All resources
People rarely buy on their first visit. Most conversions happen after multiple interactions across channels like email, social media, or paid ads. To understand which touchpoints drive results, marketers turn to attribution modeling.
It’s especially important in digital marketing, where customers move between platforms before making a decision. Attribution helps teams evaluate performance, optimize strategy, and measure campaign impact more accurately.
In this article, we’ll break down key attribution models, from GA4’s built-in options to advanced methods like Markov chains and the Shapley value, and guide you in choosing the best fit for your business.
Note: This article was originally published in 2019 but was updated to be truly a comprehensive guide in May 2025.
Also, we’ve added an emerging point for marketers here: How to evaluate the cost efficiency of the marketing attribution model.
Attribution modeling is about figuring out which marketing channels or touchpoints deserve credit for a sale. Since customers often interact with your brand several times before buying, it helps to know what’s working and what’s not.
Some models attribute all the credit to a single touchpoint, while others distribute it across multiple touchpoints. The right model can show you which channels drive real results, so you can spend your budget wisely and improve your marketing strategy. In industries with longer customer journeys, single-touch models may not be enough, making multi-touch models a better fit.
Understanding what leads to a conversion enables marketing teams to make informed decisions, utilize budgets effectively, and enhance overall performance.
Here’s how attribution modeling adds value:
There are dozens of possible attribution models, which can be grouped into various types. They can be classified in different ways depending on the logic used in their calculations.
Despite their limitations, 41% of marketers still rely on last-touch attribution, while 75% now use multi-touch models to measure performance, and very few use algorithmic models. Marketers often struggle to show the impact of top-of-funnel activities, which are typically undervalued in most attribution models.
Why so?
3 important reasons lead to the numbers above:
It’s important to remember: there’s no one-size-fits-all attribution model. The right choice depends on your goals, customer behavior, and marketing mix.
Using more than one model can give you a broader view of what drives performance, helping you understand how different channels influence each stage of the customer journey.
Let’s start with the simplest marketing attribution models, based on position, which are available in the free version of Google Analytics. These attribution models help marketers evaluate the contribution of different marketing activities to conversions.
With these models, all the value received from a conversion is attributed to the first source that led the user down the funnel, also known as the first-touch attribution model. For example, if we have a chain of four contacts, as shown below, according to the First Click model, the entire value of the conversion will be attributed to the CPC channel.
Pros: Easy to set up and use, as there are no calculations or arguments regarding the distribution of value between channels.
Cons: It overestimates the first touchpoint and ignores the rest of the journey. This model gives a limited view, missing the impact of retargeting or nurturing efforts.
Suitable for businesses focused on growing brand awareness, expanding reach, or identifying high-performing acquisition channels. It’s also a good fit for products or services with short sales cycles, where customers tend to convert quickly after the first visit. In such cases, the first point of contact plays a crucial role in influencing the decision.
Here is another example for more clarity: if a customer finds your business through a search ad, completes a subscription form on your site, receives an email, and is then retargeted with an ad on Facebook before actually converting, the search ad will receive 100% credit for this conversion.
According to this model, also known as the last interaction attribution model, the entire value of the conversion is attributed to the last channel with which the user came into contact before the conversion. The contribution of all other channels is ignored. In our example, all the values will go to the Direct channel.
Pros: A popular model that’s familiar to many marketers; ideal for evaluating campaigns for quick purchases, such as for seasonal items. It also directly shows the final touchpoint that led to a conversion.
Cons: Like all single-channel models, the last interaction attribution model ignores the role of other sources in the chain before ordering, especially if the sales cycle is long.
Suitable for businesses with a short sales cycle that uses up to three advertising channels.
With last-touch attribution, you will gain an understanding of the types of messages and channels that nudge leads over the finish line. These insights can then be used to build more effective retargeting campaigns and for the final stages of your nurture campaigns.
The entire conversion value is assigned to the last channel in the chain. However, if that last channel is Direct, then the value is attributed to the previous channel.
The logic is simple: if the user came to you from a bookmark or entered a URL, then they most likely were already familiar with your brand.
Pros: Allows you to ignore channels that are insignificant in terms of advertising costs and focus on paid sources. In addition, Last Non-Direct Click can be used for comparison with other attribution models.
Cons: Doesn’t take into account the contribution of other channels in the user journey.
Often, email is the second-to-last step before conversion. With Last Non-Direct Click, you risk undervaluing the earlier channels that first introduced the customer to your brand and led them to share their email.
Suitable for businesses that want to evaluate the effectiveness of a particular paid channel and for whom brand recognition is no longer so important.
Earlier, Google Analytics Universal reports used this model by default.
However, nowadays, it largely depends on the quality of measurement and user behavior data collection. Because privacy regulations and cookie restrictions lead to inaccurate acquisition campaign tracking.
