UTM parameters and Stripe: 6 steps to revenue by source

UTM parameters reach Stripe only when your own code carries them there.

The tags sit in the landing URL, a Stripe Checkout Session is created later on the server, and nothing joins the two unless you save the tags on the first visit and write a visitor ID onto the Stripe objects holding the payment.

Payment Links forward 5 UTM codes to your redirect URL after payment, and Stripe's documentation describes no place where Stripe saves the codes.

Without a bridge, Google Analytics 4 holds the campaign, Stripe holds the amount, and a revenue report files the payments as Direct. A bridge is the saved link between a tag and a payment, and building one takes 6 steps.

First, build the bridge in a sandbox. Finally, check 1 real $49 payment from Google Ads, LinkedIn or Pipedrive in 2026.

What happens to UTM parameters in Stripe?

A Payment Link accepts 5 UTM codes of up to 150 characters each, using letters, digits, dashes and _.

With the confirmation behavior set to redirect, Stripe appends the codes once payment completes and silently discards invalid values, per Stripe's Payment Links tracking page. Every other surface takes a field you fill yourself.

A field is a place where Stripe stores a value you wrote, such as metadata or client_reference_id. For example, a Checkout Session holds 50 metadata pairs and a Buy Button holds one 200-character reference.

Which route fits each checkout type?

A route is the way a tag reaches a Stripe object, and the table lists 4 routes for 2026. I start with the visitor ID: 1 value in client_reference_id and metadata, joined to a visit table you control.

Google's Measurement Protocol overview lists the limits of the Google Analytics 4 route: 25 events per request. A Payment Link route drops values with a dot or colon, and a Checkout code route needs 1 webhook handler.

The guide to connecting Google Analytics to Stripe revenue covers the route.

How does a click become a charge in 6 steps?

A click-to-charge chain is the path from a tagged link to a paid invoice, and a gap at any of the 6 steps leaves a $49 payment in Direct.

Each step below fixes 1 of 6 places where a tag can vanish: the link, the first request, the user row, the Stripe objects, the webhook and the sandbox test in Stripe Checkout.

The steps run in order, and a skipped step shows up as a row of unattributed revenue in the report at the end.

Step 1: How do you tag every link the same way?

Consistent tagging is the first link in the chain. Google's URL builder help says to always include utm_source, utm_medium and utm_campaign, and values are case sensitive, so google and Google count as 2 utm_source values.

The free UTM builder assembles the URL, and the UTM parameters guide covers Google's 9 supported parameters. Keep each channel's spelling lowercase, and store the approved values in 1 shared list in 2026.

Google Analytics 4 groups sessions into 19 default channels by the 3 core tags. Finally, reuse the same campaign name on every link of one send.

Step 2: Why capture tags on the server?

A server-set cookie is a cookie written in an HTTP response header. Read the tags from the first request your server receives and set the cookie in the response.

Per WebKit's tracking prevention page, Safari Intelligent Tracking Prevention deletes cookies created in JavaScript after 7 days without user interaction and caps them at 24 hours after link decoration, while a cookie set in an HTTP response escapes both rules unless a CNAME or third-party IP address cloaks the cookie.

Write the cookie only when empty, so a later channel cannot replace the first source. A server cookie can keep a 90-day or 365-day lifetime, where Safari caps a JavaScript cookie at 7 days.

The stripped parameters guide has an Express middleware for this.

Step 3: What do you save at signup?

A user row is the database record created at signup. Write the tags and the visitor ID there, so one row holds utm_source, utm_medium, utm_campaign and 1 ID, 4 values in all.

A buyer who confirms an email on a phone has no cookie there, and the email typed at Stripe Checkout can differ from the signup email, while the saved ID still matches. First, write the row once at signup.

Second, never overwrite the row with later tags in 2026. Stripe Checkout, Stripe Billing and the Stripe Dashboard then show the same visitor ID.

Step 4: What goes onto the Stripe objects?

Stripe's metadata guide lists 4 cases where Stripe copies metadata between objects, and Checkout Session to Subscription is not one, so a subscription needs subscription_data.metadata. Each invoice then holds a snapshot in parent.subscription_details.metadata.

The 3 core tags and a visitor ID use 4 of the 50 slots.

A Checkout Session expires 30 minutes to 24 hours after creation (24 by default), so create the Session when the buyer clicks Subscribe if you store last-touch tags.

The Checkout Session, the Subscription and the first Invoice then carry the same 4 keys.

