Need Data Science guidance? Talk to our consultants!
Softcrylic was approached by a client from the travel industry (Client) for assistance in analyzing their customers’ shopping behavior on their website. Our data science & analytics team analyzed the available data and came up with an approach to find high-value audiences for the Client. The goal of this project was persona-specific messaging and using audience modeling, we were able to determine high-value audiences for the client that were actionable and easy to target.
Watch the full video below (or read the main highlights) to fully understand how we were able to leverage audience modeling to meet the intended.
Main Highlights:
We have crafted our audience modeling process by working with numerous clients over the last few years. Our approach to audience modeling ensures strategists and analysts have all the necessary information to identify high value customer segments, target them via experimental design, and measure campaign performance.
The analysis workflow happens in four stages: Data aggregation/preparation, Exploratory Analysis, Modeling, and Persona Analysis.
We place a strong focus on GIGO (garbage in, garbage out) philosophy when working with huge data. This is done to ensure that the data is properly cleaned before the machine learning process starts.
Data Aggregation and Machine Learning: This is where all the data sources are brought in together from various channels and we use ML to identify audiences.
The strategy phase is where we analyze results to define personas and identify the audiences for testing.
The machine learning process we use at Softcrylic is unsupervised learning which means that the results are not pre-determined and would require human interpretation which comprises finding the contrasts between groups and parlaying the sets of contrasts into “personas.”
The personas can be further refined by incorporating additional data points by considering how the persona maps onto the business objectives.
3rd party and 1st party identifiers that collect in Tapestry allow for straightforward activation approach.
We want to leverage these clusters for when a user arrives on site. For this client, we used Adobe Audience Manager to bring in these IDs and create traits and segments. We were then able to use those segments in Adobe Target to create messages for different clusters which will result in a tailored message based on the cluster.
This is just one case study in the practical application of Softcrylic’s audience modeling and persona identification process. We have worked with clients across industries and use cases and have found this process never fails to deliver audiences that map cleanly onto high-priority business initiatives.
Discovering High Value Audiences in the Travel Industry Webinar Re-cap2022-05-172022-05-30/wp-content/uploads/2024/05/softcrylic-part-of-hexaware-light-logo.pngSoftcrylichttps://www.softcrylic.com/wp-content/uploads/2022/05/discovering-high-value-audiences-in-the-travel-industry-webinar-re-cap.jpg200px200px
Geoff Ogrin is a marketing technology senior consultant. He has over 10 years of experience in digital analytics, managing business intelligence platforms, and leading MarTech platform implementations. Geoff is passionate about solving complex problems with clients to meet their business goals.
Contact Us
We're not around right now. But you can send us an email and we'll get back to you, asap.