The Evolution and the Future of Data Modeling for Direct Mail

Key Takeaways:

  • Multiple transactional, online and offline data sources have provided increasingly deep insights into consumer behavior over the past two decades
  • That explosion of data caused the legendary “big data” issues, which have been harnessed by AI and Machine Learning breakthroughs to process large data sets into meaningful models
  • Once, it took weeks to generate a single model for testing; now, we can generate multiple models for testing in about ten working days

Listen to the podcast: Addressed for Success

Staying Ahead of the DM Data Curve

Since its inception in the 1980s, data modeling has profoundly impacted direct mail marketing to accomplish various marketing objectives across different channels. The goal is always the same — generate incremental business by finding new people who are most like those currently exhibiting profitable behavior. When it comes to prospecting, that is finding lookalike prospects who most closely resemble existing customers.

At its start, modeling focused on creating customer data profiles and segmentation analyses based on demographic information.

The combination of growing modeling expertise increased computing capacity, and the availability of enhanced data sets improved modeling effectiveness incrementally over the course of its history. But in the last ten years, we have witnessed exponential changes in data and analytics to enable increasingly effective model building.

e-Commerce, Social Media, and the Deluge of Data

The advent of digital commerce led to the development of web analytics tools designed to gather all data related to website visits and online consumer behaviors. This enabled modeling to focus on prospects’ actions, such as the customer journey. In addition, social media started generating massive amounts of data, and the challenges of dealing with “big data” are legendary. Issues like storage capacity, processing speed, and data analysis led analytics into a new era of more advanced techniques, such as machine learning and text analytics.

Starting in the mid-2010s, marketers began combining online and offline data to complete a 360-degree view of the consumer. Data models focused on collecting all data coming through from multiple systems like point of sale, CRM, social media, and website interactions. These data challenges have eased with Artificial Intelligence (AI) and Machine Learning (ML) breakthroughs to enable data analytics automation to uncover patterns and help marketers make real-time predictions.

Modeling exercises once took weeks and cost thousands of dollars. With the implementation of AI and ML, the turnaround time has dramatically reduced — in certain cases, data providers will waive costs associated with building a model in lieu of data usage commitments.

“Instead of taking weeks to generate a model, now we can generate multiple models for testing in about ten working days.”

So, how does this benefit marketers? These days, we can generate multiple models based on various input assumptions and back-test them against a customer database to predict success — all in about a couple of weeks. Whereas in the past we would be excited to evaluate a single model in a mailing, we will often evaluate multiple models from different data sources simultaneously. This helps us speed up years of learnings down to just months.

Gunderson Direct has long-lasting relationships with some of the country’s largest corporations, helping them to lower their customer acquisition costs and increase profits using address-based integrated direct marketing programs.

Drop us a line for more information on how our direct marketing expertise can help your business.

If you found this information valuable, sign up for our newsletter.

Sam is a data problem solver. His deep expertise in customer acquisition and retention comes from over 25 years of experience working at data powerhouses such as Epsilon and RL Polk and for major clients across numerous categories including financial services, retail and entertainment.

Sam doesn’t just find and analyze databases for clients, he builds them, too. He’s created targeted universes through analytic solutions, ranging from simple data profiles to robust modeling techniques —always with a focus on data quality. His skills also extend to designing and implement primary marketing research solutions.

Outside of work, Sam enjoys watching Liverpool, his favorite football team (don’t tell him it’s really soccer). He loves spending time outdoors enjoying nature, relaxing at home with his family and two dogs, and learning about other cultures through travel.

Recent Posts