So, you go to your Global Brand Hub, where you find several 100+ PDF documents full of various guidelines. You search for “colour” and find a matrix that says “Use Pantone 299C for print, #00A3E0 for digital.” The agency used #00A4E0. Is that a typo? Or a holdover from another guideline deck, “Digital-First Brand Refresh_v3_FINAL.pptx,” from last quarter?

You Slack a senior director, but they're in back-to-back meetings. You email the brand compliance alias and get an auto-reply: “We will respond within 48 business hours.” But the agency is pinging. The media slot is booked. It’s a simple, 10-second question that has blocked a time-sensitive asset.

This seemingly small frustration is actually a symptom of brand governance chaos—a massive, hidden tax on your speed, budget, and morale. Fixing this chaos requires more than just a clearer guide or a better folder structure; the real solution is to evolve from static repositories to dynamic, intelligent agents capable of delivering a single, correct answer instantly.

Agentic architectures help solve for relevant retrieval.

For the last decade, improving access to information largely meant adding a better search box to static, file-based brand hubs. However, a search box only fetches documents; it still forces the user to do the work of finding the answer within those documents. This is the critical failure point in the 4:00 PM panic scenario described above. A better solution is to move from a static repository to a system powered by orchestrated agents that retrieve data.

Unlike a search box, an agentic solution can perform tasks on behalf of the users. It can understand context, like knowing who you are (for example, a brand manager) and what you're working on (in our example above, a digital display). From there, it can reason, retrieve information across multimodal sources (PDFs, databases, websites), verify accuracy, resolve data conflicts, and compose an answer. 

It doesn't give you ten blue links to sift through; instead, it offers a single, definitive, reference-backed response. If conflicting data appears—for example, Marketing_v1.pdf says #000000 but Poster_v1.pdf says #000011—the agents use context, role and logic to determine the most accurate answer. If a clear resolution isn't possible, it flags the conflict so the user can make an informed decision: “The correct hex code for digital-first applications is #00A3E0. The #00A4E0 value is an outdated code from the Q1 refresh.” 

This shift is a powerful new driver of enterprise efficiency today. In fact, agentic assistants like this move beyond being passive tools for answering questions, evolving into Brand Intelligence Systems that retrieve accurate brand data, enforce multi-modal compliance, and generate on-brand, multi-modal content at scale.

Proving a measurable lift in efficiency for over 1,800 users. 

We recently partnered with a global technology leader facing this exact challenge. With thousands of employees and partners across the globe, they needed to provide instant, reliable and source-attributed answers to brand questions, at scale, as their MVP. We designed and deployed a bespoke, enterprise-grade AI assistant powered by orchestrated agentic workflows using a tailored Retrieval-Augmented Generation (RAG) architecture.

This orchestration ensures two things. First, the agents don’t hallucinate. Second, the system understands context (who you are, your role and your task) to deliver the right answer, often by combining multiple verified sources. 

The impact of our solution was immediate. Within four months of its rollout, over 1,800 unique users were interacting with the assistant per month, with engagement trending positively. More importantly, we proved we were solving the slow bleed of inefficiency. User sessions became measurably more efficient, dropping from an average of 1.64 messages to just 1.41, because they were getting the right answer, faster, on the first try.

Crafting solutions that integrate into real workflows. 

There are plenty of off-the-shelf and decent tools already available for building a chatbot. The real challenge lies in building bespoke systems that integrate seamlessly into daily workflows, also known as complex enterprise integrations, that are also secure, reliable, highly accurate and personalized.

This is essential not just for performance, but also for compliance, data protection and user adoption. When the experience feels like a natural extension of how teams already work, that’s when transformation sticks.

However, connecting every piece of the puzzle requires a holistic approach. The development of our solution for this specific client was rooted in a single, unified team that covered everything from initial strategy and user pain point understanding to UX and design. This was made possible by engineering teams who orchestrated models to deliver secure answers at scale, all while our QA and delivery teams ensured everyone remained focused on achieving enterprise-grade outcomes.

In practice, this meant that our strategy team mapped pain points, the AI Core team built datasets and evaluation frameworks, the UX team distilled complexity into intuitive experiences, and engineering ensured scalability, resilience and security.

