Working with Catalyst is easy. Tell us a little about your organization below. One of our strategists will get back to you shortly to learn more about your goals.
Few teams in an organization feel the impact of new business technology faster than marketing— not because marketers chase innovation, but because keeping pace with audiences’ attention and decision making processes demands it. That pattern is repeating today as marketing leaders rush to explore how they can use AI to improve everything from content to analytics. While there’s no doubt that the technology will be transformative, what’s less clear is how. At this stage, artificial intelligence represents as many unknowns as opportunities.
AI’s most profound effect on the marketing ecosystem may not be on any of these capabilities, though. Instead, it may be on the symbiotic relationship that’s evolved between in-house and third-party marketing teams over the last century—a partnership that’s evolved to deal with an increasingly complex communications landscape, boom-and-bust economic cycles, and ever more constrained budgets. In addition to allowing in-house marketing teams to quickly expand their capabilities to handle specific channels, audiences, or stages of growth, these partnerships have allowed leaders to keep their marketing footprint both lightweight and flexible.
As AI begins to decommoditize the intellectual capital and bandwidth that agencies bring to the table, it begs the question: In light of a low-cost AI “alternative”, are hub-and-spoke models that depend on outside expertise, like agencies, still relevant? Or, can we use smaller, AI-empowered teams to achieve the same objectives? In short, are marketing agencies still relevant in the age of AI?
The speed with which AI has gone from an experiment to an essential tool in the modern marketing tech stack has been nothing short of extraordinary. Not only are we finding new ways to leverage the technology almost daily, at times it feels like we’re part of a global product team, working together in an iterative cycle of feedback and development. For those of us who grew up at the advent of earlier technological shifts, from personal computing to the internet to social media, it can feel almost nostalgic. For others, these capabilities represent a very real existential threat to the marketing profession.
The way that similarly disruptive technologies have unfolded may hold important clues about what to expect from AI. For instance, we know that the hype around new technologies often outpaces their more mundane limitations. We also know that the way they evolve rarely follows a straight line, ends up where we first anticipated, or happens as quickly as we’d like. Finally, a technology’s impact often depends more on accessibility than on raw capability. It needs to be human to be transformative. When aspirations for new technologies collide with their more practical application, it can create a misalignment between planners and doers that takes years to overcome.
We’re already seeing this pattern play out across the marketing landscape. Executives who are eager to reduce costs, optimize revenue, and speed growth are rushing to leverage AI as a method of building and testing ad creatives in real time. The results, ranging from seven-fingered people to poorly worded and even offensive copy, can be amusing at first glance. They can also damage a brand’s reputation, erode authenticity, and ultimately provide little real value. While these challenges will almost certainly be solved in time, it’s a good reminder of how important it is for marketers to understand emerging technologies before relying on them.
End-to-End Support for Marketing Leaders
For marketers weighing whether AI can truly boost the bottom line, the challenge is to identify where it delivers real operational improvement, not just cost cuts.
A common assumption among executives is that agencies of the future will simply subcontract their work to AI tools, charge the same fees, and pocket the difference. That assumption naturally raises the question of whether in-house marketing teams can simply adopt these same tools and cut out the middleman. The assumption, however, is based on a flawed understanding of how agencies work, the value they provide, and their role within the current marketing paradigm.
Over the last fifty years, agencies have become a vital piece of the marketing ecosystem. That’s especially true for growing teams, which use these partnerships to bridge the gap between their own limited resources and the need to deliver the seamless omnichannel messaging audiences now expect. The gap between resources and capabilities isn’t just about a lack of channel expertise, though. Marketing agencies function as hubs, integrating the strategic, creative, and technical resources their clients need to execute at scale. That allows client organizations to execute the same kinds of strategies as better resourced competitors — with less overhead.
This symbiosis has driven major changes in how marketing leaders build their teams over the decades. One of the most profound is that the role of marketing agencies has shifted from outside consultant to an extension of its clients’ own in-house team. By embedding into their clients’ workflows, agencies can deliver the type of real-time communications that powers today’s marketing ecosystem. This integration helps organizations stay competitive, expand their reach, and stay current with emerging trends, all while keeping their marketing footprint lean.