1// The tags were saved on the user row at signup (step 3)2// Stripe unsets a metadata key whose value is empty, so skip the blanks3const tags = Object.fromEntries(4  Object.entries({5    utm_source: user.utm_source,6    utm_medium: user.utm_medium,7    utm_campaign: user.utm_campaign,8  })9    .filter(([, value]) => value)10    .map(([key, value]) => [key, String(value).trim().toLowerCase().slice(0, 500)]),11);12 13const session = await stripe.checkout.sessions.create({14  mode: "subscription",15  customer: user.stripeCustomerId,16  line_items: [{ price: priceId, quantity: 1 }],17  client_reference_id: user.visitorId,            // up to 200 characters18  metadata: { vid: user.visitorId },              // on the Checkout Session19  subscription_data: {20    metadata: { vid: user.visitorId, ...tags },   // on the Subscription, then each invoice21  },22  success_url: "https://example.com/welcome",23});

The Stripe metadata attribution setup post has the full key map, and the Payment Links attribution guide covers the per-click ID.

Step 5: Which webhooks return the payment?

Stripe events are the webhook messages announcing a payment. Handle checkout.session.completed for the first order and invoice.paid for renewals. Stripe sends a separate event when a delayed payment method such as ACH Direct Debit succeeds later, so handle checkout.session.async_payment_succeeded too.

Key each row by the object's ID, such as in_1 or cs_1, so a repeated delivery overwrites instead of adding.

For example, an invoice.paid event for month 2 carries the subscription of month 1, so the renewal inherits the first click in 2026. Stripe Webhooks retry failed deliveries, so 1 event can arrive 2 or more times.

Step 6: How do you test in a sandbox?

A Stripe sandbox is a test mode where no real money moves. In a Stripe sandbox, pay with the Visa test card 4242 4242 4242 4242, any future expiry and any 3-digit CVC, per Stripe's testing guide.

For a user with all 3 tags, the Checkout Session should then show 1 metadata key, the Subscription 4, and the first invoice's snapshot the same 4.

Fix any gap before launch in 2026, using test card 4242 4242 4242 4242 and 1 Stripe sandbox.

What does the report look like?

A revenue-by-source report is a table joining each payment to its stored tags. The hypothetical month below has one $49 monthly plan and 45 paying customers, each joined to a stored source.

Three sources have clicks and hold 35 of the 45 customers, or 77.8%, and $1,715 of the $2,205 MRR. One row holds the other 10 customers with no stored source.

Ranked by clicks, google leads with 2,000 and the newsletter has the fewest at 400.

Ranked by MRR per click the order changes: the newsletter moves from last to first at $1.47 a click, against $0.44 for google and $0.27 for linkedin.

The last row is the one to shrink, since steps 2 and 3 keep a buyer's source from going missing. Each later invoice takes the source of its subscription, so month 12 still credits the click that started it.

First click, last click or an even split is a reporting rule you can change without touching Stripe.

1select v.utm_source,2       v.utm_medium,3       count(distinct p.subscription_id) as customers,4       sum(p.amount_cents) / 100.0       as revenue5from payments p6join visitors v on v.vid = p.vid7where p.status = 'paid'8group by v.utm_source, v.utm_medium9order by revenue desc;

Does the same pattern work in Pipedrive?

Pipedrive is a CRM storing deals, leads and contacts, and in Pipedrive the tags live in custom fields covering deals, leads, contacts, projects and products, per its custom fields article.

In the deal fields API a varchar field holds 255 characters, enough for any UTM value.

Create 1 field per tag plus 1 for the visitor ID, 4 fields in all, and fill them from the step 3 values at signup. The same ID in Stripe's client_reference_id lets a payment find its deal.

For example, a $49 payment finds its deal by the shared ID in 2026.

Where does the chain leak?

What does TrackRev do with the tags?

We built TrackRev, SaaS affiliate software with link tracking and revenue attribution built in. A TrackRev link carries a channel and saved UTM tags.

In our product, the redirect logs the click on the server, appends the tags to the destination, and passes a _vid parameter the pixel stores as a vid cookie on your domain.

On Payment Links, Buy Buttons and pricing tables the pixel adds the ID as the client reference, while a Session you create in code needs the cookie value passed in client_reference_id.

We built the hourly sync to read Stripe with a read-only restricted key, and we check 4 places in order, per the table. First, the charge metadata. Last, the buyer's email in 2026.

The sync covers 6 billing systems: Stripe, Paddle Billing, Polar, Lemon Squeezy, Creem and Dodo Payments.

When does a tool earn its place?

A tool earns its place when Stripe is one of several inputs. Renewals inherit the first order's visitor through the subscription ID, and setup is on the Stripe integration page.

In our product, the pixel does not read UTM tags from a page URL, so hand-tagged traffic with no TrackRev link in the path still needs steps 2 to 5.

If one server-created Checkout covers all your sales, those steps cost no subscription; a tool earns its place with Payment Links, several partners and channels, or more than 1 billing provider.

The Free plan covers link tracking only, with 50 links and 1,000 tracked events a month, and revenue figures stay hidden. Revenue attribution needs a paid plan, from Indie at $29 a month (pricing).