The three models above are referred to as single-touch or single-channel attribution models. They give all the credit for a conversion to just one interaction, usually the last one. While they’re easy to use, they don’t show the full customer journey and can lead to wrong conclusions. These models can also be easily influenced, which makes the data less reliable.
On average, it takes from seven to nine touchpoints (the best performers can reduce this to five) for a customer to convert. Obviously, each of these touchpoints (channels) affects the probability of conversion. It’s only logical that it’s necessary to take into account the influence of all channels on the buyer in this case.
So let’s proceed to understand the multi-channel marketing attribution models.
The linear model is one of the simplest multi-channel attribution methods. It gives equal credit to every touchpoint a customer interacted with before converting. Since the value is evenly split, it’s sometimes called fractional attribution. While this model is easy to understand, it doesn’t consider the different impact each channel might have on the final decision.
Pros: Simple, but at the same time more advanced than single-channel attribution models, as it takes into account all sessions before conversion. The linear or equal model is also the most straightforward way to explain the ROI of each marketing channel to your stakeholders.
Cons: May under or overestimate the influence of a marketing channel. That is why this model is useless if you need to reallocate the budget; to divide it equally between channels is not the best option, since they can’t be equally effective.
Suitable for businesses such as B2B companies with a long sales cycle, for which it’s important to maintain contact with the client at all stages of the funnel.
The Time Decay model assigns more weight to touchpoints that occurred closer to the time of conversion. Early interactions get less value, while later ones, especially the last, are given more weight.
Pros: All channels in the chain get their piece of the cake. Works well for clients with businesses heavily dependent on relationship-building. Useful for optimizing your most influential touchpoints right before conversion, at the end of the day, clients & revenue always matter the most.
Cons: The contribution of sources that led the user to the funnel is greatly underestimated. Not as effective for businesses with shorter sales cycles
Suitable for those who are running short-term campaigns or selling expensive products that take longer for people to decide on. But if your sales process has several key steps and the cost of touchpoints is very different, a W-shaped model might give you a clearer picture by focusing on the most important moments in the journey.
This model is popular among marketers who want to focus on the final steps of the customer journey. It helps optimize the channels that drive last-mile conversions and fine-tune strategy based on timing.
This model gives 40% credit to both the first and last touchpoints, with the remaining 20% shared between the other touchpoints. But it prioritizes the first interaction and the final touchpoint before conversion as the most influential. Also known as position-based attribution, it highlights how users first find your brand and what drives them to convert.
Pros: Gives the greatest value to the channels that, in most cases play the most important role: those that attract the customer and motivate the conversion.
Cons: Sometimes, mid-funnel sessions play a bigger role than they appear, like helping users add items to the cart or subscribe. The Position-Based model tends to undervalue these touchpoints.
Suitable for businesses for whom it’s equally important to attract a new audience and convert existing visitors into buyers
Important! The first click, linear, time decay, and position-based attribution models are no longer available by default in Google Analytics 4 (starting July 2023). If you need these attribution models and don't want to give them up, you can continue to use them in OWOX BI. With OWOX BI you can get those simple models in a few clicks (as well as more advanced algorithmic models)
When choosing a model, many marketers prefer the one that looks easiest and is most understandable, even though it underestimates all sessions in the chain except for the last click.
In our opinion, there are three main reasons behind this:
To improve accuracy, data should be collected from multiple sources, online and offline, using tools like UTM tags, event tracking, and custom parameters. Budget and skill gaps can also limit how well attribution is set up.
Finally, the quality of your data matters. With stricter privacy rules and the loss of third-party cookies, using clean, first-party data is key for accurate attribution. And while platforms like Google or Facebook offer their own models, they only look at their own data, giving you a limited view.
For basic campaign checks in tools like Google Ads or GA4, simple models like Last Non-Direct Click work fine. But if you want to understand how different channels work together, you need to combine data from GA4, ad platforms, and your CRM into one system.
Algorithmic attribution models help track customer journeys across channels and assign credit more accurately. They also reveal how channels influence each other and improve decision-making. Tools can automate data integration and make these models easier to manage.
Algorithmic attribution models include Data-Driven (default in Google Analytics 360 and GA4), Markov chains, Machine Learning Funnel Based Attribution by OWOX, and custom algorithms.
Earlier, data-driven attribution was only available in Google Analytics 360, but with GA4, it’s now available to all users.
Unlike rule-based models, this one doesn’t follow fixed rules. Instead, it uses your actual data to calculate how much each channel contributed to a conversion, based on the Shapley method. It doesn’t depend on the order of touchpoints, only on whether a channel helped drive the conversion or not.