For many senior executives, the rush to AI stems from a desire to upend this relationship, to remove the ‘middleman’ and empower in-house teams to deliver the same results for a fraction of the cost. However, these strategies often prioritize cost savings at the expense of effective operationalization. That often means overlooking known weaknesses and ultimately increasing risk rather than reducing it. For example:
LLMs, like other types of automation, deliver significant value on well-defined, repetitive tasks. However, they tend to struggle with more complex tasks that require stakeholder judgment or adaptation to unexpected variables. That is notable on its own, but also ignores a somewhat darker byproduct: When in doubt, chatbots tend to present themselves confidently, even on tasks they’re not good at or don’t have an answer to. That’s because when AI encounters the unknown, it often fills in the gaps by making things up.
Almost all of the strategy, messaging, and brand-building marketers deal with on a daily basis is rooted in some sort of larger context. That context can come from personal experience, pop culture, or any number of external influences that shape how we communicate and interpret meaning. For humans, that context runs quietly in the background of our lives, but it’s one of the most important factors in our ability to navigate complex communication; even then, we sometimes get it wrong. While AI has access to the entirety of human history, news, and information at its digital fingertips, it still often misses the localized nuance that drives real human decisions making.
Leaders sometimes assume a single AI can integrate seamlessly across their entire tech stack. In reality, achieving that kind of transformation would force organizations to rethink their processes, maintain full technical teams, and continually adapt to new capabilities and limitations. The cost of this effort would quickly overcome any savings. The scope of these efforts may vary, but the vision of a fully automated workflow within the next decade is just as unrealistic here as it is elsewhere — and for the same reasons.
Brook’s Law—a well-known principle among developers—counters the idea that adding more resources to a project automatically leads to better or faster results. The saying will probably gain new relevance in the coming years as execs seek to grow the bottom line by throwing more agents at each challenge, instead of addressing the underlying issues that cause them. While it’s certain that AI will add value, experience tells us that the way its applied will matter. For the time being, this most likely includes human pilots commanding a set of AI drones.
One of the more revealing insights in the discussion about AI’s ability to replace marketing roles is that we’re at least as unsure of how to use the technology as we are certain about how not to use it. Most marketers tend to treat AI like a SaaS tool, using it to handle a set of closely related tasks. But the open-ended nature of generative AI means that, instead of being handed a defined use case, users have to create their own guard rails — a challenge many teams aren’t used to facing. The idea of an “everything app” that can handle tasks ranging from keyword research to press release development sounds enticing, but its lack of structure can create as much work as it performs.
Marketing agencies, like most other organizations, are integrating tools like Claude, ChatGPT, and Gemini into their daily workflow. The proprietary nature of these tools, along with security and reliability concerns, means agencies will likely take an enterprise-scale approach to AI transformation, though. That is, most agency efforts won’t focus on leveraging consumer-facing tools, but on developing proprietary agents that reflect their unique services, values, business practices, delivery models, and expertise.
AI at this level will likely resemble how agencies use cloud architecture platforms like Azure, AWS, or Google Cloud.While generative AI is breaking down many of the technical barriers that once limited growing organizations’ access to this technology, developing and maintaining these systems will still fall outside the technical and financial capacity of most teams. Even for enterprise teams with the resources to build AI tools from the ground up, the economics of creating campaign-specific agents often won’t add up.
If this paradigm sounds familiar, it should. The same forces have long defined the symbiotic relationship between organizations and the agencies they rely on. Agencies’ ability to scope, build, and maintain AI-driven marketing tools (and to fine-tune them across multiple use cases) will ultimately deliver greater value for clients than developing them in-house. That, in turn, will help organizations keep their marketing footprint flexible, lightweight, and up to date. It will also create a competitive moat for agencies, built on the quality of their proprietary agents and their ability to optimize them quickly.
Proprietary agency tools will encompass a wide range of applications. Central themes will include integrating business and marketing data, accelerating iteration, and customizing models for specific objectives. These will include tools that allow agencies to:
If the value of today’s agencies is based on having the right answers, marketing partners of tomorrow will be judged by the questions they ask—knowing how to write prompts that don’t just get the answer they want, but uncover fundamental truths about an organization’s approach based on its unique market, brand history, and competitive landscape. This includes knowing when to question the technology, shift context, or start over.
