You Are Making Every Marketing Decision On Numbers You Have Never Checked

Every sale traced back to what actually created it

A client books a call. Where did they come from? The honest answer in most businesses is a shrug, or a guess, or whatever the last touch happened to be.

They found you through a podcast, read three emails, saw a LinkedIn post, searched your name, landed on a page, and booked. Your reporting recorded the last step and lost the rest. So the podcast gets no credit, the emails get no credit, and next quarter the budget goes somewhere that looked better on a dashboard.

I build the tracking that captures where a person actually came from and keeps that source attached to them through registration, purchase and everything after.

How do I know if my tracking is broken?

Six symptoms. If two of these are familiar, the numbers you are managing on are wrong.

A large share of your traffic is reported as direct, which usually means the real source was lost in transit
Your platforms disagree. The CRM, the analytics and the payment processor each report a different number for the same month
Someone buys and you cannot say which post, email, referral or campaign they first arrived through
A sale that closes eight weeks after first contact is invisible, because the attribution window closed and nothing else was recording
Your CRM has contacts with an empty source field, and nobody knows when that started or how many
You know your organic content brings people in and you have never been able to put a number on it

None of these announce themselves. Tracking does not error, it just quietly reports something plausible.

Two of those and your reporting is fiction with good formatting. The audit tells you how far off it is.

Your systems already know the answer, your reports do not

The data is rarely missing. It is sitting in systems that were never introduced to each other.

Your webinar platform knows who attended and for how long. Your CRM knows which contact became a customer. Your payment processor knows what was actually kept after refunds. Your email tool knows who clicked before any of that happened. Each one holds a piece of the same story and none of them can see the others.

So reporting defaults to whichever system speaks loudest, which is usually whichever one has a dashboard already open. The picture is not wrong because the data does not exist. It is wrong because nothing joins it up.

The build is mostly plumbing rather than analytics. Capture the source once, carry it through every system the person touches, and reconcile what those systems report against what actually happened. Once that chain holds, the reporting question stops being difficult.

Who this is for

You are allocating budget or effort across channels and cannot say which ones produce customers
Your platforms each report a different number and nobody has reconciled them
Your sales cycle is longer than any attribution window, so late conversions vanish
You put real work into content, email or organic search and have no way to value it
You inherited a setup from a previous agency and have no documentation of what fires where

Who this is not for

You have one channel and one offer. If everything arrives from one place, attribution answers a question you do not have. Basic analytics is enough
You want a dashboard rather than accurate data. A dashboard built on broken tracking is a faster route to a confident wrong decision
You want it set up and never reviewed. Platforms change their requirements, and a setup nobody checks drifts out of accuracy within a year
You want tracking that ignores consent requirements. Consent handling is part of the build, not an obstacle to route around

Sound like you? Say it upfront. There is no point building measurement for decisions you are not making.

What is included in a tracking build?
Audit of what existsEvery tag, event and integration currently firing, what it reports, and where it disagrees with your other systems. Most builds start by discovering that two things are double counting and a third stopped working months ago.
Source capture at first touchWhere somebody came from gets recorded the moment they arrive, from any channel. Campaign parameters, referrers, organic search, and direct visits with a known prior touch. Not just the last click before the form.
Source persistence through the journeyThat source stays attached to the person through opt in, registration, purchase and into the CRM record. A sale months later still points back to where it began.
Event and conversion mappingWhat counts as a conversion at each stage, named consistently across every platform, so the same event means the same thing everywhere. Tracking everything produces noise rather than clarity.
Server side and browser recordingConversions recorded from your own systems as well as from the browser, so events blocked by ad blockers or privacy settings still get counted.
CRM and payment reconciliationConnecting what your CRM and payment processor know back to your reporting, so recorded conversions match actual sales rather than approximating them.
DocumentationA written map of every event, where it fires, what it reports to, and how to verify it. This is what keeps the setup maintainable after I leave.

What gets tracked?