According to Wikipedia, the Shapley value (which belongs to cooperative game theory) is the optimal distribution of winnings between players. It’s a distribution in which the gain of each player is equal to their average contribution to the well-being of the group as a whole.
To understand how a Data-Driven model works, consider a specific example. Suppose we have two chains that lead to transactions:
We’ve specifically used short chains in our example so as not to complicate an already complicated formula.
Now, we’ll determine how much each channel brought in separately and how much they brought in together:
V1 (Facebook + Direct) = $500
V2 (Direct) = $300
V3 (Facebook) = 0
The Shapley value of a channel is calculated using the following formula:
Where:
If we insert the values from our example into this formula, we get the following:
F1 = (1 - 1)! × (2 - 1)! / 2! × (0 - 0) + (2 - 1)! × (2 - 2)! / 2! × ($500 - $300) = 0 + $100 = $100
F2 = (1 - 1)! × (2 - 1)! / 2! × (300 - 0) + (2 - 1)! × (2 - 2)! × ($500 - 0) = $150 + $250 = $400
F1 is the value of the Facebook channel.
F2 is the value of the Direct channel.
We’ll now explain this in simple words for those who got scared by the formula :) Let’s start with Facebook.
This channel hasn’t brought us anything on its own, so the first element we’ll have is 0.
Facebook, in combination with Direct, brought in $500, and Direct alone brought in $300.
We subtract the money earned by Direct from the amount that the combination of channels has brought in, then divide the result by two: ($500 - $300) / 2 = $100. This is our second element.
Now add $0 + $100 = $100. This is the value of the Facebook channel.
Next, consider the value of the Direct channel. It brought in $300 independently. Divide that by two and you get $150. The Facebook + Direct combination brought in $500, which divided by two gives us $250. We add these numbers and get $400 as the value of the Direct channel.
DDA is suitable for anyone who wants to determine which campaigns and keywords work as efficiently as possible and utilize this information to allocate the marketing budget effectively. Not suitable for businesses that need to know not only the value but also the position of the channel in the chain.
If your business has specific tracking needs, you might consider creating a custom attribution model. This allows you to set your own weights for each touchpoint and include past performance data for deeper insights.
A Markov chain is a method that looks at how removing a touchpoint affects conversions. It helps measure the impact of each step in the customer journey, especially when channels work together. Originally used in fields like weather forecasting and betting, Markov models are now common in digital marketing.
As a multi-touch attribution model, it assigns credit based on the influence of each touchpoint, making it easier to understand which steps matter most for conversions.
To understand how Markov chains work, consider a specific example from e-commerce. Suppose we have three chains:
C1 → C2 → C3 → Purchase
C1 → Unsuccessful conversion
C2 → C3 → Unsuccessful conversion
C1, C2, and C3 are sessions with three different sources – for example, Google CPC, Facebook, and email.
Fill in the following table:
In the first column we have the customer path – three chains, in our example.
The second column shows how the path will look inside the model. We added the entrance to the funnel (the Start stage) and the exit from the funnel (Conversion or Null, meaning failed conversion).
In the third column, we divided the channels into pairs, since we need to evaluate all possible transitions from one step to the next.
Then we need to calculate the probabilities for each of the possible transition options and put them in a separate table. These figures are considered empirically; that is, real data about user actions is analyzed – for example, from Google Analytics. This is done using a programming language such as R or Python.
The numbers in the table above are just examples. To make it clear what these numbers mean, let’s show them on a chart:
In this chart, we see all possible transition options from our example. Everything starts with the start phase. From there, a third of users go to channel C2 and two-thirds go to C1. Furthermore, half of the users from channel C1 exit the funnel, while the other half proceed to C2, then to C3. Finally, 50% of the remaining users make a purchase. There are a few more options, but you understand the principle.
Note that in our example, there are essentially only two paths to conversion, and both of them go through channel C2.
How are such sessions evaluated? With the removal effect. That is, we delete each of the sources in turn and see how its absence would affect the number of conversions:
For example, if we remove the C1 source from our example, we lose 50% of the conversions. How did we figure this out?
Calculating the value of channels is carried out in three stages:
1. First, we need to calculate the conversion probability for each channel. More precisely, we need to figure out how many conversions we get if we remove a specific channel from the chain.
The conversion probability (P) for each channel is calculated using the following formula:
P1 = (0.33 × 1 × 0.5) = 0.167
P2 = (0.33 × 0 × 0.5) = 0
P3 = (0.33 × 1 × 0) = 0
Let’s take a closer look at the first formula. This is our probability of conversion for channel C1. We remove the C1 channel from the model and multiply all the remaining transition probabilities from the chains that lead to the purchase. That is, we multiply 0.33 by 1 by 0.5. As a result, we get 0.167, or 16.7%. This is the conversion percentage we would get if we removed the C1 source from the funnel.
