The Evolution of Audience Segmentation for Enterprise Organizations

The landscape of audience segmentation within enterprise organizations has undergone significant changes over the last two decades.

Early 2000s: Choosing Between Technical Expertise and Siloes

Many enterprise organizations still fall within the confines of the first era of options:

SQL

Initially, SQL was the sole method organizations had for constructing audience segments. Some first-gen martech tools emerged, including Unica, SAS Marketing Automation, and Teradata CIM. These tools still find a place in many tech stacks today. The use of a home-built CDP in Snowflake can replicate similar patterns, even though Snowflake would not be classified as martech software.

Channel-Specific

This is the most basic option available. Delivery tools like ESP are less commonly employed for audience segmentation in enterprise organizations. However, this practice persists despite the silos it creates between different channels. This approach continues to dominate the advertising sector, where many still build their audiences within walled gardens. The primary advantage of channel-specific segmentation is a reduced reliance on SQL expertise.

Cross-Channel

Marketing Automation Platforms (MAP) cater to owned channels like email, SMS, and direct mail, while Data Management Platforms (DMP) are geared toward advertising. MAPs are notorious for encountering scalability issues and only providing a fraction of the enterprise data to marketers.

Mid-2010s: Centralized Segmentation For Marketers

The emergence of Customer Data Platforms (CDPs) marked a new generation of martech:

Omnichannel

This level of advancement allows for the creation of audience segments across both owned and paid channels. Modern composable CDP solutions take it a step further by enabling direct integration with data warehouses. This presents a groundbreaking opportunity, reminiscent of the data access that "SQL in a warehouse" once provided.

Intelligent

Building upon the previous option, this approach incorporates AI/ML to identify audiences, moving beyond the manual definition of segments through rule-based methods.

2025: The Next Gen(-eration and -erative) Audience Segmentation 

If we summarize how audience segmentation is performed by marketers today, using any of the previously mentioned tools, there are really two options:

Option 1: The Infamous (Jira) Ticket

Business users often lack SQL knowledge, which is required to access audiences. They create a ticket, which spares them from understanding the underlying data. But oh man, it can take weeks to receive the assigned ticket back with the audience data.

Option 2: The No-Code UI (DIY)

We have grown to love martech. Business users have direct access to the data, enabling them to create audiences themselves with no more "wait and see" delays. Audiences are accessible in minutes. However, the downside now is that business users must understand the complex underlying data and schemas, particularly in enterprise organizations.

Now, envision a scenario where a natural language prompt, similar to a Jira ticket for marketing operations, empowers organizations to create and access audience segments seamlessly!.

I’ve predicted last year that we will enter a new phase of disruption for audience segmentation, coming from Generative AI. GenAI could offer the ability to combine the best of both options—The Ticket and The No-Code UI—without their limitations. I estimate that, by 2025, enterprise organizations will gain—in a production environment—this new segmentation UI.

By taking a look at the roadmaps and recent announcements from MarTech vendors, I may be pretty on target with that estimate.

That being said, I still believe that it will take longer before having a significant positive impact in the organization. There remains too many unknowns (trust, transparency, security, prompt engineering expertise) to believe that, as soon as 2025, a GenAI audience segmentation UI would perform better than our current MarTech UI options. So remain cautious about promises when evaluating software today.

The opportunity remains extremely exciting! 

🔑 Users will very soon gain a new way to discover and create audiences, leveraging the power of GenAI.