Conversion Model: Choosing the Right Attribution Model for Your Business

1. Understanding Attribution Models

1. What Are Attribution Models?

Attribution models are frameworks that allocate credit to different marketing touchpoints based on their influence in driving conversions. These models help answer questions like:

- Which channels or interactions contributed most to a sale?

- How should we distribute credit when a customer interacts with multiple touchpoints before converting?

Consider a user journey: A potential customer might discover your product through a Google search, then click on a Facebook ad, and finally make a purchase after receiving an email newsletter. Attribution models help attribute value to each of these touchpoints.

2. Common Attribution Models:

- Last-Touch Attribution (Last Click): This model assigns all credit to the last touchpoint before conversion. It's simple but often oversimplifies the customer journey. For example, if a user clicks on a paid ad right before converting, the ad gets all the credit.

- First-Touch Attribution (First Click): The opposite of last-touch, this model gives full credit to the first interaction. It's useful for understanding initial awareness but ignores subsequent touchpoints.

- Linear Attribution: In this model, credit is evenly distributed across all touchpoints. It acknowledges the entire journey but doesn't consider varying impact.

- time Decay attribution: Closer touchpoints receive more credit. For instance, a user who clicked an ad yesterday gets more credit than one who clicked a month ago.

- Position-Based (U-Shaped) Attribution: Here, the first and last touchpoints get more credit, while middle interactions receive less. It reflects the idea that awareness and closing actions matter most.

3. Examples:

- Imagine a user journey: organic search → social media ad → email → purchase. Let's apply different models:

- Last-Touch: The purchase gets all credit.

- First-Touch: Organic search receives all credit.

- Linear: Each touchpoint gets equal credit.

- Time Decay: Email and purchase get more credit.

- Position-Based: Organic search and purchase receive more credit.

4. Challenges and Considerations:

- Data Availability: Accurate attribution requires comprehensive data on user interactions.

- Channel Interplay: Users often engage with multiple channels. Models should account for cross-channel effects.

- Business Goals: Different models suit different objectives (e.g., brand awareness vs. Direct sales).

- Custom Models: Some businesses create hybrid models tailored to their unique context.

5. Best Practices:

- Analyze Multiple Models: Compare different models to understand nuances.

- Segmentation: Analyze attribution by user segments (e.g., new vs. Returning customers).

- Experiment: Test different models and assess their impact on decision-making.

In summary, attribution models are powerful tools for understanding user behavior and optimizing marketing efforts. By considering diverse perspectives and using data-driven insights, businesses can choose the right attribution model to drive growth and improve ROI. Remember, there's no one-size-fits-all solution; context matters!

2. Common Attribution Models

1. Last-Touch Attribution (or Last Interaction Model):

- In this model, all credit for a conversion is assigned to the last touchpoint before the conversion. It assumes that the final interaction is the most influential.

- Example: Imagine a user who clicks on a Facebook ad, then later searches for the product on Google and makes a purchase. The sale would be attributed solely to the Google search.

- Pros: Simple, easy to implement.

- Cons: Ignores other touchpoints that contributed to the conversion.

2. First-Touch Attribution (or First Interaction Model):

- Here, all credit goes to the initial touchpoint that introduced the user to the brand or product.

- Example: A user clicks on a banner ad, visits the website, and eventually converts. The banner ad receives full credit.

- Pros: Highlights the importance of awareness-building channels.

- Cons: Ignores subsequent interactions that may have influenced the user.

3. Linear Attribution:

- This model distributes credit equally across all touchpoints in the customer journey.

- Example: If a user interacts with a display ad, then clicks on an email link, and finally converts via organic search, each touchpoint gets 33% credit.

- Pros: Fairly represents the entire journey.

- Cons: Oversimplifies the impact of different touchpoints.

4. Time Decay Attribution:

- Credit is assigned based on the recency of the touchpoint. The closer to conversion, the more credit it receives.

- Example: A user sees a TV ad, clicks on a social media post, and then converts after a search. The search gets the most credit.

- Pros: Reflects the reality that recent interactions often have a stronger impact.

- Cons: May undervalue early touchpoints.

5. U-Shaped (Position-Based) Attribution:

- This model allocates 40% credit to the first and last touchpoints, with the remaining 20% distributed evenly across intermediate touchpoints.

- Example: A user discovers a brand through an influencer, engages with content on social media, and converts via a paid search ad. The influencer and paid search both receive significant credit.

