Executive Summary
Recently, Banzai surveyed nearly 100 marketing professionals and senior executives to understand their use of marketing attribution and the value it provides. The findings show broad adoption of attribution models and widely held beliefs that they are effective, not only for strategic decision-making but also to increase C-Suite confidence in the marketing function overall. This study also reveals some “hidden” benefits of attribution that are less obvious than the need to report on the impact of marketing on ROI. That said, practical challenges in implementation continue to persist.
Marketing Attribution: (Still) Not Universally Used
About 2/10 of respondents do not use an attribution model in their reporting, despite increased pressure in the C-Suite to quantify the impact of marketing efforts in recent years. We expect this number to decline in the years ahead, with more standalone attribution tools in the market and increasingly sophisticated analytics that are native to marketing automation platforms.
Marketing Attribution: A Diverse Approach to Models
Among the vast majority of marketers who do use attribution models, there are many different methodologies in place. There is a lack of consensus on a “best” approach to modeling since the most appropriate model type depends on specific needs.
First touch attribution is used as a primary model more than any other approach (35% of respondents). The focus on first touch attribution reflects a broad trend of putting more weight on the initial point of customer engagement.
"First touch is, frankly, easier. And it at least provides us a window into how a prospect or account first came into our orbit."
- James Furbush, AccessOne
"First touch attribution provides valuable insights into initial brand interactions with prospects."
- Ciaran Connolly, ProfileTree
“A lot of our clients use first-touch because their focus is new user acquisition, and tend to rely heavily on lifecycle marketing to nurture users after they're first acquired to increase revenue and convert the initial audiences.”
- Nima Gardideh, Pearmill
Last touch attribution is the second most widely used modeling approach, adopted by 20% of respondents. Last-touch attribution is valued for providing insights into immediate conversion drivers, particularly at e-commerce companies.
“Using last touch shows that the paths used to get to that point have been effective in leading to the goal action/behavior.”
- Marianne Hynd, National Institute for Social Media/OMCP
13% of respondents use a different modeling approach or combination, suggesting that there is a small but sophisticated group whose methodology varies based on specific needs.
“We use a mixture of attribution models. It often depends on what it is we're measuring. For example, if we want to see how deals first interacted with our website, we'd use first click. When it comes to measuring the impact of our content efforts, we'd often gravitate towards a multi-touch model to assess how leads/deals have interacted with our content across their entire journey. We apply different attribution models to different scenarios and generate a better understanding of the customer journey.”
- Katie Holmes, Ruler Analytics
An Easier ROI Story Than In the Past
Nearly 8/10 respondents who use marketing attribution modeling reported that their choice of model makes it easier to prove marketing ROI. High satisfaction suggests a more mature state of marketing measurement than expected by many industry observers.
Not surprisingly, marketers at mid-sized and larger companies are more likely to feel that attribution has made their ROI story easier than those at small companies. The difference may be due to larger budgets (with more data for modeling reliability), more sophisticated tools or more experience adapting the modeling approach based on specific needs.
The Hidden Benefits of Marketing Attribution
Respondents were asked about the value of using marketing attribution, and type of model used. The benefits go beyond the classic use case of “better allocating marketing resources.” Here are the top ones cited.
Improved Visibility and Data Richness
Attribution provides more visibility into customer journeys and gives the ability to track campaigns that reach customers at specific times. Some respondents reported that multitouch models in complex B2B journeys allow for a more detailed understanding of the funnel.
Sales Cycle Insights
Attribution models also provide clarity on the impact of different channels on the sales cycle, helping to understand what drives qualified opportunities and conversions. Users of models such as U-shaped or linear attribution appreciate the full-picture view these models provide, from the generation of a lead to the nurturing and closing stages.
"We have a long (1 to 2 year) sales cycle and understanding all of the touch points... are important.
- Bethany Walsh, Alkymi
“We can more clearly see what, if any, impact a particular channel had on the sales cycle - knowing that it naturally takes multiple channels and touchpoints to create and convert opportunities.”
