Multi-Touch Attribution: Who Really Benefits from MTA?

• Multi-touch attribution helps marketers understand the role every customer engagement played in driving that purchase, sign-up, submission or visit
• Multi-touch attribution is most valuable for brands with complex consumer journeys that end with a specific action
• There are three types of multi-touch attribution: Even or Linear Attribution Model; Time Decay Attribution Model and Algorithm Based/Custom Attribution Model

---

When analyzing your campaigns, how do you assess the value of each of the different touchpoints along the consumer journey? After all, understanding what led to a conversion generally isn’t as simple as looking at the last action a consumer took before clicking “buy.”

This is what Multi-Touch Attribution (MTA) sets out to accomplish. However, what might seem like a simple undertaking is anything but. Plenty of marketers – even those with years of experience – can get thrown off when trying to understand exactly what MTA can accomplish, and what it can’t.

Below, we’ll take  multi-touch attribution back to basics, and help you decide whether it is valuable for your brand as you strive to prove the effectiveness of your advertising tactics.

A brief introduction to multi-touch attribution

Multi-touch Attribution, or MTA for short, means so many different things depending on who you ask, but when we think about it at Viant, we try be conscious of the what it’s actually like seeing ads in the wild. What is the actual experience our customers have before ultimately making that purchase?

As marketers, we all know that consumers aren’t doing U-turns in the middle of the street to check out the furniture store they just sped by after seeing our roadside billboard. That clever commercial during a playoff game isn’t realistically inspiring customers to pick up their iPhones that very moment to call about a new auto insurance quote. We understand that advertising is more subtle than that — the consumer journey is complex and nuanced.

What is multi-touch attribution?

From the very first impression to the final conversion action and all of the touchpoints in between, multi-touch attribution considers each and every engagement a consumer has with a brand, assigns a value to them, and helps marketers understand the role each played in driving that purchase, sign-up, submission or visit.

Essentially, MTA is very much like the scorecard in a round of golf. Every golfer plays 18 holes in a round, but each shot, hole, lucky bounce and bad break adds up to a final score. The scorecard tells us what kind of round (consumer journey) each golfer (consumer) had during their Sunday afternoons and tells us how and where each golfer can improve (optimize).

What’s the difference between first-touch, last-touch and multi-touch attribution?

In their simplest terms:

• First-touch attribution is the model that gives 100% of credit for a purchase to the very first impression (most often) a user is exposed to in their journey to converting.
• Last-touch attribution, you guessed it, is the exact opposite of first-touch in that it’s the very last impression, often click, that’s allocated with 100% of credit for the purchase. Certain players in digital advertising capitalize on this immensely.
• Multi-touch attribution is an attempt to put a bit more thought behind the consumer journey. Rather than given 100% of credit for a purchase to any single exposure, MTA considers the whole picture. Customers are exposed to thousands of ads every single day. MTA seeks to make sense of all of that chaos and prove which marketing tactics actually drive performance when given a fair shot.

When should you use multi-touch attribution?

The most important thing a marketer should consider when evaluating MTA as a potential measurement tactic is to truly ask themselves: What is the goal of my campaign? Is the focus on building brand awareness or driving site traffic? Or is the marketer’s goal more performance-focused, where sales or ROI are the most important indicators of success?

Marketers seeking to inspire users to take a specific type of action are the ones who will get the most out of MTA as a measurement solution. This isn’t to say that brands focused on building awareness wouldn’t get any value from MTA, simply that there are simpler metrics and measurement tactics that can prove out the same results for those marketers.

But for those with complex consumer journeys that end with a specific action like a purchase, MTA can offer a wealth of knowledge beyond basic metrics like click-through rate, cost per acquisition or even sales/revenue driven.

Why is multi-touch attribution important?

Answering this question and helping our clients see the value of each model is the crux of our measurement solutions and, truly, the entire concept of MTA. There are certain players in the digital advertising measurement space who thrive on marketers’ misconceptions of the various attribution models and their inherent value to different types of advertisers. Without naming any names, it’s typically those who dominate a specific channel in digital advertising, like social media or paid search, for example.

Even worse for marketers is when those players who own all elements of the digital advertising process — ad serving, reporting, planning and buying, web browsing, etc. — set themselves up for constant success when considering something like last-touch attribution simply because they own the platform that is the digital representation a user’s final action before making purchase or converting. Paid search and social media are almost always over-represented and over-credited when considering a last-touch/single-touch model.

Multi-touch attribution is the best solution to give answers on which channels are actually driving performance for marketers.

Why is attribution modeling difficult?

Modeling is such a challenge because of the amount of nuance marketers must consider. How is a marketer to know how much credit should be allocated to a CTV impression vs. a paid search click? These questions and the amount of work it takes to answer them are why the majority of marketers opt for their ad server’s recommendation to use last-touch as the attribution model of choice. And why not? That ad server almost always stands to benefit from any marketer using last-touch as their default selection.