- Pros: Balances the importance of initial and final interactions.

- Cons: Assumes that middle touchpoints are equally influential.

6. Custom Attribution Models:

- Businesses can create their own models based on their unique goals and insights.

- Example: A company might assign higher weight to touchpoints that lead to high-value conversions (e.g., subscriptions) and lower weight to touchpoints that result in low-value actions (e.g., newsletter sign-ups).

- Pros: Tailored to specific business needs.

- Cons: Requires data analysis and customization.

In summary, choosing the right attribution model depends on your business objectives, data availability, and the complexity of your customer journey. By understanding these common models and their nuances, you can make informed decisions that drive growth and optimize your marketing efforts. Remember, attribution is not a one-size-fits-all solution; it's a dynamic process that evolves as your business does.

3. First-Touch Attribution

1. Understanding First-Touch Attribution:

- Definition: First-touch attribution assigns 100% of the conversion credit to the initial touchpoint that led a user to your website or campaign. It's like giving credit to the first domino that set the entire chain in motion.

- Nuances: First-touch attribution assumes that the first interaction is the most influential in driving conversions. It's often used in scenarios where brand awareness and lead generation are critical.

- Example: Imagine a user clicks on a Facebook ad, lands on your website, and signs up for your newsletter. First-touch attribution would credit the entire conversion to that initial Facebook ad click.

2. Pros of First-Touch Attribution:

- Simplicity: First-touch is straightforward. It's easy to implement and understand.

- Focus on Awareness: If your goal is to measure brand awareness or top-of-funnel performance, first-touch attribution provides clarity.

- Historical Context: In historical data analysis, first-touch helps identify trends over time.

3. Cons of First-Touch Attribution:

- Oversimplification: By ignoring subsequent touchpoints, first-touch attribution oversimplifies the complex customer journey.

- Neglecting Mid-Funnel and Closing Touchpoints: Customers often interact with multiple touchpoints before converting. First-touch ignores these crucial interactions.

- Misaligned with Reality: Rarely does a single touchpoint solely drive conversions. Customers research, compare, and engage with your brand across channels.

4. Impact on decision-Making and strategy:

- Budget Allocation: First-touch attribution may lead to overinvestment in top-of-funnel channels (e.g., awareness campaigns) and underinvestment in mid-funnel and closing channels (e.g., retargeting).

- Content Optimization: Understanding the first touchpoint helps optimize content for initial engagement.

- Segmentation: Segmenting users based on their first interaction can inform personalized marketing strategies.

5. Balancing First-Touch with Other Models:

- multi-Touch models: Consider using multi-touch attribution models (e.g., linear, time decay, or position-based) alongside first-touch. These models distribute credit across touchpoints.

- Customization: Create hybrid models that blend first-touch with other models based on your business goals.

In summary, while first-touch attribution provides simplicity and clarity, it's essential to recognize its limitations. Businesses should explore hybrid models and consider the entire customer journey to make informed decisions. Remember, attribution isn't one-size-fits-all; it's about finding the right fit for your unique context.

4. Last-Touch Attribution

One of the most common and simple attribution models is the one that assigns all the credit to the last touchpoint before the conversion. This means that the final marketing channel that the customer interacted with before making a purchase or signing up for a service is considered the most effective and influential. This model is also known as the last-click or last-interaction model.

Some of the advantages and disadvantages of using this model are:

- Advantages:

1. It is easy to implement and understand, as it does not require complex calculations or assumptions.

2. It is useful for measuring the performance of direct response campaigns, such as email marketing or paid search, that aim to drive immediate conversions.

3. It can help identify the most popular and persuasive channels that customers use to make their final decisions.

- Disadvantages:

1. It ignores the impact of other touchpoints that may have contributed to the customer journey, such as awareness, consideration, or loyalty campaigns.

2. It may undervalue the role of channels that are more effective at generating initial interest or nurturing leads, such as social media, content marketing, or display advertising.

3. It may overvalue the role of channels that are more likely to be used as the last step, such as branded search or direct traffic, which may not reflect the true influence of the marketing mix.

An example of how this model works is:

- A customer sees an online banner ad for a new product and clicks on it, but does not buy anything.

- A few days later, the customer receives an email from the same company with a discount offer and clicks on it, but still does not buy anything.

- The next day, the customer searches for the product name on Google and clicks on the first result, which is the company's website. The customer then buys the product.

