Attribution is about answering the simplest and hardest question in marketing: Which channel caused this conversion? The answer is rarely simple because the customer journey in 2026 crosses many channels and devices. According to Google, the average customer has over 20 touchpoints before conversion (Google, "Customer Journey to Online Purchase," 2024). The choice of attribution model determines how you distribute credit for the conversion among these touchpoints — and thus how you allocate budget.
What attribution is and why it matters
Attribution is the process of assigning credit for a conversion to the various marketing channels, campaigns, and touchpoints that contributed to it. If a user first sees a Facebook ad, then searches on Google, clicks an organic link, and finally converts after clicking an email campaign — who gets the credit?
The answer depends on which attribution model you use, and this is not an academic question. Your attribution model directly affects budget allocation. If you use last click attribution, the email gets all the credit in the example above, and you may undervalue Facebook's contribution to introducing the user to the brand. Use first click, and Facebook gets everything — even though the user might have found you anyway.
Marketing budgets in businesses typically range from 5-15 percent of revenue (Virke, "Handelsrapporten," 2025). Incorrect attribution can lead to cutting channels that actually drive growth, or over-investing in channels that only harvest conversions other channels have sown. Proper attribution understanding is therefore directly linked to profitability.
Attribution model types: A thorough review
Last click gives 100 percent of the credit to the last channel the user interacted with before conversion. This is the simplest model and has historically been the default. The advantage is that it is easy to understand and implement. The disadvantage is that it ignores everything that happened earlier in the customer journey — it gives zero credit to the channels that created awareness and interest.
First click gives 100 percent of the credit to the first channel the user interacted with. This model values brand building and awareness but ignores all subsequent touchpoints that pushed the user toward conversion. Linear attribution distributes credit equally among all touchpoints — if you see four channels in the journey, each gets 25 percent. Simple, but does not account for the fact that different touchpoints have different influence.
Time decay gives more credit to touchpoints closer to conversion, with credit diminishing exponentially backwards in time. This model assumes that more recent interactions are more important. Position-based (also called U-shaped) typically gives 40 percent to the first touchpoint, 40 percent to the last, and distributes the remaining 20 percent equally among all intermediate touchpoints. This model values both discovery and conversion.
- Last click: 100% to last touchpoint — simple, but ignores the entire customer journey
- First click: 100% to first touchpoint — values awareness, ignores conversion effort
- Linear: Equal distribution among all touchpoints — fair, but undifferentiated
- Time decay: More credit to recent touchpoints — assumes recency matters more
- Position-based: 40/20/40 — values first and last touchpoints
- Data-driven: Machine learning based — uses actual conversion data to calculate contribution
Data-driven attribution in GA4: How it works
GA4 uses data-driven attribution (DDA) as its default model. DDA uses machine learning to analyze all conversion paths in your data and calculate the actual impact each touchpoint has on conversion probability. Instead of using fixed rules (like last click or linear), DDA looks at what actually happens in your data.
Technically, DDA works by comparing conversion paths with non-conversion paths. If users who see a Meta ad before converting via email have a higher conversion rate than users who only convert via email, the Meta ad receives credit for having increased the probability. The model uses Shapley values (from game theory) to distribute credit fairly based on each touchpoint's marginal contribution.
For DDA to work well, you need sufficient data volume. Google recommends a minimum of 600 conversions over 28 days for a reliable model (Google, "Data-driven attribution requirements," 2025). For conversion events with low volume, DDA will fall back to a rule-based model. Check your conversion events in GA4 under Admin > Attribution settings to see which model is actually being used.
Google Ads attribution vs. GA4 attribution: Why the numbers differ
One of the most frustrating aspects of digital marketing is that Google Ads and GA4 report different conversion numbers for exactly the same campaigns. This is due to fundamental differences in how the two systems measure and attribute conversions.
Google Ads uses its own attribution model (which you select in Google Ads under Tools > Conversions) and counts conversions per click. If a user clicks an ad three times before converting, Google Ads can assign the conversion to whichever click the model selects. GA4 uses its attribution model (DDA by default) and distributes credit across all channels — including organic, direct, and other sources.
Additionally, the two systems have different conversion windows. Google Ads typically has a 30-day click window and a 1-day view window, while GA4 has a 90-day conversion window by default. Different windows mean that a conversion that happens 45 days after a click will appear in GA4 but not in Google Ads. Accept that the numbers will never match — use Google Ads for campaign optimization and GA4 for the big picture. For more on GA4 setup, see our GA4 guide.
Meta attribution: View window and click window
Meta (Facebook/Instagram) has its own attribution system that operates independently of GA4. The default attribution setting in Meta Ads Manager is 7-day click and 1-day view. This means that if a user sees an ad (without clicking), and converts within 24 hours, Meta takes the credit. If the user clicks the ad and converts within 7 days, Meta takes the credit.
View-through conversions are especially controversial. Meta counts them because ad exposure can influence the purchase decision without a direct click — and this is technically correct. The problem is that it can significantly inflate conversion numbers, especially for brands with high organic conversion. A user who would have converted anyway may be counted as a Meta conversion simply because they saw an ad.
