Marketing project management has evolved through three distinct phases: waterfall task management, agile experimentation, and now AI-powered agents. In this article we cover the history, the problems with each approach, and what the current wave of AI agents in project management tools means for growth marketing teams.
People, Process, Technology
The People, Processes, and Technology (PPT) framework is widely used around the world to describe the 3 key elements of a successful company or team. Originating from the world of change management, and adapted from Harold Leavitt’s 1964 paper Applied Organisation Change in Industry, it has been adopted as a business practice “near-mantra” that is now widely used around the world in information technology, cybersecurity and management consulting. It has become the classic framework to assess and improve team and organisational performance.
Table of contents
Open Table of contents
- Marketing project management and the technology paradox
- How waterfall marketing project management became the norm
- The problem with waterfall marketing project management
- Agile marketing project management & experimentation
- The next shift: AI agents in marketing project management
- What AI agents mean for marketing teams
- Comparing AI agents across project management tools
- The right tool for growth marketing
Marketing project management and the technology paradox
If you were to ask most marketing leaders what makes a great team, most would probably say people, some may say technology. Very few would talk about process.
Speak to some of the world’s best growth leaders however, and you’ll discover they all talk about process.
Growth has nothing to do with tactics, and everything to do with process. Growth is not about the terminology or the tactics, it’s about a change in our mentality, process, and team.
Brian Balfour, VP of Growth @ Hubspot Quote source
Growth is about implementing a rigorous, customer insight and data-driven process with sustained effort to remove friction
Brian Rothenberg, VP of Growth at Eventbrite Quote source
The growth process is designed to be a positive feedback loop, to find small wins and optimizations across the business and then compound those over time as fast as possible.
Morgan Brown, VP Growth @ Shopify, ex Facebook Quote source
In knowledge work our tools are the processes we use to approach and solve different types of problems
Andrew Chen, VC @ a16z, ex Growth @ Uber Quote source
You can have highly competent people and state-of-the-art tech, but fail to grow and achieve KPIs because of inefficient processes. Process is the hardest of the 3 disciplines to get right, and its impact is often underrated.
How waterfall marketing project management became the norm
The inception of waterfall project management and task-based systems dates back to Frederick Winslow Taylor’s work in the early 1900s. His theory, known as Scientific Management (or Taylorism), centred around the idea that dividing work into standardised discrete tasks was the key to increasing process and project efficiency.
Waterfall project delivery follows a series of steps, executed in a linear fashion one after the other. The steps are typically Planning, Design, Implementation, Testing and Deployment.

-
Planning: Discuss campaign theme and objectives at a high level.
-
Design: Understand individual assets required (content, graphics, landing page & ad creation, tools, nurture sequences, analytics etc).
-
Implementation: Creation of the above assets, including design and development work.
-
Testing: Ensure everything works as planned, test user journey, device types, translations etc.
-
Launch: Push live and launch to prospects and customers.
The problem with waterfall marketing project management
The problem with task-based systems is that they work best for predictable, frequently recurring projects. Unfortunately, today’s marketers operate under conditions of extreme uncertainty and constant change.
Agile marketing project management & experimentation
Marketers need to reduce the risk of spending time and money on content that people don’t read, and on campaigns people don’t engage with.

-
Perform – ongoing customer research
-
Gather – insights to generate new ideas & opportunities
-
Prioritise – insights to find highest impact ideas
-
Test – run fast, agile tests to validate hypotheses
-
Analyse – analyse the results, what works and what doesn’t
-
Systematise – create repeatable playbooks for successful activity