- According to the last-touch attribution model, the entire credit for the conversion is given to the organic search channel, as it was the last touchpoint before the purchase. The banner ad and the email are ignored, even though they may have influenced the customer's awareness and interest.

5. Linear Attribution

### 1. Understanding Linear Attribution

Linear attribution is a straightforward approach to assigning credit for conversions across various touchpoints in a customer's journey. Unlike other attribution models that emphasize specific touchpoints (such as first-click or last-click attribution), linear attribution distributes credit evenly across all touchpoints. In essence, it acknowledges that every interaction contributes to the final conversion, regardless of its position in the funnel.

### 2. Advantages of Linear Attribution

- Fairness: Linear attribution ensures fairness by acknowledging the role of every touchpoint. It avoids the bias of attributing all credit to the first or last interaction.

- Holistic View: By considering all touchpoints, linear attribution provides a holistic view of the customer journey. It captures the entire path, including awareness-building touchpoints and nurturing interactions.

- Simplicity: Linear attribution is easy to understand and implement. It doesn't require complex algorithms or data manipulation.

### 3. Limitations of Linear Attribution

- Equal Weighting: While fairness is a strength, equal weighting can be a limitation. Not all touchpoints are equally impactful. Some interactions may have a more significant influence on conversion than others.

- Ignoring Timing: Linear attribution doesn't account for the timing of touchpoints. A touchpoint closer to conversion may deserve more credit than one early in the journey.

- Context Matters: Linear attribution treats all channels equally, but context matters. For instance, a direct email from a salesperson might be more influential than a generic display ad.

### 4. real-World examples

Let's consider a few scenarios to illustrate linear attribution:

1. E-commerce Purchase:

- A customer discovers a product through a Facebook ad (touchpoint 1).

- Later, they search for reviews on Google (touchpoint 2).

- Finally, they make the purchase directly on the website (touchpoint 3).

- Linear attribution assigns equal credit to all three touchpoints.

2. B2B Lead Generation:

- A potential client attends a webinar (touchpoint 1).

- They receive a follow-up email (touchpoint 2).

- Eventually, they schedule a sales call (touchpoint 3).

- Linear attribution acknowledges the contribution of each step.

### Conclusion

Linear attribution offers a balanced perspective on customer interactions. While it has its limitations, it serves as a useful baseline for understanding the overall impact of touchpoints. Businesses can use it alongside other models to gain deeper insights into their marketing efforts. Remember that attribution models are tools, and the right choice depends on your specific business context and goals.

6. Time Decay Attribution

1. understanding Time decay Attribution:

- In time decay attribution, the credit for a conversion is distributed among the touchpoints based on their temporal proximity to the conversion event. Simply put, the closer a touchpoint is to the conversion, the more credit it receives.

- Imagine a customer journey where a user interacts with multiple touchpoints over time before making a purchase. Time decay attribution acknowledges that the influence of touchpoints diminishes as we move away from the conversion moment. It recognizes that the touchpoint just before the conversion deserves more credit than the one encountered weeks ago.

2. Mathematical Formulation:

- Time decay attribution assigns exponentially decreasing weights to touchpoints based on their chronological order. The most recent touchpoint receives the highest weight, while earlier touchpoints receive progressively smaller weights.

- The formula for time decay attribution can be expressed as:

\[ \text{Weight}(t) = \frac{1}{2^{(T - t)}} \]

Where:

- \(t\) represents the time elapsed between the touchpoint and the conversion event.

- \(T\) is a fixed time window (e.g., 30 days) within which touchpoints are considered relevant.

3. Example Scenario:

- Let's consider an e-commerce purchase. A customer's journey involves the following touchpoints:

- Day 1: Social media ad

- Day 7: Email newsletter

- Day 14: Google search

- Day 28: Direct website visit (conversion)

- Applying time decay attribution:

- The direct website visit (Day 28) receives the highest weight because it's closest to the conversion.

- The social media ad (Day 1) receives the least weight.

- The email newsletter (Day 7) and Google search (Day 14) fall in between.

4. Advantages of Time Decay Attribution:

- Reflects real-world behavior: Customers often respond more strongly to recent touchpoints.

- Fairness to late-stage touchpoints: Recognizes the impact of closing the deal.

- Simplicity: Easy to understand and implement.