The hardest part? Balancing accuracy, latency and cost across deep enterprise-grade system integrations. It took many iterations over the past three years to achieve team maturity. Key project members had already worked on five similar conversational AI deployments across industries, and that collective experience was crucial. The learning curve has been steep but transformative.

A unified system built by a holistic team frees creativity.

The 4:00 PM panic is just the surface symptom of deeper inefficiencies that off-the-shelf tools can’t fix when accuracy, latency and cost all matter. 

True success comes from integrating bespoke AI systems seamlessly into the creative process—not as add-ons, but as enablers. This is what a holistic approach delivers in practice: a unified system where strategic insights, intuitive design and enterprise-grade engineering work as one. It is this system that ultimately solves the hidden tax of brand friction, giving your most valuable creative people back their time, budget and energy to focus on the work that actually moves your brand forward.

At the heart of it all is the user. We are driven every day by the goal of building next-generation AI interfaces that are not only intuitive and meaningful but also truly smart. For brands and enterprises seeking to achieve this same level of clarity and efficiency, our bespoke agentic AI architecture can be fully tailored. It adapts to your unique workflows and data environments while respecting all governance requirements, empowering your teams with intelligent systems designed precisely around your needs.

At a glance:

Marketers can use AI to create differentiated and effective work by moving beyond generic tools and building a proprietary system trained on their own unique brand intelligence. Here’s how:

• True AI-powered differentiation comes not from the tools themselves, but from training them on your unique historical data, audience insights, and strategic knowledge.
• The key is to develop a custom agentic workflow—a network of specialized AI agents—that acts as a proprietary, intelligent system that understands and enforces your brand.
• This approach transforms your marketing archive from a static resource into a dynamic, compounding advantage, creating a defensible piece of intellectual property that competitors cannot replicate.

How can we use AI without losing our brand's unique voice?

When everyone is rushing to adopt the same AI tools with the same capabilities, how do you create work that actually stands out? Creative differentiation has long been a critical imperative for brands, which is taking on a new urgency as creative workflows become more automated—at the risk of becoming more homogenized in the process. So, the answer to the differentiation dilemma doesn't lie in mastering the same off-the-shelf tools as competitors. Instead, it lies in building a proprietary system fueled by a company's unique data, intelligence, and history.

As we move from the excitement and potential of the early hype cycle to the messy reality of implementation, many brands are finding that while generic AI tools are powerful, they can lack the brand-specific nuance required for high-stakes marketing if they aren’t trained on your brand’s unique intelligence. The initial thrill has been replaced by the practical challenge of making AI outputs not just faster, but better and—most importantly—distinct.

How can my brand's data create a competitive advantage in AI?

In overcoming these challenges, we must move our focus from the AI model itself to the fuel that powers it. When brands use the same public models with similar prompts, they will inevitably get similar, generic results—a race to the bottom that creates content not worth people's time.

True differentiation comes from the quality and uniqueness of your inputs. The most valuable assets your brand possesses are your own historical performance data, your nuanced audience insights, your risk tolerance, and your strategic knowledge. Altogether, these make up your brand’s DNA. More than just quantitative data, brand DNA includes the qualitative elements that are harder to measure but essential to your brand’s identity: your unique point of view, your strategic decisions about which markets to enter (and which to avoid), and your specific interpretation of complex industry regulations. These are the inputs that contain your competitive edge.

So, how do you activate this unique brand DNA, transforming this treasure trove of data from a static archive into a dynamic, intelligent engine? The path to doing so lies in building a custom, agentic workflow.

What is an agentic workflow for marketing?

A marketing-focused agentic workflow operates as a network of specialized AI agents, all trained on your specific brand ecosystem to work in concert. These agents function like a bespoke, in-house team of experts that thinks and acts like your brand, but with the capacity to operate at limitless scale.

At the heart of this system is the brand model. Training this model involves feeding the system your entire history: your visual identity system, your tone-of-voice playbooks, your compliance policies, and your competitive positioning. We enrich this with the very language from your past high-performing campaigns and the specific nuances of your regulatory landscape. The model learns not just the rules, but the spirit of the brand. This in turn delivers consistency, ensuring every piece of AI-assisted work is instantly and recognizably on-brand.

