How to Track Where Your Leads Come From, Properly
How to track where your leads come from: one tracker, and a form that carries the source into the lead record. Test it with a real person before you trust it.
To track where your leads come from, put one tracker on your site and make every form attach the visitor's source to the person it creates. Then follow one real person from first visit to payment. If the source doesn't arrive with the lead, nothing downstream works.
How to track where your leads come from: it all happens at the form
You can have a traffic report and a customer list and still not know where your leads come from, not even for a single customer.
Why? Because the two never met. The visits sit in one building. The money sits in another. Nothing carries the story from one to the other.
The place where it breaks is almost always the form. Someone fills in a name and an email, and the form passes along exactly that. The source, the pages they read, the link they clicked, all left behind. You just severed the connection between everything that person saw and everything they later buy. Nothing downstream can reconnect it.
Every form on your site is one of two things. An identity handoff that keeps the history, or a place where your attribution dies. There is no third option.
The minimum setup, in the order to do it
Forget the imagined project. The warehouse, the consultant, the quarter of evenings. At your scale that's actively harmful. It delays the one thing that matters, which is having any light on at all.
The honest minimum fits in an evening.
| Step | What you do | How it breaks |
|---|---|---|
| One tracker | A single script on every page you already run | Pages you forgot exist have no script |
| Carry the source | Campaign details persist past the landing page | Source lost on the second click |
| Form handoff | The form attaches where your leads come from to the new person | Name and email arrive, source does not |
| Outcome in one place | The sale is recorded against the same person | Revenue and traffic never get related |
| Self-report | One optional "how did you hear about us?" field | Left off, so the invisible half stays invisible |
Five rows. That's the whole thing.
Free guide · 26 pagesStop paying for what does not work.The Attribution Playbook: how to find which marketing produces money.Get the free guide →Choose the tracker carefully
You only need one. Insist on three properties.
It must be first-party. The data goes to your own system, not locked inside a platform that shows it back to you only in aggregate.
It must persist the original campaign details for the whole session. Not just the landing page. Otherwise the source vanishes the moment someone clicks through.
And it must record custom events, not only page views. In most funnels the interesting moments are actions, not pages.
This doesn't mean rebuilding your site or moving your hosting. It's a tag on pages you already run.
Keep the outcome next to the identity
Once someone is known, their journey continues. They subscribe, they pay, they renew, or they leave.
For attribution to work, the record holding that outcome has to be the same record holding the touchpoint history. Or joined to it by a stable key.
When identity and outcome live in two systems that were never introduced, you get the classic result. Perfect traffic reporting. Perfect revenue reporting. No way to relate them.
Record the value too, not just the fact of a sale. And know how a lost or cancelled customer is marked, so your rates aren't quietly inflated.
Test it with a real person
Don't trust the setup because the script loaded. Prove it with your own hands.
- Arrive at your site from a tagged link.
- Click through to a second and third page.
- Submit a real test form.
- Open the new person record. Is the source there?
- Repeat for every form on the site, not just the main one.
- Follow one real customer end to end: first visit, form, payment.
Also try it from a device and network that aren't yours. And find out what happens when someone signs up from a second device, because that's where the history can get lost.
Where this goes wrong is rarely the first form. It's the one you forgot about.
Check your funnel counts it properly
If you chart steps, there's a trap worth knowing. A real funnel counts, at each step, only the people who completed every previous step. Not everyone who happened to do that action.
Independent counts produce a chart shaped like a funnel that means nothing. The people at step one and the people at step three may be different people. If your tool does independent counts, your drop-off percentages are fiction and you'll improve the wrong step. Verify before you act on any of it.
Consent, honestly
Depending on where your visitors are, you'll need consent before some tracking runs, and a meaningful share will say no.
Design for that. Get the consent handling right, prefer first-party signals that survive browser limits, and accept that your numbers describe a large sample, not everyone.
That is plenty for the job. Comparing channels against each other works fine on a representative sample. Reconciling to your accounts doesn't, and you shouldn't try.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →Then stop
This is where most people overbuild. A second analytics tool for comparison. A heatmap suite. A session recorder. A warehouse.
Don't. Two tools measuring the same thing differently don't give you confidence. They give you a standing argument about which is right, and you're both sides of it.
Get one honest pipeline running end to end. Look at it for a full sales cycle. Add something only when you have a specific question the current setup can't answer.
Once leads carry their source, you can start reading what they say. That begins with what marketing attribution actually is, and the review habit that turns it into decisions is in how to review marketing results every month.
Where your leads come from starts before the form
Before the form there's the visit. Before the visit there's the link someone clicked. If that link carried no tag, the visit arrives as direct traffic and the story starts with a blank.
That's why tracking and tagging are one job, not two. Tag everything you control, once, and keep tagging. Anything you place in a post, an email, a community reply or a partner page needs it. The naming rules are in how to set up UTM tags for a small business.
Then the tracker holds the details through the session. The form attaches them to the person. The payment lands on the same record. Four links in a chain. If any one is weak, the chain is weak, and you won't see which one from the outside.
What you can read once it works
When it works, a lead record tells you where the person first arrived, which path they took, and what they eventually paid. That's what lets you ask the three questions of every channel: how many, at what rate, at what value.
It also lets you do something most solo operators never get to. You can look at the sources that produce customers who stay, not just customers who arrive. The cheapest channel to acquire from is often not the cheapest to serve.
And you can answer the awkward question honestly. Which of the things you do on a Tuesday evening is earning its place?
A caution to finish. Your numbers will describe a sample, and some of your best marketing happens where no tracker follows. That's fine. Treat it as a compass. Judge it as a ledger and it fails, and you'll go back to guessing, which was worse.
Free guide · 26 pagesStop paying for what does not work.The Attribution Playbook: how to find which marketing produces money.Get the free guide →The questions to ask about your own setup
Before you build anything, answer these honestly.
- Is the script on every page, including the ones I forget exist?
- Does the source survive past the first page?
- Does every form pass it on, or only the main one?
- Can I follow one real person from first visit to payment?
- Would I know if one of these quietly broke?
If any answer is "I'm not sure", that's your next task. Not a new tool. A test of where your leads come from.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →Frequently asked questions
Why does my lead list show no source?
Most likely your form submits only a name and an email and drops the history the tracker collected. Submit a test form from a tagged link and check the new person record. That tells you where the link breaks.
Do I need more than one tool?
No. One tracker that is first-party, keeps campaign details through the session and records custom events is enough. A second tool measuring the same thing mostly creates arguments.
What if visitors decline tracking?
Some will. Your numbers then describe a large sample rather than everyone. That is sufficient for comparing channels against each other, though not for matching your accounts to the cent.
How do I catch leads that never click a tagged link?
Add one optional question to your main form asking how they heard about you, with free text allowed. It catches the invisible half better than any tracker, as long as you read it beside tracked data.
The full method, with the checklists, is in The Attribution Playbook, a free 26-page playbook.
How trapped are you, really?
Take the Corporate Suffocation Index, the free solopreneur readiness assessment.