5. Considerations and Limitations:

- Short vs. Long Conversion Paths: Time decay works well when conversion paths are relatively short. For longer paths, it may overemphasize recent touchpoints.

- Choosing the Time Window: Setting an appropriate time window is critical. Too short, and early touchpoints are ignored; too long, and recent touchpoints dominate excessively.

- Context Matters: Different industries and products may require customized time windows.

6. Conclusion:

- Time decay attribution provides a balanced view of touchpoint influence, considering both early and late interactions. It bridges the gap between linear attribution (which treats all touchpoints equally) and position-based attribution (which focuses only on first or last touchpoints).

- As you choose your attribution model, consider your business context, customer behavior, and the impact of time on touchpoint effectiveness. Time decay attribution might be the right fit if your conversion paths are relatively short and recent touchpoints matter significantly.

Remember, attribution models are not one-size-fits-all. Evaluate your specific business needs and experiment with different models to find the one that aligns best with your goals.

7. Position-Based Attribution

1. Introduction to Position-Based Attribution:

- Position-based attribution, also known as U-shaped attribution, recognizes that not all touchpoints are equal contributors to a conversion. Instead, it focuses on three critical positions:

- First Interaction (Awareness): The touchpoint where the user first interacts with your brand or product. It creates awareness and initiates the customer journey.

- Last Interaction (Conversion): The touchpoint immediately preceding the conversion. It often gets the most credit for the sale.

- Middle Interactions (Consideration): All other touchpoints between the first and last interactions. These play a role in nurturing and guiding the user toward conversion.

- Position-based attribution acknowledges that the first and last interactions are pivotal, but it also values the middle interactions that influence the decision-making process.

2. Weighting Scheme:

- In position-based attribution, the first and last interactions receive a fixed percentage of credit, while the middle interactions share the remaining credit.

- Commonly used weight distribution:

- First Interaction: 40% credit

- Last Interaction: 40% credit

- Middle Interactions (Combined): 20% credit

- This distribution reflects the idea that the initial touchpoint sparks interest, the middle interactions nurture it, and the final touchpoint seals the deal.

3. Example Scenario:

- Let's consider an e-commerce purchase journey:

- User A discovers a product through a Facebook ad (first interaction).

- Later, they research reviews and compare prices on various websites (middle interactions).

- Finally, they click on a Google search ad and make the purchase (last interaction).

- In position-based attribution:

- Facebook ad and Google search ad each receive 40% credit.

- The middle interactions (review sites, price comparisons) share the remaining 20%.

- This model acknowledges the entire journey, giving due credit to all touchpoints.

4. Strengths and Limitations:

- Strengths:

- Reflects the reality of multi-touch journeys.

- Balances the importance of initial awareness and closing the deal.

- Provides a more holistic view than simplistic models.

- Limitations:

- Assumes equal importance for all middle interactions (which may not always be true).

- Fixed weight distribution might not suit every business context.

5. Business Applications:

- Paid Advertising Optimization: Position-based attribution helps allocate ad spend effectively by considering the entire funnel.

- Content Marketing: Understand which content touchpoints contribute most to conversions.

- Product Development: Identify critical touchpoints for product adoption.

In summary, position-based attribution offers a balanced perspective on the customer journey, recognizing the unique roles of different touchpoints. By incorporating this model, businesses can make informed decisions about resource allocation and optimize their marketing strategies. Remember, it's not just about the last click; it's about the entire path that leads to conversion!

8. Custom Attribution Models

One of the most important decisions that marketers face when analyzing their conversion data is how to assign credit to the different touchpoints along the customer journey. Different attribution models can have a significant impact on how the performance of each channel is measured and optimized. However, none of the standard attribution models can capture the full complexity and uniqueness of each business and its customers. That is why some marketers opt for creating their own custom attribution models that suit their specific needs and goals. Custom attribution models allow marketers to assign custom weights to each touchpoint based on their own criteria and logic. This way, they can account for factors that are not considered by the standard models, such as seasonality, customer lifetime value, or channel interactions. Custom attribution models can also be more flexible and adaptable to changing business environments and customer behaviors.

There are several steps involved in creating and implementing a custom attribution model. Here are some of them:

1. Define the objective and scope of the model. The first step is to determine what the purpose and the scope of the custom attribution model are. For example, the objective could be to increase conversions, revenue, or retention. The scope could be to focus on a specific product, segment, or campaign. The objective and scope should be aligned with the overall business goals and strategy.