For a more conservative picture, we recommend comparing Meta-reported conversions with GA4 data for the same campaigns. The difference between the two numbers gives you an indication of how much of Meta's reported conversions are incremental. For a complete walkthrough of Meta tracking, see our Facebook Pixel and CAPI guide. For detailed PPC reporting, see our PPC reporting guide.
Cross-channel attribution: Walled gardens and cookie limitations
The fundamental problem with digital attribution in 2026 is that no single platform has full visibility into the customer journey. Google sees Google data, Meta sees Meta data, and neither sees what the other does. These walled gardens mean that each platform reports its own version of the truth, and the sum of all platforms' conversions is almost always higher than the actual number of conversions.
Cookie limitations worsen the situation. Safari removes third-party cookies entirely and limits first-party cookies set via JavaScript to 7 days. Chrome has announced changes to cookie policy multiple times. In practice, this means that long customer journeys (common in B2B and for expensive products) are not captured completely by any platform.
For businesses with customer journeys that span weeks or months, this is a real challenge. The solution is not to find the one perfect attribution tool, but to accept that all attribution is imperfect and use a combination of approaches: platform reports for tactical campaign optimization, GA4 for the big picture, and incrementality testing for strategic decisions.
Incrementality testing: The gold standard for attribution
Incrementality testing answers the question that no attribution model can answer: How many of these conversions would have happened anyway — without the marketing? This is done by splitting the audience into a test group (that sees the ad) and a control group (that does not), and comparing conversion rates.
Holdout tests are the simplest form: you pause advertising in a geographic area or for a user segment for a period, and compare conversions with a similar area/segment where advertising continues. Geo experiments are a more sophisticated variant where you use GeoLift or similar frameworks to control for seasonal variations and external factors.
Meta offers built-in incrementality testing (Conversion Lift Studies) for advertisers with sufficient volume. Google offers similar functionality through Brand Lift and Conversion Lift. The downside is that these studies require significant advertising volume and time (usually a minimum of 2-4 weeks) to produce statistically significant results. At LeapJet, we recommend running incrementality tests at least once every six months for the largest channels.
Marketing mix modeling (MMM) for budget allocation
Marketing mix modeling is a statistical approach that uses historical data to estimate the effect of different marketing channels on sales or conversions. Unlike digital attribution that tracks individual user journeys, MMM analyzes aggregated data over time — for example weekly advertising spend per channel compared with weekly revenue.
MMM has experienced a renaissance in recent years, partly because cookie limitations make digital attribution less reliable. Google has launched Meridian, an open-source MMM framework, and Meta has Robyn. These tools make MMM accessible to businesses that previously could not afford expensive consulting projects.
For businesses with advertising budgets over 500,000 NOK per year, we recommend considering MMM as a supplement to digital attribution. MMM requires a minimum of 2-3 years of historical data and a certain degree of variation in media spend to produce meaningful results. Combined with incrementality testing, MMM provides a robust foundation for strategic budget allocation that is not dependent on cookies or platform-specific reporting.
Practical attribution strategy: Which model should you use
There is no single perfect attribution model, but there are right models for different purposes. For daily campaign optimization in Google Ads, use Google Ads' own attribution (DDA if you have enough volume). For a big-picture view across channels, use GA4 with DDA. For strategic budget allocation between channels, use incrementality testing and MMM.
A practical framework is to think in three layers. Layer 1 (tactical): Use each platform's own reports to optimize campaigns within each platform. Layer 2 (tactical-strategic): Use GA4 to understand the customer journey across channels and identify synergies. Layer 3 (strategic): Use incrementality testing and MMM to validate that channels actually deliver incremental value.
For most businesses, the most important thing is to master layers 1 and 2 before investing in layer 3. Ensure that GA4 setup, conversion tracking, and UTM tagging are in place and consistent — without reliable data, no attribution model is meaningful. For tracking system setup, see our conversion tracking guide. For dashboard setup, see our marketing dashboard guide.
From attribution to contribution analysis: The next step
Attribution focuses on distributing credit for conversions, but a more mature approach is contribution analysis: understanding the unique value each channel adds to the entire marketing system. Instead of asking who caused this conversion, you ask what would have happened if this channel did not exist?
Contribution analysis uses incrementality testing, MMM, and controlled experiments to estimate the causal effect of each channel. It is fundamentally different from attribution because it attempts to measure causality (cause-effect) rather than correlation (association). A channel can have low attribution credit in DDA but high incremental value — for example branded display that increases conversion rates across all other channels.
For businesses that already have robust attribution in place, the next step is to build a culture where attribution is seen as an indicator, not a truth. Combine attribution data with business intuition, customer insight, and experimentation. No model will give you the perfect answer, but a combination of approaches gives you a decision-making foundation that is far better than gut feeling. See our glossary for definitions of attribution model, conversion tracking, and data-driven attribution. For services related to analytics and optimization, see our SEO and analytics services.