It’s a real challenge to sit down and take the time to think through how much credit should really be assigned to CTV vs. paid search vs. paid social or even a simpler concept like an impression vs. a click. But to those marketers who are willing to put in the time and work with a partner like Adelphic, the benefits are there to be realized.

What’s required of a campaign in order for MTA to get up and running?

There are a number of boxes that need to be checked in order to launch a successful MTA-measured campaign. Here are the three that, when leveraging MTA are the most important:

1. Ensuring Viant’s site-tracking pixels are placed in advance and are aligned with a customer’s on-site journey: without having pixels properly placed, there will be no site-side data to connect the dots from first impression to final action (conversion) for each customer. This goes for any DSP. Often times there are separate teams who manage a client’s website tagging, so involving that team well ahead of launch is a must.

2. Tagging all media within the campaign that should be measured by MTA: If running media in Adephic, there’s no additional tagging work required because that media will be tracked at the individual impression level directly within Viant’s logs. If a marketer wants to track activity running in other DSPs or with other partners, that media must be tagged well in advance in coordination with Viant’s Technical Account Management team.

3. Finally, while less about the logistical setup checkboxes, marketers will get the most out of MTA if they have a firm understanding of their campaign goals and what success looks like in addition to a solid strategy around how to use the data MTA produces: We’ve found that the clients who get the most out of MTA are those who have a very firm sense of how their media tactics should perform (e.g., Display vs. Video vs. Social) and how they should optimize once statistical significance is reached.

Who shouldn’t be utilizing multi-touch attribution ?

In the same vein as how to know whether MTA is an appropriate measurement solution, those marketers focused more on driving brand awareness or simply driving site traffic are the first examples that come to mind when I think of who shouldn’t be using MTA.

For example, a marketer focused on driving homepage visits for a new CPG brand would get just as much out of their reporting solutions if they use click-through rate and cost-per-acquisition as their primary KPIs. Those two metrics, both of which are offered in all major ad servers, will easily tell them which media partners and tactics are over-performing, and which are under-performing.

Marketers without the ability to place pixels on their clients’ sites or those who are managing a a small/specialized component of the larger brand media strategy should also exercise caution when considering MTA. Pixeling is the primary requirement to activate an MTA measurement campaign, but MTA reporting also requires a strong volume of activity in order to produce meaningful results.

What are the different types of MTA?

You may get a different list of the MTA types, depending on who you ask! While there may be some slight differences in the names and descriptions of each type, the key concepts are largely the same and fall into a few main groups:

1. Even or Linear Attribution Model

The most straightforward attribution models in the MTA space fall under this bucket. These models simply assign even credit to each touchpoint in a consumer’s journey to help marketers understand performance beyond first and last-touch attribution. How the credit is assigned can vary slightly, but ultimately the majority of touchpoints receive the same amount of credit for a given conversion.

2. Time Decay Attribution Model

Slightly more complex than Even Weighting MTA models but based on more mathematical weighting, credit is allocated based on where in the consumer journey each touchpoint took place. More credit is assigned to interactions that occur closer to the conversion event.

3. U-Shaped Attribution Model

The U-Shaped attribution model assigns more credit to the first and last touchpoints in the customer journey. The remaining credit is distributed evenly among the remaining touchpoints.

4. Algorithm Based/Custom Attribution Model

Those marketers with dedicated Data Science resources are the only ones who will likely utilize these models, given they are typically based on in-depth analysis incorporating years of media activity to prove the specific value of the various types of touchpoints a customer can experience during their journey. For example, these models can assign mixed credit across a variety of media tactics based on impressions, clicks, views and other actions.

What stands out about Viant’s MTA capabilities?

For MTA to be feasible as a concept, marketers need to be able to tie two key sets of data together: media activity (impressions, clicks, etc.) and conversion activity (pixel fires). Without the ability to link a customer’s purchase to the impressions they were exposed to, MTA simply isn’t possible to perform.

Leveraging Viant’s device graphs, registered user database and Household ID, our MTA solution not only connects the dots between media and pixel logs, it provides attribution reporting with minimal gaps in the data. Viant’s solution is not beholden to a single digital identifier like cookies to tie media exposures to conversions, which means that the consumer journey isn’t lost when tying log-level data together.

With the capability to perform attribution using multiple identifiers, Viant’s MTA offering is also fully cross-device, unlike the majority of alternative solutions in the space. This means that marketers are able to evaluate, understand and optimize their campaigns based on a truly holistic picture of their campaigns’ performance with confidence.

Still want to learn more about whether MTA is right for you?

Viant’s guide Measurement Tactics for a Cookieless World will help you understand how you can prove your media dollars are working – now, and when cookies are phased out.