2. Identify and collect the relevant data. The next step is to identify and collect the data that will be used to build and evaluate the custom attribution model. This could include data from various sources, such as web analytics, CRM, email marketing, social media, etc. The data should be accurate, complete, and consistent across the different sources. The data should also include information about the touchpoints, such as the channel, the timestamp, the content, the action, etc.

3. Choose the attribution method and logic. The third step is to choose the attribution method and logic that will be used to assign credit to each touchpoint. There are different methods and logic that can be used, such as:

- Rule-based: This method uses predefined rules to assign credit to each touchpoint based on certain criteria, such as the position, the time, or the frequency of the touchpoint. For example, a rule-based model could assign 40% credit to the first touchpoint, 20% credit to the last touchpoint, and 10% credit to each of the middle touchpoints.

- Data-driven: This method uses statistical or machine learning techniques to assign credit to each touchpoint based on the data. For example, a data-driven model could use regression, clustering, or neural networks to estimate the contribution of each touchpoint to the conversion outcome.

- Hybrid: This method combines both rule-based and data-driven approaches to assign credit to each touchpoint. For example, a hybrid model could use a rule-based model to assign credit to the first and last touchpoints, and a data-driven model to assign credit to the middle touchpoints.

4. test and validate the model. The fourth step is to test and validate the custom attribution model using historical or simulated data. The model should be compared with the standard models and the actual results to evaluate its accuracy, reliability, and validity. The model should also be tested for its sensitivity, robustness, and scalability. The model should be able to handle different scenarios, such as changes in customer behavior, channel mix, or conversion rates.

5. Implement and optimize the model. The final step is to implement and optimize the custom attribution model in the marketing analytics platform. The model should be integrated with the existing data sources and reporting tools. The model should also be monitored and updated regularly to reflect the latest data and insights. The model should be used to measure and optimize the performance of each channel and campaign, as well as the overall marketing strategy.

9. Implementing and Analyzing Attribution

Once you have chosen the right attribution model for your business, you need to implement it and analyze its performance. This is a crucial step to optimize your conversion funnel and allocate your marketing budget effectively. However, implementing and analyzing attribution is not a one-time task, but a continuous process that requires careful planning and execution. Here are some tips to help you with this process:

- 1. Define your goals and KPIs. Before you implement any attribution model, you need to have a clear idea of what you want to achieve and how you will measure it. For example, do you want to increase conversions, revenue, customer lifetime value, or retention? What are the key indicators that reflect your success? How will you compare different attribution models and channels? Having a well-defined goal and a set of KPIs will help you focus your efforts and evaluate your results.

- 2. choose the right tools and platforms. Depending on your business size, complexity, and needs, you may need different tools and platforms to implement and analyze attribution. For example, you may use Google Analytics, facebook Ads manager, or a third-party attribution software to track and attribute your conversions. You may also need to integrate different data sources, such as CRM, email, or offline sales, to get a holistic view of your customer journey. Choosing the right tools and platforms will help you collect, process, and visualize your attribution data more efficiently and accurately.

- 3. Implement your attribution model consistently and correctly. Once you have chosen your tools and platforms, you need to make sure that you implement your attribution model consistently and correctly across all your channels and touchpoints. This means that you need to use the same attribution model and parameters for all your campaigns and sources, and ensure that your tracking codes and tags are working properly. You also need to account for any potential issues, such as cross-device tracking, cookie expiration, or data discrepancies, that may affect your attribution accuracy. Implementing your attribution model consistently and correctly will help you avoid any errors or biases in your attribution data.

- 4. Analyze your attribution data and insights. After you have implemented your attribution model, you need to analyze your attribution data and insights regularly and systematically. This means that you need to monitor your KPIs and performance metrics, such as conversion rate, ROI, CPA, or ROAS, for each attribution model and channel. You also need to compare different attribution models and scenarios, such as last-click, first-click, linear, or time-decay, to see how they affect your results and decisions. You also need to look for any patterns, trends, or anomalies in your attribution data, such as seasonality, spikes, or drops, and investigate their causes and implications. Analyzing your attribution data and insights will help you understand your customer behavior and preferences, and optimize your marketing strategy and tactics.

Startups should be - if you graph their financial performance, it should be what's called a J curve. You start out at zero. you're not making any money; you're not losing any money.

Chamath Palihapitiya