Within the agentic workflows we build for our clients, we deploy specialist agents to support the core brand model. For example, our Insights Agent surfaces relevant trends and conversations, enabling brands to move at the speed of culture and engage authentically with relevant trends. Meanwhile, Strategy Agents act as powerful accelerators for growth, helping brands identify opportunities and activate new audience segments by analyzing market data through the lens of their unique strategic goals. Insights that in the past could take one of our CPG clients 18 months to identify and activate are now reviewed and responded to in close to real time.

How does an agentic workflow become a defensible asset?

This approach transforms AI into a strategic asset that serves as a powerful engine for differentiation. While competitors rent access to generic models and compete on similar prompts, you’ve built distinct intellectual property that produces creative work no one else can.

This "brand brain" is a living system that grows smarter and more effective with every campaign it analyzes and every data point it absorbs, a virtuous cycle that turns your brand’s history into compounding advantage. In practice, this means the agents work in concert to uncover unique creative opportunities. An Insights Agent might identify a burgeoning cultural trend through the specific lens of your brand's DNA, surfacing an angle invisible to your competitors. This is passed to the Strategy Agent to propose a micro-campaign, which is then validated by the brand model to ensure it aligns perfectly with your tone. This interconnectedness is what allows you to move on distinct cultural moments with speed, relevance, and safety.

What’s even better is that insights from a successful campaign automatically inform the strategy for the next one. The nuances of your highest-performing creative become ingrained in your brand model’s DNA, making the entire system better at finding the unique intersections between what your brand stands for and what the culture is talking about, leading to increasingly differentiated work over time.

Where should we begin in building a custom AI workflow?

In the age of AI, the most durable competitive advantage won't come from a tool you can buy, but from an internal capability you build. Activating your unique data through a custom agentic system is how you make your own intelligence scalable, turning your brand’s history into a strategic asset.

Making this operational and strategic shift is a significant undertaking, and every brand is at a different stage of its journey. That’s why we engage with our clients flexibly. For some, it begins with advisory to navigate the change. For others, we stand up managed agent services to take the burden from in-house teams. And for those ready to invest in self-service, we help execute full custom builds. Regardless of where you start, the most critical step is recognizing that your brand’s history is the most powerful dataset you own.

As EVP of Innovation, Products & Solutions Development, I’ve had firsthand experience helping brands innovate through our brand model practice. This service, strengthened by our recent partnership with Adobe, delivers three key components: fine-tuned models, updated brand prompt guidelines for the AI era, and a prompt bank to jumpstart AI usage. Together, this enhances our ability to help brand marketing teams quickly plan, create, manage, activate and measure on-brand content.

As you can imagine from that description, building brand models relies on more than just technology. Change management and the people element are just as crucial. With that in mind, here is a recipe to bring data, people, processes and technology together to create powerful, bespoke brand models—and some of the successes we’ve seen so far with their implementation.

Data is essential to building brand models—but is only part of the equation.

First, it’s no secret that artificial intelligence relies on vast amounts of data for training, including customer insights, media data and historic brand IP. Therefore, it may seem as though the more data you have, the better—but it’s not that simple.

While a data lake is nice to have, its benefits will diminish over time. Brands seeking to build AI models tailored to their specific business needs require a constant stream of fresh data to adapt and improve over time. This begins with the adept orchestration of data pipelines: the process of designing, managing, and optimizing the flow of data from various sources to ensure it’s current, relevant and seamlessly integrated into your AI systems.

But equally critical is fostering a collaborative environment where your team can effectively leverage the technology and ultimately keep these models up to date. A Forrester report by Forrester VP, Principal Analyst Jay Pattisall—titled “Brand AI Models Will Reinvent How Marketing Creates Business Value”— discusses brand models in detail and hits on the role that people play in filling in those gaps left by historic data. There, he quotes me saying, “[Foundation] models are a picture from the past with data from the past. The future will be continuing to iterate with the model, and that’s the people.” Let’s dig deeper into how aligning your team around these data-driven initiatives can drive transformational change.

