Customer journeys rarely follow a straight line. People discover a business on social media, think about it, compare options, come back through search or email and finally buy. Attribution tries to share credit across those steps. Understanding how it works helps you read reports correctly and avoid cutting channels that quietly do a lot of the work.
Common attribution models
| Model | How credit is assigned | Tends to favour |
|---|---|---|
| Last click | All credit to the final touchpoint | Search and direct visits |
| First click | All credit to the first touchpoint | Discovery channels such as social |
| Linear | Equal credit to every touchpoint | Long journeys |
| Time decay | More credit to recent touchpoints | Channels close to purchase |
| Position based | Most credit to first and last, some to the middle | Discovery and closing channels |
| Data driven | Credit based on patterns in your own data | Depends on the data |
The Academy's glossary explains attribution and attribution windows.
Why social is often undercounted
Social media often starts journeys rather than ending them. Someone sees a post, does not click, then searches for the brand days later. Last click models give all the credit to search. Social also suffers from cross-device journeys, in-app browsers and privacy limits that break tracking. The result is that social can look less valuable in reports than it is in reality.
Data-driven attribution
Google Analytics 4 uses data-driven attribution by default for many reports, assigning credit based on patterns in your conversion paths. It needs enough data to work well. For small businesses with few conversions, results can be unstable. The Academy's guide to GA4 for social media explains where to find attribution settings and reports.
Platform attribution
Ad platforms such as Meta and TikTok report conversions using their own attribution windows, often counting conversions that happen within a set number of days after a click or view. Their numbers usually differ from website analytics, and adding them together double counts sales. Use platform numbers to optimise within a platform and analytics to compare channels.
Attribution windows
An attribution window is the period after a click or view during which a conversion is credited. A seven-day click window counts conversions within seven days of clicking. Longer windows credit more conversions; shorter windows are stricter. Check which window each report uses before comparing.
Reading attribution sensibly
- Compare models in your analytics tool to see how credit shifts.
- Look at assisted conversions, where social appears earlier in the path.
- Use UTM tags consistently so social traffic is labelled.
- Ask customers how they heard about you.
- Treat every model as an estimate, not the truth.
The Academy's guide to UTM tracking explains labelling links.
Beyond attribution
Attribution only sees what can be tracked. Methods such as incrementality tests and marketing mix modelling estimate the true effect of channels, including untracked influence. The Academy's guides to incrementality testing and marketing mix modelling explain them. The Academy's glossary explains incrementality.
A small business approach
For most small businesses, a simple combination works: UTM tags on every link, a look at GA4's conversion paths, a question on enquiry forms asking how people found you and a monthly note of what customers mention. Together, these give a reasonable picture without expensive tools.
Organic social and attribution
Organic posts rarely carry links, and many people never click. Reach and engagement may influence later purchases without any trackable path. Platform signals, such as Instagram's documented watch time and completion, likes, shares (sends per reach), saves, relationship history✓, determine how far organic posts travel, but the sales effect often shows up indirectly. Watch for rises in branded search and direct traffic after strong social periods.
Talking about attribution with leadership
Explain that attribution models are estimates, show how social contributes earlier in the journey and use more than one source of evidence. Avoid claiming every sale for social, which damages credibility, and avoid accepting last click numbers as the full story.
A worked example
A furniture shop sees that last click attribution credits social with five sales in a month, while first click credits it with twenty. Customer surveys show that a third of buyers first saw the shop on Instagram. Branded searches rose during a month of strong Reels. Taken together, the evidence suggests social plays a much bigger role in discovery than last click shows, even if search usually closes the sale. That combined picture supports keeping investment in social content.
Choosing a model for reporting
Pick one model for regular reporting, so month-to-month comparisons are fair, and look at others when making big decisions. Many businesses report on the default data-driven model in GA4 and check first click occasionally to understand discovery. Write down which model you use, so anyone reading the report knows what the numbers mean.
Attribution and offline sales
Many sales happen offline, in a shop, over the phone or after a quote. Attribution tools miss these unless you connect them. Simple steps help: unique discount codes for social, a how did you hear about us question and a note in your booking system. These offline signals often reveal that social does more than analytics alone suggests.
Common mistakes
- Treating last click as the full story.
- Adding platform and analytics conversions together.
- Ignoring attribution windows when comparing reports.
- Claiming all sales for social.
- Forgetting offline and untracked influence.
Learning attribution
The Academy's second analytics module in the third level covers attribution models and conversion paths, and the advanced measurement module in the fourth level covers incrementality and modelling. The free assessment shows where you would start.