The good news is that there are still plenty of untapped opportunities to use AI in ways that deliver the outcomes executives are looking for. Many of them, however, focus less on replacing intellectual capital with AI and more on driving profits through process improvements such as smoother communication, better collaboration, and deeper shared understanding. These improvements shift the focus from automation for its own sake to using AI as a tool for alignment — connecting people, processes, and decisions more effectively.
Anyone who’s worked with a marketing agency, for instance, knows that the biggest challenge often isn’t the quality of output, but the sheer volume of back-and-forth required to get there. Alignment between internal and external teams often forces stakeholders to step outside their normal workflows. That creates friction. Because AI can understand and interpret tasks, it’s easy to imagine a future where one of its most valuable uses is facilitating information flow across teams and bridging disconnected technology stacks and processes. That includes streamlining the cycle of scoping, updates, iteration, and analysis that make these partnerships work.
One of the most compelling applications for AI in the agency ecosystem may be in helping agencies extract meaning from an ever-growing mountain of client data. While dashboards in tools like Tableau or Power BI help agencies tell better stories, they’re still constrained by the KPIs they were built to track. The ability to query a client’s data in plain English could enable deeper, more responsive, real-time collaboration that connects cause and effect. Even so, dashboards and the KPIs that define them remain vital. AI allows agencies to surface a wider range of feedback more quickly, enabling faster insights and more responsive decision-making.
AI will give agencies the ability to analyze and deliver research insights that would otherwise take weeks to produce. That includes enabling strategists to explore and extract meaning from data rather than simply report on it. This shift also means external marketing teams will likely become even more embedded in their clients’ day-to-day operations. With deeper insight into marketing needs and budget cycles, agency teams can use AI to identify hidden pain points before they become problems, helping clients and partners work together more efficiently.
Finally, AI will allow agencies to codify and share institutional knowledge across teams. This could, for example, take the form of an AI-powered “living brief” that archives past campaign decisions, creative adjustments, and performance insights — surfacing them when marketers plan their next push. New stakeholders — whether a junior strategist at the agency or a rotating marketing lead on the client side — can query why certain audiences were prioritized, what messaging resonated last quarter, or which channels consistently underperformed. That transparency drives efficiency, accelerates onboarding, and keeps both sides aligned on brand voice and strategic priorities.
While it’s tempting to think of AI in terms of flashy new capabilities, the most effective implementations will almost certainly blend into existing workflows — quietly strengthening collaboration, deepening insight, and accelerating iteration between teams. The impact of that shift will eventually show up in the bottom line, but it starts by supporting more strategic, effective relationships that maximize every marketing dollar.
Over the next few years, AI will drive a tectonic shift in how we work. Some see it leading to a massive reduction in human labor; others believe the time and bandwidth it frees up will fuel even faster innovation. For marketers, this shift means rethinking nearly every aspect of the job — from the creative process and workflow management to how audiences discover and build relationships with the brands they love.
To keep up with this change, marketers should view agencies not as cost centers, but as springboards for their own AI transformation.Over the next few years, the most valuable agency partners will be those that define themselves by more than subject-matter expertise. They’ll use their own AI tools to craft deeper, more authentic, and more personal relationships with audiences. Marketing agencies that understand this distinction won’t just survive the age of AI — they’ll define it.
As marketing teams continue to explore AI, executives are asking whether the agency relationships that have powered the communication ecosystem for decades are still relevant. Learn more on this week's post.... 
For many organizations the tension between sales and marketing teams can be palpable. Greater collaboration between the two starts with deeper understanding about their respective roles. Learn more about how sales and marketing teams can drive stronger growth on this week's Insight....
AI has had an enormous impact on the digital marketing ecosystem over the last year that's only growing wider. On this week's Insight, AI in Digital Marketing...
Original content is even more important to getting found in the age of AI. In this post, learn how to build and scale a high-volume content engine....
Standing up a full-time marketing team may be one of the most underestimated milestones in the journey of any organization. Because marketing isn’t typically deemed “mission-critical”, it's usually the first to take a hit during cut-backs and the last to receive funding when things are going well....