Organize people and pipelines into efficient, automated workflows.

We help brands re-engineer their workforce for AI readiness with a consultative phase focused on change management and breaking down silos. This involves pragmatic advisory to understand where you are in your AI journey and implementing a workflow infrastructure for your tech stack and teams, because ultimately the job of the orchestration layer—or the part of the tech stack that manages, coordinates and optimizes interactions between services, application and processes—is to amplify the best of the technology you have and the best of the people you have.

By leveraging technologies like Monks.Flow, for example, we can streamline how people and AI work together. Monks.Flow is an AI-powered professional managed service that connects employees, vendors, AI and enterprise software into efficient, automated workflows. This platform-agnostic environment is designed to work seamlessly across an organization’s existing tech stack, helping businesses achieve benefits like the integration of tools, people and processes; building business intelligence; and maximizing the impact of their marketing efforts to drive growth. Those who marry skillful use of the technology with deep brand expertise can provide strategic oversight throughout this workflow, thereby achieving stronger outcomes than if you relied on the technology alone.

A perfect example of this is an incredibly promising brand lift study we did with Forever 21. The fashion retailer leveraged Monks.Flow and Meta’s AI Text Generation to elevate the creative strategy of their shopping campaigns. We created 50 variations of Forever 21’s business-as-usual (BAU) creative and used Meta’s AI to generate copy variations powered by Llama 3. This conversion lift study aimed to measure the impact of these AI strategies on campaign performance. By incorporating AI to generate diverse copy and imagery, we were able to test and optimize for the best-performing creative. The results were impressive, showing significant improvements in click-through rates and conversion rates compared to BAU assets alone.

Talent is key to domesticating brand models and AI-generated output.

Just as it's crucial to prepare and collaborate with your internal workforce, it's equally important to engage with partners within your vertical for AI consultation and implementation. These experts are deeply immersed in your market of choice, understanding its nuances and dynamics, and can leverage this knowledge to fuel your strategic AI efforts. Moreover, a partner who knows your brand inside and out can provide tailored insights that align perfectly with your business objectives, ensuring a cohesive and effective AI strategy.

For instance, I spoke with April Huff, who has successfully incorporated AI into her role as VP of Research & Insights at Media.Monks to build audience personas and aggregate data from focus groups. As a researcher and strategist, she seeks intriguing differences in focus group data, because those unique viewpoints often trigger the best creative ideas.

She emphasized that because AI models excel at predicting the next most likely word when generating text, there’s a tendency toward homogenization of experiences—which can be valuable in market research, but not always. “What makes humans interesting is that they will  say something that is unpredictable,” she says, underscoring the role of talent who can parse these nuances and add depth to AI-generated insights.

One example of how humans teased a surprising creative insight from AI data is our work with Sephora on its “mAI colpevoli” campaign, which launched on International Day for the Elimination of Violence Against Women. We used AI to analyze online data and create three scripted monologues that retold typical episodes of gender-based violence.

The team realized the AI-generated scripts had something in common: written as monologues, they each featured a speaker who blamed herself for what happened to her. The output highlighted the pervasiveness of victim blaming, mimicking societal biases, and we realized we could use this insight to encourage audiences to reflect on their own propensity for blaming victims (or themselves) of injustices. By marrying AI insights with strategic oversight, we were able to create a powerful, purpose-driven campaign that supported women's freedom and self-expression, demonstrating how deep brand expertise and AI can work together to drive transformative change.

The secret to brand model success: people and tech working together.

The integration of artificial intelligence within the enterprise is a multifaceted endeavor that goes beyond mere data accumulation or technological foundations. To truly unlock the potential of AI, brands must focus on creating self-evolving models that adapt and improve over time, driven by continuous data generation and strategic human oversight. This approach ensures that AI systems remain relevant and effective in an ever-changing market landscape.

Ultimately, the successful implementation of AI hinges on the symbiotic relationship between advanced technology and skilled human operators. It requires breaking down organizational silos, fostering collaboration, and engaging with partners who understand both the market dynamics and the brand’s intricacies. As brands navigate the complexities of AI adoption, focusing on these strategic elements will pave the way for sustained success and innovation.