8 Attribution Models Compared: Which One Should You Use?

A user clicks a Google ad, reads a blog post a week later, opens an email campaign, and finally types your URL directly to make a purchase. Which touchpoint gets credit for the conversion? The answer depends entirely on your attribution model -- and choosing the wrong one can lead you to overinvest in channels that are not driving growth while underinvesting in the ones that are.

This guide covers eight attribution models, from the simplest (last-click) to the most sophisticated (Shapley value). For each model, we explain how it works, when to use it, and how Gurulu implements it. By the end, you will know which model fits your business and why.

What Is Attribution?

Attribution is the process of assigning credit for a conversion to the marketing touchpoints that influenced it. A touchpoint is any interaction between a user and your brand: an ad click, a social media post, an email open, a direct visit, an organic search result. Most users interact with multiple touchpoints before converting, and attribution models define the rules for distributing credit across that journey.

The 8 Models

1. Last-Click Attribution

The final touchpoint before conversion gets 100% of the credit. This is the default in most analytics platforms because it is simple to implement and easy to understand. The problem is that it completely ignores everything that happened before the last click. If a user discovered you through a Google ad, was nurtured by an email sequence, and converted by typing your URL directly, last-click gives all credit to direct traffic -- which tells you nothing useful.

Use when: You have short, single-touchpoint sales cycles, or as a baseline to compare against other models.

2. First-Click Attribution

The first touchpoint in the journey gets 100% of the credit. This model values discovery and awareness: it tells you which channels bring new users into your funnel. It is useful for understanding top-of-funnel performance but blind to everything that happens after initial discovery. A user might discover you through organic search and then need six more touchpoints before converting -- first-click ignores all of them.

Use when: You want to optimize for awareness and new user acquisition, or your funnel is top-heavy with most drop-off at the discovery stage.

3. Linear Attribution

Every touchpoint in the journey gets equal credit. If there were four touchpoints before a $100 conversion, each gets $25. Linear is fair but naive -- it assumes every interaction contributes equally, which is rarely true. The blog post that educated the user and the retargeting ad that reminded them to come back probably had very different impacts on the decision.

Use when: You have long, complex sales cycles with many touchpoints and no strong hypothesis about which stages matter most.

4. Time-Decay Attribution

Touchpoints closer to the conversion receive more credit than earlier ones. The decay follows an exponential curve with a configurable half-life (Gurulu defaults to 7 days). A touchpoint that occurred one day before conversion gets roughly twice the credit of one that occurred eight days before. This model reflects the intuition that recent interactions are more influential than distant ones.

Use when: Your sales cycle is measured in days or weeks and recent touchpoints are genuinely more influential (common in e-commerce and SaaS trials).

5. Position-Based (U-Shaped) Attribution

The first and last touchpoints each receive 40% of the credit, and the remaining 20% is distributed evenly among the middle touchpoints. This model recognizes that discovery (first touch) and conversion (last touch) are typically the most important moments, while middle touchpoints play a supporting role. Gurulu allows you to customize the split percentages.

Use when: You believe both acquisition and conversion channels matter more than nurture channels, which is true for most B2B and high-consideration B2C funnels.

6. Data-Driven Attribution

Instead of applying fixed rules, data-driven attribution uses machine learning to determine how much credit each touchpoint deserves based on your actual conversion data. The model analyzes converting and non-converting journeys to identify which touchpoints appear more frequently in successful paths. It adapts to your specific business rather than relying on generic assumptions.

Use when: You have sufficient conversion volume (Gurulu requires at least 300 conversions per month for reliable data-driven attribution) and want the model to reflect your actual user behavior.

7. Markov Chain Attribution

Markov chain attribution models the user journey as a series of state transitions. Each touchpoint is a state, and the model calculates the probability of reaching conversion from each state. The credit assigned to a touchpoint is based on the removal effect: how much the overall conversion rate would drop if that touchpoint were removed from all journeys. This provides a causality-adjacent measure of each channel's contribution.

Use when: You want a mathematically rigorous model that accounts for channel interactions and you have enough data to estimate transition probabilities reliably.

8. Shapley Value Attribution

Borrowed from cooperative game theory, Shapley value attribution computes the average marginal contribution of each touchpoint across all possible orderings of touchpoints. It is provably the fairest way to distribute credit: no other method satisfies the four axioms of efficiency, symmetry, linearity, and null player. The downside is computational cost -- Shapley values are exponential in the number of touchpoints -- but Gurulu uses sampling approximations to make it practical.

Use when: You need the most theoretically sound attribution and have complex, multi-channel journeys where fairness of credit distribution matters (common for enterprise marketing teams with large budgets).

Attribution in Practice

Theory is one thing; practice is another. Here is a concrete example showing how the same user journey yields different credit allocation under each model:

// Example: User journey with 4 touchpoints
// Google Ads → Blog Post → Email → Direct Visit → Purchase ($100)

// Last-click:     Direct Visit = $100
// First-click:    Google Ads   = $100
// Linear:         Each gets    = $25
// Time-decay:     Direct $40, Email $30, Blog $20, Ads $10
// Position-based: Ads $40, Direct $40, Blog $10, Email $10
// Shapley value:  Computed from marginal contributions

The differences are dramatic. Under last-click, you would conclude that direct traffic is your most valuable channel. Under first-click, you would invest heavily in Google Ads. Under Shapley value, you would see a more nuanced picture where every channel contributes. The "right" answer depends on what question you are trying to answer.

How Gurulu Implements Attribution

Gurulu offers all eight models in the dashboard, selectable from a dropdown on any conversion report. You can switch between models instantly to compare how credit shifts across channels. The data-driven, Markov chain, and Shapley value models run as background computations that update daily. Rule-based models (last-click, first-click, linear, time-decay, position-based) compute in real time.

Attribution in Gurulu is built on the identity resolution system. Because Gurulu can stitch sessions across devices and browsers, attribution models see the complete user journey -- not just the touchpoints from a single browser. This is critical for accurate attribution: if a user discovers you on mobile but converts on desktop, cookie-based tools attribute the conversion to direct traffic because they cannot link the two sessions. Gurulu attributes it correctly to the original mobile touchpoint.

Practical Recommendations

Start with position-based. If you are new to attribution modeling, position-based is the best starting point. It gives appropriate weight to discovery and conversion while acknowledging the middle funnel. It is easy to explain to stakeholders and produces reasonable results even with limited data.

Graduate to data-driven. Once you have enough conversion volume, switch to data-driven attribution. It will show you whether your assumptions about channel importance were correct -- and they often are not. Many teams discover that channels they were underinvesting in are actually critical conversion drivers.

Use multiple models for budget decisions. Never base budget allocation on a single attribution model. Compare at least three models side by side. If a channel shows up as important across multiple models, you can invest confidently. If it only shows up under one model, dig deeper before committing budget.

The Bottom Line

Attribution is not a solved problem, and no single model is universally correct. The value of attribution modeling is not in finding the "true" answer -- it is in understanding how different assumptions about credit allocation change your view of channel performance. The teams that get attribution right are the ones that use multiple models, question their assumptions, and update their approach as their data and business evolve.

Gurulu makes this practical by offering all eight models on the same data set, powered by cross-device identity resolution. You can switch models in seconds, compare results, and make informed decisions about where to invest your marketing budget.