The next shift: AI agents in marketing project management
Marketing project management is undergoing its biggest transformation since the move from waterfall to agile. Every major platform is now building AI agents — autonomous software that can analyse data, make recommendations, and take actions within your workflow.
This matters for marketing teams because AI agents are changing what project management software can do. Instead of passively tracking tasks, these tools are starting to actively participate in work — triaging requests, summarising campaign results, and suggesting next steps.
How the major platforms are approaching AI agents in 2026:
AI agents as workspace members
Most platforms now treat agents as first-class participants. Asana calls them “AI Teammates”, ClickUp calls them “Super Agents”, Notion calls them “Custom Agents”, and Jira lets you @mention agents in comments and assign issues to them. The agent has a profile, memory, and operates within the same permission model as human users.
No-code agent builders
Nearly every platform now offers a way for non-technical users to create custom agents. Asana has AI Studio, Monday.com has monday agents, Wrike has Agent Builder, and Smartsheet has Smart Hub. The pattern is the same: describe what an agent should do in plain language, pick triggers and schedules, and the platform wires it up.
MCP as the integration protocol
The Model Context Protocol (MCP) is emerging as the standard way AI agents connect to external tools. Jira has gone all-in on MCP — Rovo agents connect to third-party tools like Amplitude, Figma, and Intercom via MCP, with enterprises driving 93% of MCP Server usage. Notion lets Custom Agents connect to external apps like Linear and HubSpot via MCP connections. Basecamp took the most radical approach — “agent first, agent native” — where any AI agent that can run shell commands can operate in Basecamp via their CLI. Teamwork has an MCP server that lets external agents create projects, assign tasks, and produce reports.
Triggers: manual, scheduled, and event-driven
Most competitors now offer all three trigger types. Notion Custom Agents run on schedules or triggers. Wrike agents chain together and fire on workflow events. ClickUp Autopilot Agents handle both on-demand and automated tasks. The trend is towards autonomous agents that run without human intervention — but always with human checkpoints for critical decisions.
Multi-model access
ClickUp Brain lets users toggle between GPT-5, Claude, o3, and others. Notion offers a model picker. This is becoming standard — the agent designer picks the right model, or the user chooses.
What AI agents mean for marketing teams
The AI agent wave sounds impressive. Marketing teams should ask two questions before buying in:
1. Are these agents built for marketing, or bolted on?
Most AI agents in project management tools are general-purpose. They can triage tickets, summarise meetings, assign tasks — useful for any team. But very few can analyse campaign performance against a hypothesis, suggest the next experiment to run, or connect the dots between your analytics data and your marketing strategy. For growth marketing teams, a generic AI agent that helps with task management is nice. An agent that understands experimentation methodology is transformative.
2. Does adding agents increase or decrease complexity?
Every AI feature is another thing to configure, monitor, and debug. For horizontal platforms that already suffer from feature bloat, adding agents can make the complexity problem worse, not better. The best implementations are narrowly scoped — purpose-built agents that do one thing well within an opinionated workflow.
The real work is done in the shadows, alone, behind closed doors. Small, incremental improvements, as you put in the time.
Comparing AI agents across project management tools
How each major platform approaches AI agents:
| Platform | Agent approach | MCP support | Key differentiator |
|---|---|---|---|
| ClickUp | Super Agents — autonomous, long-running, memory-based | No (native) | Multi-model picker, Codegen acquisition |
| Linear | Agents as teammates — assigned to issues, mentioned in comments | Yes | Deep dev workflow (PRDs to PRs) |
| Monday.com | Monday agents + Sidekick | No (native) | Vibe coding (build apps in plain language) |
| Asana | AI Studio + AI Teammates | No (native) | Work Graph context, step-by-step checkpoints |
| Notion | Custom Agents — autonomous, scheduled, 20min+ runs | Yes | MCP-first external tool access, 24/7 agents |
| Jira | Agents in Jira — assignable, @mentionable | Yes (MCP-first) | MCP ecosystem, enterprise permissions |
| Airtable | Omni AI + Cobuilder | No (native) | Data-centric — operates across 100M+ records |
| Smartsheet | Smart Agents, Project Manager agent | No (native) | Enterprise governance, Smart Hub |
| Wrike | AI Agents — triage, intake, risk agents | No (native) | Agent chaining (link multiple agents) |
| Trello | Butler + Atlassian Intelligence | Inherits Rovo MCP | Simpler rule-based automation |
| Basecamp | CLI + Agent Skills — “agent first, agent native” | Yes (CLI-based) | Bring-your-own-agent |
| Teamwork | MCP server + AI assistant | Yes (MCP server) | MCP server for external agents |
| Planview | Anvi agents — portfolio insights, resource planning | No (native) | Cross-portfolio data fabric |
The right tool for growth marketing
Process isn’t glamorous, but getting it right has the biggest impact on your team and company growth. The evolution from waterfall to agile was about reducing risk through faster iteration. The evolution from agile to AI-augmented workflows should be about reducing the manual overhead of running experiments — not adding more features to configure.
For growth marketing teams, the ideal tool is one that combines structured experimentation with intelligent automation, purpose-built for the marketing discipline rather than adapted from a generic project management platform. To explore alternatives built specifically for growth marketing, see our guides to alternatives for ClickUp, Asana, Monday.com, Notion, Jira, Wrike, and other tools.
Growth Method is the only work management platform built specifically for growth marketers. We help companies implement a systematic approach to grow leads and revenue. Book a call today to see how Growth Method can transform your growth marketing efforts.