MetricRaven

Attribution

What is marketing attribution?

Marketing attribution is the practice of crediting sales and signups to the marketing channels, campaigns and touchpoints that led to them. It tells you which marketing brings paying customers, so you know where to spend more and what to cut.

Updated

What marketing attribution means

Every customer arrives somehow. They clicked an ad, found a blog post on Google, read a newsletter, or typed your URL after a friend mentioned you. Marketing attribution connects each sale to those touchpoints and adds up the result per channel and campaign.

The output is a table you can act on: search brought $12,000 this month, the newsletter $4,000, paid social $2,500. Without it, you know your total revenue and your total marketing spend, but not which part of the spend earned which part of the revenue.

Attribution has three ingredients:

  • Touchpoints: the tracked visits a person made before buying, each with a source. Sources come from UTM parameters, click IDs and the referrer. A visit with none of these is direct traffic.
  • Conversions: the outcomes you care about. For most businesses that should be a payment, not a pageview or a form submission.
  • An attribution model: the rule that splits credit for each conversion between the touchpoints, within an attribution window.

Attribution models at a glance

The model decides who gets credit when a customer touched more than one channel. The common ones:

ModelCredit ruleBest for
First touch100% to the first touchpointFinding what brings new people in
Last touch100% to the last touchpoint before the saleFinding what closes
LinearEqual share to every touchpointLong journeys where every step matters
Time decayMore credit the closer a touch is to the saleShort sales cycles and promotions
Position based40% first, 40% last, 20% split across the middleValuing both discovery and closing
Data-drivenCredit from a statistical model of your conversion pathsLarge accounts with lots of conversions

First touch and last touch are single-touch models. The others are multi-touch attribution models. See first-touch vs last-touch attribution for when each single-touch model fits.

Worked example

A customer finds you through a Google search, comes back a week later from your newsletter, then clicks a retargeting ad and buys a $600 annual plan. Here is how much each channel gets under each model (time decay uses a 7-day half-life, with the visits 14, 7 and 0 days before the sale):

TouchpointFirst touchLast touchLinearPosition basedTime decay
Organic search$600$0$200$240$86
Newsletter$0$0$200$120$171
Retargeting ad$0$600$200$240$343

Same customer, same $600, five answers. Under first touch, search looks like the hero and the ad looks wasted. Under last touch, the ad gets everything. Neither is wrong; they answer different questions. Pick one model for budget decisions and keep it, so month-to-month comparisons mean something.

Why attribution should start from revenue

Many tools attribute signups, leads or "conversions" defined inside an ad platform. These are easy to count, but they don't pay the bills. A campaign that brings 300 free signups and no customers is worse than one that brings 30 signups and 10 customers.

Revenue-based attribution connects each payment from Stripe, Paddle, Shopify or another provider to the visits of the person who paid. It lets you compare channels on revenue per visitor, ROAS and customer acquisition cost, which are the numbers that decide budgets. A revenue attribution model is simply an attribution model applied to payments instead of proxy events.

How to measure marketing attribution

  1. Tag your links. Add UTM parameters to every link you control: ads, emails, social posts, partner links. Keep values lowercase and consistent.
  2. Track visits on every page, including landing pages on subdomains and checkout, so the journey isn't cut in two.
  3. Link payments to visitors. Pass the visitor ID into checkout, identify users when they sign up, or match by email after purchase. This is the step most setups skip, and without it there is nothing to attribute.
  4. Choose a model and window before you look at the numbers, and write them down.
  5. Read revenue per channel and campaign, then divide by spend where there is spend.

What attribution can't tell you

Attribution only sees tracked touchpoints. A podcast mention, a conference talk or a link pasted into a private chat (dark social) usually shows up as direct traffic or not at all. Ad views without a click are invisible to your site. Customers who switch devices or clear storage look like new people unless something links them.

Attribution also shows correlation, not cause. A brand search ad gets credit for buyers who were going to buy anyway. To learn whether a channel truly adds sales, pause it for a while or run a holdout test, and watch total revenue.

Common marketing attribution mistakes

  • Attributing proxy events instead of payments. Leads and add-to-carts reward campaigns that never pay.
  • Adding up ad platform reports. Each network counts the same sale under its own rules, so the sum is more than you earned.
  • Switching models every month. Trends become impossible to read.
  • Untagged links. Email and partner traffic without UTM tags lands in Direct or Referral and looks worthless.
  • Ignoring unattributed revenue. Payments you can't connect to a visit are a measure of your tracking gaps. Watch that share and work to shrink it.

How MetricRaven handles marketing attribution

MetricRaven attributes payments, not proxy events. Connect a payment provider and every payment is credited to the customer's first tracked visit: the earliest pageview across every visitor ID linked to that customer. That is first-touch attribution, applied the same way in every report, so channels, campaigns, referrers, landing pages and countries never double count a sale.

  • Visits are linked to payments when checkout passes the visitor ID (Stripe Checkout), when you identify users at signup, or when you attribute by email. Payments with no tracked visit appear as Unattributed.
  • Each visit gets one channel (Paid, Email, Affiliate, AI, Social, Organic Search, Referral, Campaign or Direct) from its UTM tags, click ID and referrer. The rules are in Campaign & referral links.
  • With Meta Ads or Google Ads connected, the ad report adds assisted revenue: payments from customers who clicked an ad but first arrived some other way. It shows what a campaign helped close without double counting.
  • Paths to payment shows the pages buyers went through across all their visits, and the Pages card can credit revenue to every page viewed before paying.

MetricRaven doesn't offer linear, time-decay, position-based or data-driven models. It keeps one model and shows the touches around it, so you can see the whole path without the numbers changing with a setting.

Questions

What is marketing attribution in simple terms?
It is working out which marketing led to each sale. You track where every visitor came from, connect their payment to those visits, and add up revenue by channel and campaign.
What is the best marketing attribution model?
There is no single best one. First touch shows what brings new customers, last touch shows what closes, and multi-touch models spread credit across the journey. Pick the one that matches your question and use it consistently.
Why don't my ad platforms and my analytics agree?
Each ad platform credits sales under its own rules, often including people who only saw an ad, and every platform claims the same customer. Analytics that credit each payment once, from your own payment records, report less.
Can I do marketing attribution without cookies?
Yes, within limits. Cookieless tracking still records each visit's source. To connect visits on different days to one buyer, link them at checkout, signup or by email, or use a stored ID after consent.
Full glossary

Attribute your real revenue to every channel

Connect your payment provider and MetricRaven credits each payment to the visit that first brought the customer, by channel, campaign, referrer and page.

14-day free trial, no card required.