Illustration showing a computer with an image
Internet Marketing

Challenges of Attribution in the Digital Age

11. 09. 2019

What is Attribution

In terms of marketing, attribution refers to pinpointing the specific marketing channel that creates a sale or lead.

With the move from traditional to digital advertising, accurate attribution is more important than ever. In the early days of digital advertising, it was almost impossible to track where a sale was coming from. The lack of sophisticated tracking technology made it difficult to measure results, which left companies feeling wary about investing money in digital marketing. If they couldn’t measure ROI, they didn’t want to commit precious marketing dollars to an unreliable marketing channel. As the medium evolved, however, platforms such as Google Analytics and Adobe Analytics can now provide advertisers with information about how their advertising investments performed.

Although tracking technology has drastically improved, marketers continue to focus on optimizing campaigns and improving overall ROI.

At 9thCO, we understand the importance of attribution and we know that generating real revenue is what matters most to clients. As a result, we don’t simply focus on generating a lead for a client. Instead, we ensure those leads result in actual sales.

Importance of Proper Attribution

The Struggle

No matter what channel you choose—TV, out of home, radio or digital—the focus of your advertising campaign is always performance. Performance goals can refer to increasing brand awareness, in-store sales or lead generation. No matter your channel or performance goals, the success of your campaign must be measurable.

An ongoing struggle in advertising is that attributing the reasons for a customer performing a desired action is difficult. For example:

  1. A customer browses the web on Sunday and receives an image ad from a furniture manufacturer.
  2. On Monday at work, the customer searches for “best leather sofas”.
  3. The customer clicks on an advertiser’s organic search result, which leads to a leather sofas product page. On Thursday at home, the customer searches again for “advertiser’s leather sofas” on a personal desktop and clicks on the advertiser’s paid search ad and completes a sofa purchase.

In this example, the purchase journey looks like this:
Display Advertising –> Organic Search –> Paid Search.

Without proper attribution modelling, the advertiser would give 100% of the credit for the sale to their paid search channel. However, the data did not properly attribute all touchpoints, with which the customer interacted, before making the purchase decision. In this case, how much credit should be given to the display ad for the purchase? Proper attribution would allow the advertiser to understand how much the display ad influenced the buyer.

Making Informed Decisions

Data helps marketers make informed decisions. However, making sure that data is accurate is crucial.

A performance-focused agency like 9thCO understands that collecting accurate data ensures accurate attribution. The end result is stronger marketing decisions and more effective advertising campaigns.

What are Attribution Models?

Image showing the four types of attribution models: Last click, Time decay, Linear, Position-based, First click

 

Google platforms currently have six main attribution models that can help marketers accurately collect and interpret the data they collect from users:

 

Last click
Gives 100% of the credit to the last consumer touchpoint.
First click
Gives 100% of the credit to the first consumer touchpoint.
Linear
Splits the credit evenly throughout all touchpoints in the consumer purchase journey.
Time decay
Gives more credit to the touch points that occurred closer in time to the time of conversion.
Position-based
Gives 40% credit to each the first and last touchpoints and splits the remaining 20% evenly through all the touchpoints in the middle of the purchase journey.
Data-driven
Uses machine learning and previous consumer data to dynamically distribute credit for each touchpoint in the consumer journey. With large amounts of historical and current data, data-driven attribution is the most accurate attribution model available. However, the eligible data requirements needed for this model aren’t easily accessible as it requires;
  • 15,000 clicks on Google Search
  • 600 conversions within 30 Days

Choosing an Attribution Model

These attribution models allocate more weight to the first interaction and are focused on growth. They give more credit to higher funnel clicks from broad keywords, such as “sofas”, or display ads. Consumers who interact with your brand through these sources probably haven’t heard of your brand or may not have considered your brand when beginning their shopping journey.

