As every holding company and agency races to assemble its own version of an AI-driven platform, Horizon Media is pitching its version, Blu, as a sort of anti-black box consultancy that helps clients not only create campaigns, but also find groups of clients they might otherwise have missed.
Overseen by Bob Lord, president of Horizon Media, and led by Domenic Venuto, Horizon’s chief product and data officer, Blu is essentially a connected marketing platform that uses a variety of LLMs to help independent agency clients determine broader business goals through the lens of media (the creative contributions will come later – and we’ll talk about that later), only on steroids.
Blu was first launched in 2020 as an advanced data platform using various machine learning elements. But it has evolved into a simple form of advice for clients who use it. These customers include retailers like Bob’s Discount Furniture, grocery chains like Wegmans and telecommunications companies like Spectrum. Venuto described it as “native AI,” meaning built from the ground up, rather than on top of existing systems.
It is intended to leverage the various core LLMs, given the speed at which they update themselves, while using TransUnion data as the data backbone. Lord said a key part of speeding up the process without sacrificing data quality is organizing the data in a flat lattice structure. Older data sets were arranged in rows of tables, making them harder for AI to read. “In the new AI landscape, designing data so that the LLM can access it easily is actually the magic of speed,” Lord said.
The system is meant to operate within each customer’s data walls, Lord said, which he sees as an important distinction from the competition. “Customer A owns their data — it’s on their own cloud… As the platform learns, it learns the way Customer A does. Therefore, these models stay within the customer, which gives them a competitive advantage… My competitors don’t do that. They will reuse the models for customers on their behalf and aggregate them.”
Blu is built with transparency in mind, where all sources that LLMs derive information from are listed alongside the quick results. “If you’re not transparent about the source of the data, where it came from, what it is, how it was trained, what the model looked like, you’re not going to have good acceptance, because then people won’t trust the system,” Lord said.
And the main focus is on building audiences, either by better understanding existing customers or finding potential customers, said Zach Claisse, vice president of data solutions at Horizon, who calls himself a “customer architect” and sits between the business solutions team and the product team.
With each client, Blu can also be used as a business consultancy, analyzing the client’s business, competition as well as existing and potential customers. For example, a supermarket client can use Blu to determine where to open new stores based on customer potential, then develop a media plan for that store opening. “When we think about our media strategy, we come up with the most optimal plan that will actually move the business forward, not just hit the front-end media metrics,” Claisse said.
“This is where we move from just being an agency of record to a growth agency, where we have the tools and capabilities to answer business questions that extend beyond just media, the traditional media questions that we were responsible for,” Venuto added.
At least one client who uses Blu corroborates the broader appeal the platform offers them when it comes to advice. “We can create our own audiences just for analysis purposes – creating dashboards and developing modeling so we can answer questions that aren’t just for targeting with this or that ad,” said Lauren Pearson, director of CRM at Bob’s Discount Furniture. “Instead, we can answer questions that address the internal needs of our business, for example: What is this type of customer interested in beyond just furniture? Where are they shopping? What is their browsing behavior? … This really shifts the horizon towards a kind of business consulting rather than just an agency.”
Eventually, the platform will also encompass creative contribution, Venuto said, and pilot a partnership to access and optimize creation by connecting through APIs or MCPs – and agent options. “We see an opportunity to use technology to play with creative assembly and disassembly, scoring and generation, and use those signals to feed media and vice versa,” Venuto said.