Source attributionFirst touch and last touch both captured and stored on the contact record, so they can be compared rather than trusting whichever your platform defaults to.
Funnel step eventsPage views, opt ins, registrations, webinar attendance, checkout starts, purchases, upsell acceptance. Named consistently across every platform.
Webinar behaviorRegistration, attendance, watch duration and drop off point, passed into the CRM so follow up and reporting can both use it.
Revenue eventsPurchases, refunds, subscription renewals and failed payments, so reported revenue reflects what was kept rather than what was charged.
Ad platform conversionsMeta pixel and Conversions API, Google Ads conversion tracking and enhanced conversions, for the channels where you are spending.
Consent and privacy handlingConsent mode, cookie banners and regional requirements configured as part of the setup rather than bolted on afterwards.

What has this produced?

How much does attribution setup cost?

From $1,000Tracking auditA full review of what is firing, what is double counting, what is missing, and where your platforms disagree. Delivered as a documented findings list you can act on or hand to your developer.
From $2,000Standard setupSource capture and persistence into the CRM, funnel step events, GA4 configuration, server side recording, and consent handling. One funnel, one primary channel.
From $4,000Full attribution buildMulti channel and multi funnel. First and last touch stored on the contact record, webinar behavior capture, CRM and payment reconciliation, and reporting that reconciles across every system.

Setups involving custom platforms, or untangling a previous agency's undocumented configuration, are quoted separately after the audit.

Terms are 50% upfront, balance on delivery.

The audit is worth doing on its own. It frequently finds that recorded conversions and actual sales are further apart than anyone assumed, which changes what you should be investing in and where.

And if your tracking turns out to be fine, that is a real answer. It means the problem is somewhere else, and you can stop suspecting your data.

Send me one month from your CRM, your payment processor and your analytics. The gap between those three is the starting point.

How long does a tracking setup take?

Add a verification period on top of any build. Tracking cannot be confirmed correct on the day it goes live, because it needs real conversions to flow through before the numbers can be compared against your CRM and payment processor. Two weeks of live data is a reasonable check.

Scope	Timeline
Tracking audit	3 to 5 days
Standard setup	1 to 2 weeks
Full attribution build	3 to 5 weeks

How the project runs

InventoryEverything currently firing gets cataloged, including the tags nobody remembers adding. Duplicates and orphans get identified before anything new goes in.
Decide what countsWhich events matter, what each one means, and how sources get classified. Naming gets standardized here so the same thing means the same thing in every system.
ImplementSource capture, persistence, event tracking, server side recording, CRM connection and consent handling. Built, not just specified.
Verify against realityRecorded numbers get compared against CRM records and payment processor data. The build is not finished when it goes live, it is finished when the numbers agree.

Questions people ask before hiring me

My developer already installed analytics. Is that enough?

It captures visits. Attribution is a different job: keeping a person's original source attached to them through every step until they buy. Standard analytics installations almost never do that, which is why so much traffic ends up filed as direct.

Can you tell me exactly what produced each customer?

Closer than you have now, not perfectly. Real journeys have several touches, so any single answer involves a choice about how credit gets assigned. What you get is first touch and last touch both recorded, so you can compare them rather than trusting a default.

How do I measure organic content, referrals and social posts?

Same mechanism as anything else, provided the source is captured at first touch and carried through. This is usually where the biggest reporting gap sits, because content gets no tracking discipline and then gets blamed for not producing results.

What if my sales cycle is months long?

That is exactly where standard attribution windows fail. Storing source on the contact record rather than relying on a platform window means a sale closing in month four still points back to what started it.

Will this survive the next privacy update?

Server side recording is more durable than browser only, because it does not depend on the browser cooperating. Durable is not permanent. Requirements change, which is why the setup is documented and worth reviewing annually.

Is this compliant under GDPR?

Consent governs what you may collect and pass on, regardless of where it gets sent from. Consent handling is built in rather than worked around. If your current setup is not compliant, that comes up in the audit.

Do you fix tracking on funnels you did not build?

Yes, and it is common. Inherited setups are usually undocumented, which makes the audit stage longer and more useful.

Check Them Before You Spend Against Them

Every marketing decision you make this year runs on these numbers. Worth knowing whether they are real before you make them.