First Click/Position Based

These attribution models allocate more weight to the first interaction and are focused on growth. They give more credit to higher funnel clicks from broad keywords, such as “sofas”, or display ads. Consumers who interact with your brand through these sources probably haven’t heard of your brand or may not have considered your brand when beginning their shopping journey.

Last Click/Time Decay

Alternatively, if you currently have a high position in your market, focusing on the most recent click that led to a sale or lead will be more effective. These interactions are very specific, and the keywords could be a brand name or “leather recliner sofas in Toronto”.

The Challenges of Attribution

The challenge for online advertising platforms such as Google, Facebook and Microsoft is that marketing strategies are always evolving to meet the complex wants and needs of competitors and consumers.

For instance, although e-commerce is the dominant force in marketing, consumers prefer to purchase products in store. When an advertiser acquires an online click from a new user experiencing the brand for the first time, how can the advertiser attribute that click to specific results if the user goes in store to purchase that sofa?

Attribution is possible by tracking the user’s location since data can be collected from a mobile device showing they visited a store and completed a purchase. While this strategy is effective, it’s problematic because it compromises a customer’s right to privacy. Advertisers like 9thCO and platforms like Google strive to respect and protect that right.

We know that attribution is important to ensure marketing efforts are working, but how do we accurately attribute and measure our efforts without compromising our customers’ privacy and security?

Solutions to Attribution

9thCO leverages several strategies to ensure client campaigns continue to perform and their brands continue to grow, while accurately measuring the effectiveness of each campaign element.

Change Attribution Models

The first step in achieving accurate attribution is to change the attribution model used in Google Ads through the following steps:

  1. Log into your Google Ads account.
  2. Select Tools > Measurement > Conversions.
    image highlighting step 2 of the changing attribution process
  3. Select the conversion action which will change the attribution.
    image highlighting step 3 of the changing attribution process
  4. Click ‘Edit Settings’.
  5. Select the new attribution model under Attribution Model.
    image highlighting step 5 of the changing attribution process
  6. Click ‘Save’ and ‘Done’.

View Data through Attribution Models

Changing the attribution model in Google Ads will change how the data for your PPC campaigns are interpreted through that platform. However, to determine which channels (social, programmatic or organic) are driving real performance, you can change the attribution modeling in Google Analytics by following these steps:

  1. Log into Google Analytics.
  2. Navigate to the side bar and select Conversions > Attribution > Model Comparison Tool.
    image highlighting step 2 of the changing attribution modeling process
  3. Select the conversion for the data you want to view.
    image highlighting step 3 of the changing attribution modeling process
  4. Select the lookback window (maximum is 90 days).
    • It is recommended to use 90 days to get a full picture of conversions with high time lag. Use 30 days or less depending on your industry, product or service.
      For example; if your industry has a short sales cycle, maybe pet food or cosmetic products, you might want a shorter window
      image highlighting the location of the Lookback Window
  5. Select the attribution model that you want to test against the default attribution model.
    image highlighting step 5 of the changing attribution modeling process
  6. Change the date range to span at least one month. A larger date range may be required if there isn’t enough data in a one month period.
    image highlighting step 6 of the changing attribution modeling process
  7. View the table.
    image highlighting step 7 of the changing attribution modeling process

Analyzing the data in the table shows the effect of changing your attribution modeling from Last Click to Position-Based Attribution. The biggest shift is that conversions credited to Direct traffic drop by 13.66%, whereas display is credited for 37.75% more conversions. This difference exemplifies how improving the accuracy of your attribution can lead to strategies that have very different outcomes (i.e. stopping display advertising vs. adding more budget to display).

This tool allows you to view the data through alternate models, but Analytics cannot currently change the attribution it uses to process your data.

Conclusion

Attribution is a complex and crucial aspect of measuring the performance of your brand’s digital marketing strategy. Although accurate attribution is difficult, understanding the nature of attribution and the tools available to ensure your results are as accurate as possible will ensure you’re maximizing the ROI on your marketing dollars.

Add new comment

All fields required.

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.