AI-Powered Segmentation Revolutionizing Marketing with Neuralift

Welcome, everyone, to the latest edition of *AI Marketer Skills*. Today, we're excited to have Jonathan Mendez from Neuralift here to share insights on segmentation and more. I've known Jonathan for a long time, probably since his time founding and running Yieldbot—a fantastic performance ad network he built and managed for nearly a decade.

**(00:51)**  

Jonathan, it’s great to see what you’re doing now in this AI-powered era. Before you dive in, just a quick reminder to everyone: please feel free to use the chat, raise your hand, or come on camera if you have questions. We want this to be interactive and make the best use of your time today. So, Jonathan, over to you.

**(01:24)**  

Thank you, David. I appreciate everyone being here. Whether you’re joining during lunch in New York, breakfast on the West Coast, or at the end of your day across the Atlantic, I’m thrilled to have you here. First, I want to thank David for creating the *AI Marketers Guild*. We've been building Neuralift for about a year, and the forum David has established here is incredibly valuable for connecting people in marketing AI.

**(01:53)**  

As many of you know, AI is transformative. If you’ve worked in the field or are involved in projects, you’ve likely found it eye-opening. Having a community like this that David has organized, with resources like our Slack channel, helps foster connection and knowledge in the marketing industry. David, thank you for your vision and for bringing us together.

**(02:21)**  

Reflecting on how far we've come with data, I remember when getting data was as simple as pulling it from Urchin, which later became Google Analytics. Since then, data collection has evolved tremendously, especially with first-party data and data privacy. Today, it’s more critical than ever for brands to collect and leverage data effectively. While many of us work with giants like Google, Meta, and Amazon, they each have vast data and advanced AI, so we need tools that allow us to be competitive.

**(02:56)**  

With AI, we now have ways to level the playing field. Neuralift was built to help brands and agencies leverage data to compete more effectively. I'm excited to show you what we’ve been working on. I'll start by sharing a bit about Neuralift, and then we’ll dive into a demo.

**(03:32)**  

Our concept for Neuralift came to me during my time as Chief Digital Officer at a large travel brand. We had access to extensive data sources—Google Analytics 4 streaming in real-time, Google Cloud Platform, Snowflake for transactional data, and a CDP for lookalike modeling. Yet, with all that data, creating actionable value was still incredibly challenging. I realized we could use AI to build precise and insightful customer segments.

**(04:07)**  

I reconnected with some folks from Yieldbot, where we had developed extensive machine learning applications. Given the advances in AI since Yieldbot, we saw an opportunity to use AI for customer segmentation that would truly understand audiences and their unique preferences. That’s how Neuralift started, and today I’m excited to share our journey with you.

**(04:40)**  

Let’s talk about segmentation itself. In marketing, we use segmentation for various purposes, whether it’s personalization, targeting, or loyalty programs. On the paid media side, we refer to “audiences,” but the concepts overlap. Here’s an early screenshot from Google Analytics showing default segments. Segmentation, simply put, is grouping people with similar affinities. The goal is to match the right person with the right product, offer, or message at the right time.

**(05:46)**  

Segmentation has numerous applications, from planning and creative testing to personalization and pricing. AI has allowed us to take segmentation to a new level. Recently, Coca-Cola, Bain, and OpenAI conducted an *AI for Personalization* project. Their study showed that AI can provide faster insights, allow for rapid iteration, and serve as a creative tool that inspires strategic thinking.

**(08:24)**  

When it comes to segmentation, the current approach with many SaaS tools results in different segmentations across channels—email, personalization, loyalty, custom audiences on platforms like Google and Meta, and more. This fragmented approach can lead to inconsistent messaging. At Neuralift, we unify customer data across these channels, allowing brands to create comprehensive, cross-channel segments.

**(09:29)**  

Currently, most segments are created with SQL. But this traditional approach requires prioritizing which data columns to include, often introducing bias and limiting the potential insights. Neuralift, on the other hand, can handle vast datasets—millions of rows, thousands of columns—so you’re no longer forced to simplify or reduce data.

**(11:02)**  

Additionally, current segments are often built to optimize KPIs, but without grounding in true incrementality. Neuralift, however, grounds segmentation in key metrics. With precision as a core focus, we can enhance operational efficiency, improve ROI, and create value more rapidly, aligning with the fast pace of marketing and advertising.

**(14:06)**  

Now, let's dive into the product. Neuralift starts by ingesting a CSV file from the brand. After securely uploading the data, our deep learning models process it. We don’t need any personally identifiable information (PII) for our outputs. The system then clusters the data, producing segments with names, summaries, insights, personas, marketing recommendations, and underlying data for activation.

**(16:53)**  

Let me show you a live example. Here is a customer dataset with 2.5 million rows. Each entry is linked to a customer ID, along with data on device preferences, purchase patterns, loyalty, app usage, and transactional details. This kind of dataset is a mix of behavioral and transactional data for an e-commerce retailer in women’s fashion.

**(18:36)**  

Now, here’s the Neuralift dashboard, where this dataset has been segmented. The system identified 29 segments based on the data. These segment names, summaries, and insights were generated by AI and processed in under two hours. This efficiency offers a huge advantage in terms of speed to value, as every part of this was generated without manual intervention.

**(20:23)**  

To make this actionable, we provide a view of key metrics, allowing brands to see how segments distribute across KPIs like loyalty, churn, or new customer acquisition. For example, if a brand defines "at-risk" customers as those who haven’t purchased in the last six months, we can show the distribution of at-risk customers across segments.

**(22:42)**  

Let’s look deeper into one of the segments. For instance, we have a segment called “Evening Elegance Weekend Fashionistas.” This group predominantly shops in the evenings and weekends, has a high app usage rate, and tends to be acquired through paid search. The AI provides a summary and specific insights on this group’s behavior, enabling targeted marketing actions.

**(26:46)**  

We also provide tailored marketing recommendations for each segment. For instance, based on a segment’s behavior, we might suggest targeted messaging or campaign ideas. In the future, we’re collaborating with partners to automate these recommendations directly into campaign management systems.

**(30:12)**  

For brands interested in building models such as next-best-action, Neuralift segments can serve as a starting point. Brands can take Neuralift’s highly precise segments and layer their own models for even more tailored actions.

**(36:19)**  

On the topic of reliability, the insights provided by Neuralift are based on data calculations and verified through strict checks to ensure accuracy. We validate every insight against the underlying data to maintain high accuracy.

**(45:12)**  

For the personas, these profiles are generated using the underlying data and refined using frameworks that have proven effective in marketing over time. This blend of AI and marketing best practices enables us to create personas that offer genuine marketing value.

**(47:54)**  

Finally, Neuralift includes a generative co-pilot that creates structured outputs for specific channels—whether that’s SMS, email, social ads, or search ads. The co-pilot can generate targeted campaign ideas with data-backed messaging for each segment.

**(50:56)**  

In terms of cost, Neuralift pricing varies based on the size of the brand and the dataset. However, the ROI is immediate, especially when compared to the cost of hiring data analysts or a team to manage segmentation manually. Neuralift offers a much faster and more efficient solution.

**(54:24)**  

If you’d like to follow up, feel free to connect with me on LinkedIn, or email me at jon@nuro.ai. I’m happy to answer questions or provide further demos if you’re interested. David, thank you for having me, and thanks to everyone for joining!