Last week, VettaFi Investment Strategist Cinthia Murphy moderated a discussion on marketing attribution featuring Ozzie Solares, global head of integrated media & analytics at J.P. Morgan Asset & Wealth Management, and Mitchell Home, head of digital products at VettaFi. The conversation centered around how to structure effective attribution frameworks. 

View  Attribution in action: Driving growth with media.

The challenge

In an early poll, attendees were asked how confident they were in connecting media spend to results. Forty-seven percent of respondents said they were not very confident, and 41% said they were somewhat confident. Only 6% said they were very confident. 

“This is a challenging space,” said Solares. “The landscape isn’t exactly doing us any favors, and it continues to evolve.”

Murphy added that the abundance of data available today compared to prior years makes it seem like it could be easier to connect dots and find accurate marketing attribution. While the volume of data creates opportunities, it can also create issues. “The landscape itself is increasingly fragmented,” Solares added.

Home noted that media spend can often take time to show value, even though asset managers typically look at spend monthly. This means marketers often need to “justify” their existence. Without good attribution frameworks, this can be particularly challenging, even though the right kind of spend can have a compounding effect over time. 

Where to start with attribution frameworks

“Before you talk about what your attribution framework is going to be, you need to take inventory of yourself and your business,” Home explained. Determining what data matters and why, as well as what core questions you want to understand, is critical to setting up a good attribution model that fits your organization. 

The first thing to do is take stock of your data. Do you have access to the right data? Do you have the right talent and infrastructure in place to effectively leverage it? Once addressed, you should align all stakeholders through clearly defined campaign goals and objectives. All departments should agree on the key metrics. From there you can organize your data and assets. Institute data processes and standardize the tracking of campaign assets.

It is worth noting that looking at the broader data landscape is also important. There are regulatory challenges and layers of transparency when it comes to different data sources.

With a foundation in place, you can determine your attribution model based on data maturity and goals. 

Models to consider

As organizations look at their data and their needs, they need to figure out how they are going to attribute their media spend.

In terms of complexity, the most straightforward is a single-touch method, which gives 100% credit to one source. For example, someone who ultimately converts might do so after clicking on an ad. That ad gets the full credit for the conversion in a single touch model.

Multitouch is more complicated, as it distributes credit across sources depending on pre-set criteria. A data-driven model goes a bit deeper to calculate the precise influence and contribution of each touchpoint. “Data is king here. You have to really understand what your data capabilities are,” Murphy added.

Solares noted, “As you mature, you are going to find yourself using a myriad of models.” He continued, "Each model lends itself to a different use case.” Understanding the entire scope of your interactions is critical, according to Solares. “It is easy to give credit and subsidize lower funnel,” Solares noted. Multitouch attribution can help give more credit to the upper funnel.”

Overall, you’ll need to select the attribution model that best fits your data maturity. Given the limitations of the single touch model, Solares pushed for firms to explore multi-touch and algorithmic models, though it is worth noting that more sophisticated approaches take time and expertise. Still, the commitment of these resources will result in a clearer attribution picture that makes it easier for leadership to understand the value of combined upper- and lower-funnel efforts.

Fragmentation & how to set yourself up for success

Home discussed that even as technology has advanced, data availability and fragmentation constantly changes. He pointed to Twitter as an example, where data used to be transparent and accessible, which isn’t the case anymore since the app changed ownership.

Solares noted that marketing mix modeling is more forgiving when it comes to the availability of data, as it uses aggregate sales and marketing data. While the multi-touch attribution model focuses on a bottom-up analysis of individual touchpoints in a customer journey, marketing mix modeling is more top down, analyzing aggregated data and leveraging online and offline sources. This approach accounts for seasonality, competition, and other factors while a multi-touch attribution model is limited in how it considers external factors. Multi-touch can be useful for short term ROI and tactical optimization, but a marketing mix model will be more useful for strategic planning and budget allocation, as it is more robust and broad in scope.

How to navigate leadership conversations

The session then discussed how to approach attribution conversations with your leadership team. At the end of the day, businesses have limited resources and are looking for the best return on any investment. As you develop your business case for best attribution practices, this is important to remember. “When engaging with sales, they are self-interested, because they want to be able to drive revenue,” Solares said. This makes sense, as sales departments often get paid purely out of what they mark off as a sales success. Marketing departments, meanwhile, also need to show their impact, which could result in competing interests when it comes to attribution. “The key thing is marketing is a partner to sales,” Solares added.

Home concurred: “You don’t want to be thinking about it as competing with sales for attribution; you want to be partnering with them.”

Conversations around attribution can be challenging, but they boil down to storytelling and good data. Remembering the decisions CEOs and leaders are making can help better tell the story. It is the CEO’s job to look at the landscape and the available resources, and allocate for the biggest impact. “I love the concept that this all comes down to storytelling in the end,” Murphy said.

There are often multiple pieces of marketing in play, multiple data vectors, and a wealth of information to sort. Good attribution and marketing requires lots of little things to all go right together. It can be easy for something simple to go awry and provide bad (or incomplete) information, given the amount of balls in the air. ”Don’t let perfect be the enemy of good. Some attribution and some confidence is better than none at all,” Home added.

Solares emphasized the importance of acknowledging the realities of your data and your organization's internal communications. “Spending a lot of time on the foundational stuff pays dividends down the road,” shared Solares.

Key takeaways

As asset managers look for attribution structures that makes sense for them, here’s what to remember:

• Foundational steps pay dividends. Taking the time to be 100% honest about the data you have and the data available to you, coupled with understanding what your organization needs, is critical for success. Attribution is complex. Taking time to plan and implement is key.
• Data is part of storytelling. As you look to get buy-in for marketing programs or initiatives, remember that data and storytelling need to be intertwined. Don’t neglect one for the other; use them in tandem to get leadership on to your page.
• Attribution constantly evolves. Data sources change over time, as do organizational needs. Your attribution strategy needs to constantly evolve, and your data needs to constantly mature.
• Other departments are partners, not rivals. As every department competes for limited resources, remember that the entire enterprise grows or fails together.