Marketing Attribution
Understand What Reports
Can And Cannot Prove

A customer sees a social post, reads an article, asks a colleague about your business and returns a week later through a search ad. The ad platform records a conversion. Your sales team remembers the referral. The website report shows a search visit.

All three observations can describe part of the same purchase. Trouble begins when the business treats each report as a complete explanation and adds the claims together as though they represent separate customers.

Attribution assigns credit to marketing interactions according to a set of rules. It can help organise evidence and compare performance. It does not automatically prove which activity caused a sale or what would have happened without it.

A New Analytics Product Does Not Remove The Need For Definitions

On October 14, 2020, Google introduced the new Google Analytics, building on its App + Web property. The announcement emphasised machine learning and a more complete view across devices and platforms. It also recommended creating a new property alongside existing analytics to start collecting data. Google's October 2020 announcement.

For a business reviewing this change, the first question is still what it wants to measure. A new interface cannot decide whether a brochure download, an enquiry and a completed purchase have the same commercial value.

Document those definitions before comparing reports or changing implementation. Otherwise, a difference in measurement can be mistaken for a change in business performance.

Start With One Agreed Conversion

Choose the main business outcome for the review. For an online store, that may be a completed purchase with a unique order number. For a service business, it may be accepted work recorded in the sales system.

Keep supporting actions visible without treating them as equivalent. A contact button click is evidence of interest, not proof that a conversation happened. A submitted form is an enquiry, not a confirmed customer.

Write down when each outcome is recorded and how duplicates are handled. If refreshing a confirmation page creates another conversion, the campaign report may be counting browser behaviour rather than additional business.

Understand Why Two Platforms Can Claim One Sale

Advertising platforms may use different attribution windows, interaction types and credit rules. A platform can claim a conversion after an ad interaction even when another channel appears later in the customer's journey.

That is a reason to read the reporting definitions, not automatically an indication of fraud or a broken account. Inspect the actual settings and separate clicks from other credited interactions where the reporting allows it.

Suppose an imaginary customer clicks a social advertisement on Monday and a search advertisement on Thursday before placing one order. If both systems' rules include the purchase, each can report a conversion. The business still has one order. Its order system is the starting point for counting purchases.

Keep Campaign Naming Consistent

Use a written convention for campaign links. Decide how the team will name sources, mediums and campaigns, then reuse those conventions across email, social posts and partner promotions.

For example, a webinar campaign might use one campaign name while distinguishing the email newsletter from a partner post through the source and medium fields. The exact labels matter less than a consistent, documented approach your reporting can interpret.

Avoid putting personal information into campaign parameters. Do not add tracking tags to ordinary internal navigation merely to count clicks; that can complicate source interpretation. Use the appropriate interaction measurement for the analytics setup instead.

Test links before launch and keep a small register. Untagged or inconsistently named activity becomes harder to untangle after a campaign ends.

Reconcile Reports With A Simple Example

Imagine a month with twenty unique orders. The search platform claims twelve conversions and the social platform claims eleven. The website analytics report assigns ten purchases to paid search, six to social and four elsewhere.

Those numbers do not mean the business made forty three sales. They are different views of twenty orders, potentially with differences in timing, tracking coverage and credit rules.

Use each record for the question it can reasonably answer. Do not force them into agreement by quietly changing definitions until the totals look tidy.

Add Customer Context Without Treating Memory As Perfect

Ask new customers how they first heard about the business and what helped them choose. Keep the questions brief and allow more than one influence. A dropdown that demands one channel can erase a complex decision.

Treat the answer as another piece of evidence. People may remember a recommendation but forget an earlier advertisement, or name Google when they mean a branded search after a referral.

Combine the response with available sales notes and campaign timing. The purpose is to understand the journey better, not to declare one source of data universally correct.

Separate Attribution From Incremental Impact

Attribution asks how to allocate credit for observed outcomes. Incremental impact asks how many additional outcomes happened because of an activity. These are related but different questions.

A well designed comparison can help investigate incremental impact, such as a controlled campaign test with suitable groups. Small businesses should be cautious about drawing strong conclusions from tiny samples, different territories or periods with unrelated changes.

If you temporarily change advertising, record other influences such as promotions, stock availability and operating hours. A before and after comparison can suggest a direction without proving that advertising alone caused the difference.

Build A Report That Supports A Decision

Show completed business, acquisition spending, qualified enquiries and the main reporting limitations. Use consistent dates and identify whether revenue is booked or collected. Keep platform attributed results in their own clearly labelled view.

Review the economics alongside the attribution. Our paid search budgeting guide explains how customer acquisition allowances connect with lead quality and close rates.

Your next move: choose one month and reconcile the marketing reports against unique orders or accepted projects. The discrepancies will show you which definitions and tracking steps need attention before the next budget decision.

Written for the 2020 archive series. Platform references reflect the assigned period.