- Kenneth Burke, Text Request
Highlighting Indirect Revenue Contribution
Many respondents mentioned that attribution helps communicate the value of marketing efforts to leadership and stakeholders. For example, attribution can make it clearer to C-level executives that marketing activities can contribute to the sales process, despite not driving direct revenue.
"It makes it clearer to C-level and Finance that although not every marketing activity will drive direct revenue, they all contribute."
- Anna Kocsis, Ungate
Client Retention and Demand Capture
Identifying the most effective channels for client retention.
Flexibility and Granularity
AI and custom attribution tools allow for tracking at the visitor level and linking marketing touchpoints directly to leads and deals, with flexibility to switch between models for different business goals.
Not a Walk in the Park. Challenges Exist.
Despite high satisfaction with attribution overall, marketers continue to be vexed by a number of challenges. Below are the top themes that emerged.
Attribution Model Bias
There is a risk of focusing on metrics that favor the attribution model rather than the actual effectiveness of marketing efforts.
“You're judged or rewarded based on the model, then you start having wrong priorities."
- Yulia Olennikova, N.Rich
"A first-touch model doesn't value the touches after the first."
- Joe Kevens, PartnerStack
We advise marketing professionals to avoid the trap of over-investing in methodologies that serve narrow interests and are not aligned with true business drivers. It is best to gain alignment with transparency, share the pitfalls of attribution and even use it directionally if needed.
Incomplete Customer Journey Representation
Attribution models often fail to capture the full scope of customer interactions over extended periods. Although “full customer journey” was cited as a benefit, there are some practical challenges with implementation.
"It really doesn't paint a genuine picture of all the touches that happen when a deal takes 18 months to close."
- James Furbush, AccessOne
Related to this is the difficulty of tracking users who are not direct buyers but do influence purchases.
Cross-Channel and Cross-Device Attribution
Tracking and attributing conversions accurately across different channels and devices is challenging. For example, usage of different browsers by an individual can muddy the waters.
"The biggest issue for us is cross-channel attribution."
- Brock Murray, seoplus+
“Phone call first-touch attribution is the most difficult, as phone calls can be tricky to track if proper number swapping and tools such as Call Rail are not implemented properly.”
- Daniel Lynch, Empathy First Media
Tracking Complexity and Manual Effort
The process of tracking attribution can be labor-intensive and complex.
"It shows all the data points, and sometimes we don't need all of it, we just want to understand where the conversion comes from."
- Alex Milner, AEMedia
Privacy Regulations and Data Limitations
Changes in privacy laws and data tracking limitations are making it more difficult to attribute performance.
“Cookie consent and GDPR-related tracking limitations making it harder and harder to attribute performance."
- Anna Kocsis, Ungate
Offline Interaction Attribution
Accounting for offline interactions within primarily online attribution models presents a significant challenge. This is particularly important when in-person events are a large component of the marketing mix.
Skills and Talent Needed
Despite more intuitive UI/UX in marketing attribution tools, success still is all about the data. Specialization and skills are required for proper set up and management that are lacking in some organizations.
“Marketing attribution requires a combination of marketing operations and data analysis. Hiring for this more specialized skill set tends to come after generalist roles are hired and beginning to scale channels.”
- MK Getler, Loop & Tie
Implications for Marketing Professionals
Manage expectations
It's important for marketers to be transparent about the capabilities and limitations of marketing attribution, so findings can be put in context with colleagues. In doing so, you’ll inspire confidence, gain trust and build deeper relationships with senior management.
Align Your Model with Organization Goals
Marketing leadership tends to gravitate towards first touch models, which gives more weight to the beginning of the customer journey. Marketers should consider how this model selection aligns with broader organizational goals.
Small Companies: Build a Strong Foundation that Can Scale
Directionally, larger companies report a higher efficacy of marketing attribution models. For small companies, this suggests preparing to build fundamental capabilities first so attribution methods can be scaled with business growth.
Give Context When Reporting Results
The qualitative responses in this study highlight the complexity of attribution in diverse customer journeys. Marketers should consider these insights when designing their tracking systems and attribution models, and provide appropriate context when sharing the results.
Respondent Profiles
The profile of respondents is detailed as follows: