership (documented UTM link matrix)
METHOD · Count of distinct channels with pre-built tracked links in the campaign’s written attribution plan.
VERIFIED · 2026-09
CLASS · repository
✓ every number on this site opens its receipt
CLAIM · utm-matrix
How we countThe work behind the numbers
The number
281,949 veterinary board records across 30 states and DC behind the clinician verification gate
- ClientA DTC pet-food brandGoogle Tag Manager and GA4 rebuilt, Google Ads conversions cut to the one real store purchase, Meta Pixel events corrected, and a landing system that carries every click ID into the store.
- ClientA California new-car dealershipSeasonal sales-event campaigns with GA4 lead events carrying the shopper’s language, the Google Ads conversion on a dedicated thank-you route, and one documented link matrix spanning search, social, email, QR, print and radio.
- Venture we operateTriaPetA written tracking plan before the first campaign, key events recorded by the server with identity stitched from first visit through triage to signup, and the Google Ads link confirmed in the property.
04 · How it starts
Audit what fires today. Write the plan. Build it. Prove it against orders.
The same order every time. You see what the current tags actually count before anything changes.
Every engagement starts at step one, because the tracking in the accounts on this page was broken when we got it. We would rather show you an unflattering number than a pretty dashboard.
Step 1: The free Google Ads and tracking audit
Read-only access to the analytics property, the tag container, the ad accounts and the store. We list every conversion action and key event, what each one actually counts, and where the order history disagrees.
Out a written audit, yours to keep
Step 2: A written plan
If there is a fit, you get a tracking plan and a written plan with the scope, a timeline and an estimate: which events, which parameters, which platforms, in what order. If there is not a fit, we say so.
Out a tracking plan, plus scope, timeline and an estimate
Step 3: Tags built and tested
GA4 key events, Google Ads conversions, server-side tagging, click-ID passthrough and the UTM matrix, built in a container you own and tested on staging before they fire in production.
Out a container the next person can read
Step 4: Reconciled against the system of record
Platform conversions compared with the store’s or CRM’s own records over the same window, and the gaps explained in writing. Absent data is labeled, never rendered as zero.
Out platform numbers that match the orders
Step 5: Handed over, documented
The tracking plan, the container and the UTM matrix documented for your team, with offline imports running on a schedule, so the next campaign is judged on the same honest count.
Out documentation your team can run with
05 · Straight answers
Questions people ask before letting anyone touch their tags.
Do you need admin access to GA4 and Google Tag Manager?
Not to start. The audit runs on read-only access to the analytics property, the tag container, the ad accounts and the store. Changes come after you have read the audit and agreed to the plan, and they go into containers and properties that stay in your name.
What is server-side tagging, and do I need it?
Instead of every tag firing from the visitor’s browser, events go to a server container you own, which forwards them to Google and Meta. It holds up when consent banners, ad blockers and short cookie lifetimes stop the browser from reporting. Whether you need it depends on how much of your traffic those affect, and the audit measures that gap before recommending it.
Our landing pages and store are on different domains. Can attribution survive that?
Yes, if the click IDs and UTMs are carried across the hop on purpose. We pass gclid, wbraid, gbraid, fbclid, msclkid and ttclid plus the UTMs into the store’s links and onto the order, and set up cross-domain measurement in GA4, so the purchase lands on the click that caused it.
Our sales close on the phone. How does Google find out?
Through offline conversion imports. The click ID is stored with the lead in your CRM, and when the lead becomes a customer the sale is sent back to Google Ads against that click, on a schedule. Campaigns are then judged on customers, not on form fills.
Why do the platform numbers and my order history disagree?
Usually several reasons at once: engagement actions counted as conversions, duplicate tags, each platform claiming the same sale, and modeled conversions filling gaps the browser left. We reconcile platform reports against the store or CRM over the same window and explain each gap in writing. The honest count is the store’s.
What does this cost?
The audit is free and ends with a written plan that carries the scope, a timeline and an estimate for your setup. We do not publish prices because an empty property and a container with years of tags are very different jobs. You see the plan before you commit to anything.
06 · The front door
Find out what your tracking actually counts.
Bring read-only access to the analytics property, the tag container and the ad accounts. We check every conversion action and key event against the orders you recorded, and hand you the written audit either way.
A written audit. No pitch unless you ask for one.