The following is a short extract of The state of Marsch 2025 Report that will be published on May 6, 2025 – #Martechday. Register for free to attend our key presentation, which you can see live or on demand. You will also obtain the first access to our report of more than 100 pages, the new Marketing Technology Panorama 2025 and the complete cover of all the inputs of the 2025 Stackie Prize.
Do not miss the Marsch package of the year, register now.
Marsch is about being complicated be complex.
Those two words sound like synonyms. But here we want to say them as technical concepts that are fundamental different from each other. To explain this distinction and what it means to the future of marketing operations, we want to use the cynefin framework.
Created by Dave Snowden in 1999, the Cynefin framework describes four different types of envy in which the meaning occurs, complicated, complex and chaotic in a 2 × 2 matrix.
Starting from the lower right, a “clear” environment for the meaning is where the cause and effect are directly and obviously linked. Do unknown and AND It will happen. It is almost a perfectly predictable machine. Anyone can intuitively understand how it works.
We use an image of a manual juice to squeeze as a visual metaphor. Take half of a citrus fruit, tighten and turn in the squeezer, get juice. There is not much request for purification or analysis of advanced analysis. In marketing, an example could be a simple and fixed automation, such as sending a standard “welcome” email to someone registered for your newsletter.
Moving towards the upper right corner, we enter the “complicated” space. In a complicated environment, there are now many moving parts, intricately interconnected. Things still work deterministic, but they continue a much more windy path. Instead of unknown and AND It will happen, it is more like: to do to and then unknown” ANDWorks Council Z It will happen depending on the steps and branches between Arizona. Only an expert who really understands the architecture of a complicated system can understand the complex decision logic between cause and effect. For a non -expert, the result may seem random. But it is not.
The visual metaphor we choose here is a Ferrari. It is a machine made with many interconnected pieces. They work together predictably. However, although it is quite easy to get behind the wheel and drive one, if something breaks and the car stops working, it may not be so easy to discover why. It is likely that you need a trained and experienced Ferrari mechanic to diagnose and repair it.
In marketing, we believe that Marsch’s battery as a whole, at least how it is structured and operated in the legs to date, is an excellence example of a complicated system. The logic of how its battery works is deterministic, a function of how all differential platforms and applications within it have programmed, configured and integrated bone. But because the general system is very large and has so many interconnected parts, it can be difficult for a non -expert to understand why it behaves in certain ways. You really need a trained and experienced marketing operations professional.
Now things get interesting when we move to the upper left, complicated to “complex.” In a complex environment, there are also many, many different and interrelated parts. But they have greater independence, and the way they interact is more dynamic and probabilistic. Without a doubt, there is still cause and effect. But it is impossible to perfectly predict a specific cause result in a specific effect.
Our visual metaphor here is a tropical jungle. It is an ecosystem of many different flora and fauna that interact with each other in all kinds of fascinating ways. The causes begin the effects, which serve as causes of other effects, which cause other things … and so on in the great circle of life.
People can definitely affect a tropical jungle, unfortunately, generally not for better. But we can also be administrators of the ecosystem and participate, at least directionally, with causes and effects in the context. For example, if you get the tail of a sleeping tiger, we can precisely guarantee what will happen later. But most likely it will not be pleasant.
In marketing, the introduction of LLM and agents of the increasingly autonomous the IA make it complex Mars in this way. LLM outputs are, for all practical, probabilistic purposes. The same notice will not give the same answer every time. It can give you something similar. But that’s not the same. If these outputs are used to make decisions or trigger results, then the chain of what happens is also probabilistic, even if the specific steps within the chain remain more deterministic IF-X-THEN-Y standards.
Most AI agents trust these LLM and their reasoning engines to boost their autonomous decisions and actions. They become more independent. Every time they have the ability to interact with each other, through protocols such as MCP and A2A. And the number of different small and large, independent, integrated, commercial agents that work in batteries are growing.
The resulting Martech battery becomes more like a tropical jungle than a Ferrari.
It is an ecosystem more than a machine.
This is not inherently a bad! In fact, a stack similar to the tropical jungle can be very adaptable, powerful and scalable. It can evolve more organically. But it requires marketing operations to evolve beyond the construction and maintenance of batteries and batteries of rigid and linear processes:
- Adopt a more modular and little coupled architecture for the battery, favoring open and interoperable applications and platforms, componable Martech, Marsch
- Improves the “observability” mechanism: transmission records, panels and anomalies alerts to identify when probabilistic results are derived from the expected limits
- Participate in safer experiments to miss, where new AI agents and agent capabilities are tested in small and limited pilots to learn how they behave (probably)
- Institute of Human Control Points in the Loop for Larger Decisions, such as the Hearing or the Budget Assignment to protect against biased or fugitive behaviors, but verify
- Create pods of “creator of meaning” interfunctional of small empowered equipment that include marketing operations, data engineering, IA/ml and compliance to evaluate and interpret emerging behaviors and make adjustments in the schedule
- Invested in greater expiration of data, especially with your underlying data infrastructure and data pipes to validate the quality of the incoming data, the drift of labels and the output coherence
- Promote a culture of experimentation: reward equipment to discover unexpected failure modes or novel victories, not only “hit goals”
To finish our explanation of the Cynefin frame, the fourth sense of sense of meaning, in the lower left corner, is “chaotic”. This is where it is simply not time to consider all possible causes and effects or how they interview. Instead, with limited information and a marked clock, you must take measures quickly, the best you can. The result is not predictable. But hopefully with the right heuristics, you will do it well.
The visual metaphor here is a flame building. Go to the exit as quickly as you can. Save the detailed analysis for the Marshal of Firefighters who investigates it later.
In marketing, it could be said that a completely not managed Marsch pile, with many chefs in the kitchen and an almost zero governance, is essentially a chaotic atmosphere.
It is important to emphasize that a complex environment is not an unmanty environment. It is only more probabilistic than we have used the leg with Marsh’s thesis adapts to a couple of decades.
Still because and we need to orchestrate experiences for employees and customs in a complex pile. But instead of a single “orchestra” teacher in the battery, we are likely to see multiple tools (applications, agents, platforms) that orchestrate different contextual experiences for those users.
For a while, we were converging towards large SAAS Marketing platforms and the IPAA automation platforms and leading workflow are the two types of orchestators. These platforms are still a large part of the Marsech battery and will certainly be the main orchestrators of AI agents.
But now there are many new categories of AI agents. The great attendees of AI turned into agents such as Chatgpt, Claude and Gemini. Specific roles agents such as 11x, Rox and Sierra. Builders of AI agents such as Crewai and Agent.AI. Agents based on the web browser such as Barten, induced and alone. And many frames, such as Langchain, Camel.AI and automate so that software developers build their own agents.
Things will be more complex. Again, that is not necessarily Bath. But it’s different.
It is more like a jazz improvisation than a precision Bach concert. Actually, more like a multitude of improvised jazz trios in parallel throughout the organization, instead of a complete symphonic orchestra with a single teacher.
This is the era of the great operations: to manage a spectacular volume, variety and speed of applications, agents and automation, all operating and interacting simultaneously throughout the organization.
That is why the future of Marketing operations and Martech management is brilliant.
Don’t forget Register in #Martechday To obtain the full report of more than 100 pages, this publication was extracted and captures our key discussion live or on demand. See you there!
