3 Types of Marketing Attribution Software to Try in 2020

Despite (and because of) the many technologies marketers use, determining what drives a sale remains one of the messiest areas of marketing.

Figuring out which channel or touchpoint convinces a customer to make a purchase is as much of an art as it is a science.

To get to the bottom of it, marketers use attribution modeling. But attribution modeling comes in many forms. read on to find out how attribution marketing can help your business.

What is Marketing Attribution?

Using attribution models, marketers learn where sales triggers lie on the customer’s path to purchase. If an outsized percent of shoppers decide to buy online after visiting a physical location, then it’s a safe bet to say the in-store experience is owed a lot of credit for those website sales.

Some attribution models look at the entire buying cycle, while others drill down into specific channels. If a company wants to evaluate the effectiveness of various email marketing messages, for instance, it might want to look at the last thing customers see before purchasing by using a last-touchpoint attribution.

Of course, it’s not necessarily the last touchpoint that secured the sale, a challenge other attribution models take into account. No attribution model or type is perfect, which is why it can be challenging at first to use. Only through experimentation can marketers determine the right model and tool for them.

The bottom line is — every communication plan template worth its salt should contain marketing attribution models to help measure and track ROI.

Types of Marketing Attribution Software

Marketing attribution software comes in three broad types:

Multi-touch Attribution (MTA)

Multi-touch attribution tools are exactly what they sound like: software that teases out which of many touchpoints is responsible for the sale. It’s important to realize that while most MTA programs include some channel data, they’re designed to test specific touchpoints.

For instance, HubSpot’s reporting dashboard can evaluate touchpoints in five ways:

• First-touch: A first-touch model assigns all of the credit to the first webpage or digital asset that led a customer to your site. This model is great for understanding what brings people to your door.
• Last-touch: Last-touch attribution is the inverse of first-touch. It gives 100% of the credit to the last thing a customer sees before making a purchase. This model is great for evaluating bottom-of-the-funnel content like CTAs and landing pages, but it’s not much help in the start or middle.
• Last-interaction: The last-interaction model overlaps often with the last-touch one, but it differs in an important way — last-interaction modeling gives all credit to the last touchpoint that produces a conversion. In the last-touch model, the final blog post a customer viewed would get the credit, whereas it would not in the last-interaction model.
• First-and-last: Unlike the three prior models, the first-and-last attribution framework splits the credit. Automation platform Ontraport, which I’ve worked with in the past, recommends this model because the first and last touchpoints tend to stick out in customers’ minds.
• Simple-decay: The most complex of the five, the simple-decay model gives the most credit to the last touchpoint and progressively less to prior ones. The question is whether those weights correspond accurately to the customer’s own experience.

Any attribution model is better than none. However, MTA models can contain gaps and undue assumptions. A customer might have seen a television ad or received a friend’s suggestions that digital tracking could not account for.

Multi-touch attribution is also getting more difficult thanks to platform restrictions. Google, Amazon, and Facebook — the three largest ad networks — have placed limitations on cross-platform tagging. Pixel-based solutions face similar limitations due to Mozilla and Apple’s recent browser updates, which eliminate tracking pixels for speed and security reasons.

Marketing Mix Modeling

Marketing (or media) mix modeling takes a very difficult approach than MTA. Rather than use a tag or pixel to follow the individual user around the web, MMM uses multivariate regressions to predict just how much of an impact certain sales and marketing tactics had on customer behavior.

A top-down model, MMM takes into account historical data from online and offline sources. It attempts to account for external influences, such as seasonality, pricing data, and broader economic conditions. MMM is most popular at enterprise companies, which typically conduct it between once a year and once a quarter.

Because MMM requires a boatload of data and complex algorithms, the space is somewhat dominated by enterprise vendors with roots outside of marketing.

Multi-channel Attribution (MCA)

Multi-channel (sometimes called cross-channel) attribution is a blend of the MTA and MMM camps. Multi-channel attribution uses individual level data, but it attempts to evaluate certain tactics, like marketing mix modeling.

MCA seeks to paint a full picture of how a consumer’s online and offline activities lead to the sale. In the simplest MCA model, any channel that the customer accessed en route to the sale gets credit.

If a customer searches for a product on desktop, reads a blog post on mobile, visits a physical store, and then finally purchases after a social media referral, each channel gets weighted credit depending on the time that user spent on it.

Tools like tracking pixels allow marketers to evaluate channels like search, social media, and ad retargeting. Additionally, people-based attribution techniques like those developed by Branch tie channel and touchpoint data to individual customers.

Today, the challenge is connecting customers’ on-and offline behavior. How do marketers do it? The techniques fall into four buckets:

• Foot traffic: Mobile marketing companies use beacon technology or, in Foursquare’s case, decision-making engines, to determine where smartphone users are. Mobile device IDs are then matched with customer profile data to give credit to certain campaigns.
• Point-of-sale data: When a user makes a credit card purchase at a physical location, the credit provider works with a data company to give credit for the purchase to prior channels.
• Customer panels: Some companies ask users to opt-in via an app that transmits data about their location or offline behaviors back to the marketing team.
• Multi-source matching: Neither purchase nor location nor declared data tell the whole story. Companies that invest deeply in MCA cross-check data for the most accurate picture of which offline channels a customer might have engaged with.

For example, companies can use point-of-sale data to begin tracking customer behavior and use those numbers to influence a future campaign.

If you’re looking for another avenue for marketing attribution software, HubSpot offers attribution reporting in the Marketing Hub. Alternatively, if you’re looking to learn more about attribution reporting, check out our course on it.

As channels and touchpoints proliferate, marketers need better tools to determine which campaigns actually made a dent. While no approach is perfect, they each have characteristics that fit your business goals. Next year, consider combining them for a full view of where your sales and marketing spend is bringing in the most money.