Determining the best attribution model for your ad campaign

This post is a guest post from Lucas Sommer, director of marketing at LeadsRX. Over the last decade, Lucas has worked with hundreds of organizations and marketers to set up their attribution and help them get useful insights from their data. As director of marketing for LeadsRx, he currently focuses his energy on understanding and optimizing their cross-channel stack as well.

Having choices is a great thing. Sometimes, when you have options, choosing one over another is a great luxury. When choosing the right attribution model for your ad campaign, making the right selection is more than a matter of luxury. It’s a matter of assessing whether your marketing dollars are being spent wisely or thoughtlessly. Marketers that use attribution models and are familiar with how they work may still experience some consternation when it comes to leveraging the one that provides the greatest insight.

When it comes to selecting an attribution model, making the right choice really boils down to the specific ad campaign you’re running and what your objectives are. Before we begin identifying campaign types and objectives, let’s first discuss what an attribution model is.

What is an attribution model?

Determining what prospect interactions lead to a final sale can be compared to searching for the holy grail. You know it exists. People have talked about it. Finding it is another story. Marketers know that a sale has occurred, but what led to that final conversion? This is where attribution comes in.

An attribution model is a means of identifying those specific interactions or touchpoints prospects experience with your company on their way toward making a purchase. As we all know, engagement may be a positive or negative experience. In an attribution model, the hope is that all interactions are positive and ultimately propel the prospect forward to making a buying decision.

There are generally two types of attribution models: a single-touch model and a multi-touch model. Both offer benefits, depending on the type of campaign being run, the length of the sales cycle, and the type of product or service being offered. With that said, at least one model is typically suitable for an ad campaign being run.

Choosing the right model

There are several things to consider when deciding which attribution model to use for your campaign. Those things include determining your marketing objectives, your budget, your website content, and other channels. At the end of the day, marketers must understand which attribution model fits the desired insight.

First-click model

First click attribution is a single-touch model that emphasizes the first touchpoint a prospect has with your brand. That touchpoint could be a Facebook ad that is eventually followed by several other interactions. While there could be one additional interaction or several additional interactions, the first one receives all the credit for the sale.

When first-click works best

While possibly missing many other interactions contributing to a sale, the first click model can be a powerful approach to finding new customers. For some campaigns, understanding which action leads to a new prospect entering the top of the funnel is most important. Refocusing more resources to campaigns that deliver on this notion, such as social media ads, eBooks, etc., can truly provide the needed insight.

Last-click and last non-direct click model

Last-click attribution works the opposite way a first-click approach does. Rather than assigning all credit to the first touchpoint, the last-click attribution model allocates all credit to the last interaction before the sale.

A last-click model can be direct or non-direct. When engagements are direct, customers will type in your company’s URL and make a purchase. There may have been other touchpoints that lead to awareness, but when the customer was ready to buy, they went directly to your site. For non-direct engagements, a customer may be referred to your site via search results or a partner site. They arrived at your site, but they did not type in the URL.

When they work best

The benefit of using either of the last-click models is understanding which interaction leads to the actual conversion. The direct model works well with campaigns that use discounts, trials, or specific calls to action that lead directly to a purchase. The non-direct model works well with campaigns that are designed primarily for awareness and brand building.

Linear model

There are campaigns where marketers may want to give every touchpoint an equal share of sales credit. In those instances, a linear attribution model works well. When searching on a related topic, your prospects may find your company, subscribe to your email newsletter, or come across a LinkedIn ad. After each of those interactions, a prospect may find a reference to your company on a partner site, click on your website, and make a purchase.

When linear works best

The linear model works best when you want to measure the effectiveness of your entire marketing strategy. This can be done through brand awareness and other campaigns where a single interaction is not the driving factor for the sale.

U-shaped model

The U-shaped or position-based model attributes 40% of the sales credit to both the first and last touches of a campaign. The remaining 20% is divided evenly among the other touchpoints in the journey, whether it be one more of five more.

When it works best

The U-shaped model is ideal for campaigns where you want to know what works well for the top and the bottom of the funnel. In other words, the engagement that garnered the initial attraction and the one that occurred last to close the sale is best for this model. When campaigns rely heavily on keywords, marketers can determine what engagements opened the door and which ones sealed the deal.

Time Decay model

The model that gives the most credit to those touchpoints that occur closer to the sale is the time decay attribution model. Initial touchpoints are less important, while those that lead up to and occur closer to the conversion are deemed more important.

When time decay works best

Campaigns that assume a long sales cycle, and require educating the prospect, work well with the time decay model. Often, awareness campaigns, partner campaigns, and activities that build your brand over time are ideal candidates for the time decay model. The time decay approach can also help marketers better understand the true length of their sales cycle.

Data-driven model

Data-driven attribution is perhaps the most sophisticated model of the ones we’ve reviewed. This approach relies on using data (whether derived from A.I. or other means) to determine each touchpoint’s importance. The data-driven model makes no general assumptions but relies on each engagement’s performance to determine attribution weight.

When data-driven works best

The data-driven attribution model is best suited for more comprehensive campaigns where a relatively large budget can accommodate enough data to analyze each touchpoint’s effectiveness. Whether the campaigns are focused on clicks, awareness, or non-direct attribution, having enough data to drive detailed analysis works well with data-driven attribution models.

As we stated initially, choice is a great thing to have, especially when you know why you’re making a choice. Using the right attribution model for the right campaign will go a long way in giving you the information you need to implement the most effective campaigns for the best ROI.