If we remove channels C2 and C3, then we’ll have no conversions at all.
2. Next, we determine the deletion effect (R) for each channel. This shows the percentage of conversions we’ll lose if we remove the channel from the funnel, and it’s calculated as follows: the unit of conversion (P) divided by the number of users at the beginning of the chain (transition probability) is subtracted from 1 (i.e. 100%).
R1 = 1 - 0.167 / 0.33 = 0.5
R2 = 1 - 0 = 1
R3 = 1 - 0 = 1
3. Finally, we calculate the value (V) of each channel. To do that, take the percentage of lost conversions (R) and divide it by the sum of all coefficients (R1, R2, and R3).
V1 = 0.5 / (0.5 + 1 + 1) = 0.2
V2 = 1 / (0.5 + 1 + 1) = 0.4
V3 = 1 / (0.5 + 1 + 1) = 0.4
Machine Learning Funnel-Based Attribution by OWOX BI helps you assess the mutual influence of channels on encouraging a customer through the funnel and achieving a conversion.
Previously, OWOX BI calculated the value of channels using a proprietary algorithm. However, we recently began using the Markov chain in our calculations.
In the Machine Learning Funnel-Based Attribution by OWOX BI, you can use the following information:
All this data can be analyzed in a complex and used to configure funnel steps (the default funnel is Enhanced E-commerce). You can add any steps and data – for example, transactions from your СRM and any other online and offline events (calls, meetings, etc.).
This model is flexible. You can customize funnel steps, add offline or CRM events, and assign weights to touchpoints based on your business needs. It’s ideal for companies with long sales cycles or rich historical data.
By using one ID across devices and channels, you get a complete view of user actions, from the first interaction to the final conversion.
It is suitable for anyone who wants to take into account every step of the user in the funnel and honestly evaluate advertising channels. Book a demo meeting with our specialists and find out what can be done for your company!
Choosing the right attribution model isn’t one-size-fits-all. It depends on how your business sells, how customers engage with your brand, and which channels you use.
Choosing the right model ensures you’re measuring marketing performance accurately. It helps you focus budget and strategy on what truly moves customers forward.
Basic models like Last Non-Direct Click miss the full customer journey. Funnel-based models give credit to every step, offering a clearer view of what drives conversions.
To measure efficiency, calculate the return on ad spend (ROAS) and compare performance across models. Run tests over 30–90 days and segment audiences to see which campaigns deliver the best return. These insights help you spend more on what truly works.
Start by choosing a KPI to measure campaign performance, like ROAS, ROI, CPA, or ad cost share. Comparing this metric across campaigns shows which one performs better.
When building your report, compare results from different models. Use Funnel-Based Attribution for one value and your current model (likely Last Non-Direct Click from GA) for the other. This helps you see which channels and campaigns truly deliver the most value, not just the last click.
The ROAS Difference shows how results change when using Funnel-Based Attribution instead of Last Non-Direct Click. It reveals how much more value a campaign adds when you consider its full role in the user journey, not just final orders.
How do you understand if the new attribution model and the ROAS you’ve calculated provide a more objective measurement of your advertising campaigns? It’s quite simple — use them while allocating your advertising budget and compare the results.
Since one user may interact with many campaigns, classic A/B testing won’t work. Instead, create non-overlapping audience segments (like different regions) that are still influenced by marketing.
Within segment A, let’s work it out “the old-fashioned way.” When allocating the budget in the B segment, let’s calculate ROAS according to the new model.
Your report will show which campaigns are overvalued (negative Cost Difference) and which are undervalued (positive Cost Difference). Attribution modeling helps you decide which campaigns should get more or less budget based on true performance.
Before shifting the budget, decide what matters most to your business:
Also, check if the channel has room to grow. If boosting its budget only increases the cost per click without more traffic, it may not be worth the extra spend.
The testing period depends on your funnel, from a user’s first visit to their final purchase. Campaigns that drive top-of-funnel traffic often take longer to show results and are usually undervalued by Last Non-Direct Click models.
In most cases, 30 to 90 days is a good testing window. The first part of that time is used to reallocate your budget based on the new model, while the later part is for tracking results.
That said, testing time can vary. E-commerce funnels may move faster than real estate or B2B. Choose a timeline that fits how long your customers usually take to convert.
Simply put, when you know the real cost behind each conversion, you can spend your budget more wisely and focus on what truly drives results.
Attribution only works if you apply it. Reports alone won’t help unless you use the data to adjust your strategy and budget.
If you want to know the real effectiveness of your marketing:
If you didn’t find an attribution model suitable for your project in this article, you can create a custom model as Answear did. Schedule a call with us and we’ll be happy to help you set up the attribution model (or a few models) you want to test for your business.
