Inside Airbridge
How Airbridge Innovates Itself as an MMP with MTA and Marketing Mix Modeling
July 27, 2022
Minah Lee

This July, Airbridge successfully hosted the Modern Growth Stack 2022 (MGS 2022) adtech & martech conference in Seoul. We brought together around 40 industry leaders and 2,000 attendees to share growth strategies and tactics in the era of digital products.

Out of all the insightful sessions, here’s a recap of “Unified Measurement Stack - MMP, MTA, MMM and Lift” by our Chief Strategy Officer, Yongchoul Han. With the advertising and marketing market changing like never before, how should marketers measure performance?

Emerging challenges in marketing measurement

Privacy changes restricting user data collection

Apple’s iOS 14.5 update included privacy-first features such as the App Tracking Transparency framework. Since then, all iOS apps have been required to ask users for permission to share their data, meaning their access to the Identifier for Advertisers (IDFA) keys has been limited. In the meantime, Google announced plans to phase out third-party cookies in its Chrome browser by 2024. Privacy policies like the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA) also pose difficulties for user tracking, targeting, and attribution.

Limitations of last-touch attribution

The last-touch attribution model gives 100% of the credit for a conversion to the single last touchpoint in a conversion path.

It has been one of the most popular attribution models as it is the more simplistic, cost-efficient and easy-to-implement model. However, it doesn’t consider earlier touchpoints that could have had incremental impact, risking giving incomplete conversion insights. The fact that it is a single-touch attribution model also makes it more difficult to draw insights for multi-channel marketing campaigns.

How Airbridge is responding to the challenges

Multi-touch attribution (MTA)

With Airbridge, you can measure the true incremental effectiveness of marketing in influencing user conversion. This is possible because our MTA model reveals the essential touchpoints and how they each contribute to the eventual conversion.

Airbridge leverages extensive partnerships, including the MTA partnership with Meta, to gather data for not only the last but also the earlier touchpoints. Our Privacy-Enhancing Technologies (PETs) allow us to measure growth while ensuring users privacy.

Marketing Mix Modeling (MMM)

MMM is a statistical analysis method for estimating return on investments in different marketing channels. It’s a proven tool to optimize marketing budget allocations and boost growth.

Airbridge uses regression or Bayesian models to calculate marketing incrementality. Powered by advanced machine learning and statistical methodologies, our model can predict future conversions based on given inputs in each campaign. It also helps companies remain compliant in the era of digital privacy by using aggregated data instead of individual user data.

Unified measurement stack: an integrated approach

Airbridge’s Unified Measurement Stack

Han recommends building a unified measurement stack which combines MTA, MMM and last-touch attribution model. By flexibly working with all three of them depending on the changing situation and the characteristics of your digital product, you can achieve the most balanced view of measurement.

Finishing up with observational and experimental analyses

Airbridge’s unified measurement stack - MMP, MTA, and MMM explained above - is based on an observational model. It is a time-efficient method that gives you full control over campaign execution and performance measurement. However, there are also limitations as statistical modeling uses assumptions to make predictions about the real world, unlike that of the experimental model.

While the experimental model is certainly more accurate as it is an actual experiment, it involves a higher cost - designing and takes longer time to conduct. You also do not have access to the results until the experiment is completed.

Hence, the ideal approach is to use the observational model for day-to-day analysis and conduct experiments every three or six months to check how the results are alike. In other words, the two methods complement each other and it is the marketer’s job to be adept at working with both.

Korean health brand Dano ran Airbridge’s incremental MTA study as well as a Meta conversion lift test. Compared to the lift study, the MTA results were only slightly lower - showing just 2% fewer app installs and 3% fewer purchases - giving Dano further confidence in its campaign, and convincing the team to have greater trust in Airbridge’s MTA model. Check out the full case study here.

Airbridge’s approach to marketing measurement

Airbridge uses a last-touch attribution model as other mobile attribution solutions do, but also works with MTA and MMM to measure true effectiveness of marketing campaigns.

By the end of his session, Han emphasized that marketers should not solely rely on a single attribution model and that Airbridge’s goal is not just to measure, but to contribute to marketing success through effective and accurate measurement.

If you want to have a deeper discussion with our team of experts on marketing measurement, contact us today.

Want to get more insights?
Get a mail whenever a new article is uploaded.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Check out our all-in-one package that fits every stage of your growth.
Minah Lee
Product Marketing Part Lead
Minah is Airbridge's product marketing part lead. She creates content, plans various marketing campaigns, and collaborates with the growth of teams, customers, and partners.
Subscribe to the newsletter for marketing trends, insights, and strategies.
Get a mail whenever a new article is uploaded.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.


准备好加快 App 的增长进程了吗?只需与我们聊聊,即可获取所需的专业知识和工具。
加入 20,000 多名领先 App 营销专业人士的行列,获取每周洞察
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.