Thoughts on Marketing Attribution

Published: 9/16/2022

Multi-Touch Attribution (MTA) – collects individual, user-level data for addressable (trackable) media and conversion events to determine each media event’s impact on a customer’s path to conversion. Because MTA requires tracking and connecting all media at the user level, it does not account for non-addressable media, like print, radio, and traditional (linear) TV, which is not tracked back to individuals.

Media mix modeling – collects aggregated data from marketing and non-marketing sources over a multi-year historical period, also factoring in external influences such as seasonality, economic data, weather, and promotions. The data is then used to develop a demand model which quantifies the historical contribution of each marketing and non-marketing input to a business outcome, like sales or conversions.

Using MTA and MMM for cross-channel attribution has proven to be difficult for reasons including:

• Many platforms like Amazon and Facebook are walled gardens and inaccessible to third-party tracking of impressions.
• Identity resolution across media platforms is relatively low.
  • Services like LiveRamp (IDL product), TransUnion, and Neustar provide a solution, while The Trade Desk has offered Unified ID 2.0 (UID2) as an open-source solution.
• Cross-device tracking is complex, and match rates are meager.
• Instrumenting a tracking infrastructure by a third-party measurement provider has proved to be fraught with breakage and data leakage.
• Both are extremely time-consuming to implement and maintain and require experience in data science and analytics.
• Media Mix Modeling (MMM) is typically two quarters behind which misses major sales cycles and introduces bias by giving the most credit to the strongest impression-based channels.

Consumers are becoming increasingly aware of data collection in their day-to-day lives. According to Marketing Evolution, 97% of consumers are concerned about protecting their personal data. This increased focus on privacy negatively impacts a marketer’s ability to collect and derive insights at a rapid pace and makes some of their previous marketing technology investments obsolete.

Cookies power conversion tracking, audience targeting, and frequency capping on desktop and mobile web. Apple and Mozilla’s decision to block tracking in Firefox and Safari does not impact Advertisers as much due to their minority market share. Still, Google’s decision to phase out cookies, in chrome, on their 65% market share by 2022 sent shock waves through the ad tech industry. These changes by Apple, Firefox, and Google have created uncertainty within the industry, and it is still unclear how advertising cookies will work in the future. These changes have impacted companies’ ad-based revenues like Apple, shifting ad spending from Facebook back to Apple Search and their new DSP. This change also has a significant impact on Facebook’s Ad revenue. Facebook is retooling their ad targeting engine to accommodate changes limiting the effectiveness of 3rd party cookies and pixels.

During this debate, marketers and their technology partners have devised various approaches to overcome the obstacles to tracking consumer behavior:

• Respect consumer privacy and use anonymized data to inform marketing decisions.
• Build a system to overcome consumer privacy by creating a universal ID assigned to every device and user.
• Preserve past models by taking what is known and extrapolating what is unknown—creating synthetic transactions to fill in the blanks.

At Alembic, we respect consumer privacy, take anonymous client data, and turn them into insights. Other approaches rely heavily on a “leap of faith” that what is “unknown” can be known by extrapolating from what is “known” with certainty. The biggest drawback of this approach, like MMM, is that it treats all consumers as a monolith (equal LTV). In reality, everyone responds differently to various advertisements, and not all traffic is equally as valuable. Creating synthetic transactions to fill in “unmeasurable data gaps” based on what is measurable reduces the power and effectiveness of the data and increases the margin of error. Small sample sizes of what is “known and measurable” yields higher variability, which may lead to bias. Furthermore, it is estimated that up to 25% of ad clicks are from bots, competitors, and click bombing. These variables combine to limit the effectiveness and believability of synthetic transactions.

What can Alembic provide?

Alembic provides insights into the impact of marketing events on data channels. Alembic detects events that generate a material change and then correlates the detected events to provide attribution touchpoints. These touchpoints drive actionable insights across Engagement, Activation, and Adoption phases. Touchpoints or detected events are like marketing breadcrumbs that reveal a correlation between actions and insights when connected. The connections between actions and insights allow marketers to make better decisions by understanding the attribution tapestry that detected and correlated events create. Anonymized aggregated data from first-party linked data channels allows Alembics customers to avoid the cookie apocalypse. 

Alembic takes a holistic approach to attribution. We see all traffic and the impact derived from the traffic on marketing efforts and ROI. We see all traffic and the impact derived from the traffic on marketing efforts and ROI. We can provide an understanding of a touchpoint, event attribution, and impact on your marketing tapestry. This approach allows us to be a leading indicator rather than a lagging indicator.

Unlike the losing battle that all MTA companies are fighting, trying to build an attribution picture from continually degrading technology – cookies and tracking pixels – Alembic pulls your first-party data, Media (News, PR, Radio, TV, and Podcasts), web, application, and customer enriched data into a real-time mixed media event correlation platform that can provide actionable attribution insights.

Conclusion

Alembic suggests running your current MTA/MMM platform side-by-side to compare. As the market shifts toward old-school attribution models, you can decide which models to embrace and which to stop supporting. Want to learn more? Book a Demo today.

Sources

1. https://www.marketingevolution.com/knowledge-center/data-privacy-issues-in-data-driven-marketing
2. https://blog.mozilla.org/en/products/firefox/firefox-rolls-out-total-cookie-protection-by-default-to-all-users-worldwide/
3. https://www.clickcease.com/blog/list-of-click-fraud-statistics/