Editorial
The Gist
-
Use case focus. Start with high-impact use cases that deliver ROI and build confidence in agentic AI.
-
Autonomy with boundaries. Define how much decision-making power agents should have, and build in safeguards.
-
People still matter. AI oversight, training and human collaboration are key to long-term success.
In part one of this three-part series on agentic AI in martech, we looked at the progression of technology in martech that has led us to the point we are at today. Now, agents are poised to be assistants to marketers. They can help with internal operations and also with external customer engagement tasks.
Let’s consider some of the potential primary use cases for agentic AI in martech. As organizations consider how to roll agents out, I think it’s easiest to bucket them according to internal and external categories. Internal agents are those that are part of a martech ecosystem but not customer-facing, and external agents are those that engage with the organization’s end customer.
Agentic AI Use Cases in Martech
Primary use cases are split between internal marketing operations and external customer engagement functions.
| Internal Agents | External Agents |
|---|---|
| Audience creation | Information and discovery |
| Journey creation | Service and support |
| Model creation | Conversational commerce |
| Content creation | Lead qualification |
| Decision determination | Orchestration and decision delivery/nudging |
While the list is much more extensive, this overview gives you a good idea of the difference in the two categories. No matter what type of agent deployment you decide to tackle first, there are some tips and tricks you can use to be successful.
Table of Contents
- Focus on Clear, High-Impact Use Cases
- Define Agent Roles and Boundaries
- Build in Oversight From the Start
- Make Sure Data Systems Are Ready
- Keep Personalization Useful and Seamless
- Track What’s Working and What’s Not
- Train Teams and Rethink Collaboration
Focus on Clear, High-Impact Use Cases
As with most things in life, developing a preliminary plan for success is critical. To gain momentum early, start with clear, high impact use cases whereby agents can deliver measurable ROI. These could be internal agents that score leads, personalized content delivery, email optimization, customer journey orchestration or real-time support via agents.
Initially, use cases should be considered where agents can augment or optimize a process rather than fully replacing it. Save replacement use cases for the point at which agentic AI has been nicely integrated into your tech ecosystem and the organization is comfortable with its workings.
Related Article: The Secret Ingredient to a Smarter Marketing Tech Stack
Define Agent Roles and Boundaries
As you begin agentic design, it’s important to consider the level of autonomy that the AI agent should have. Should they be informational in nature (i.e., only providing suggestions), decisioning in nature (i.e., able to make and automate business decisions) and/or action-taking in nature (i.e., able to execute decisions made with autonomy)?
As you account for this autonomy, include guardrails that support compliance and brand safety. Make sure decisions are explainable and transparent.
Build in Oversight From the Start
Something that must be stated but shouldn’t come as a surprise is the need for ongoing agentic oversight. As organizations roll agents out, they should create functions to continually review agent behavior and outputs, especially early on. Having the ability to override agentic action will be critical. With highly regulated industries, audit and logging capabilities for compliance will be needed.
Related Article: Balancing Agentic AI Autonomy and Boundedness in Contact Centers
Make Sure Data Systems Are Ready
Agents must be integrated with core data systems that contain clean, accurate and complete data. In some instances, this data may even need to be enriched, depending on what data collection techniques your organization uses currently. After all, agents are only as smart as the real-time data pipelines and systems they tap into.
Your organization will need clean, connected data pipelines and workflows. Ask your CDP, CMS, CRM and channel activation providers about agentic integration capabilities, including APIs or middleware provided out of the box.
Keep Personalization Useful and Seamless
As you deploy internal and external agents, they will eventually help your organization dynamically adapt experiences across channels using behavioral, contextual and intent data. As you move from broad personalization to individualization, keep certain needs in mind.
First, maintain a customer focus. Don’t just add agents to processes for the sake of automation, especially if they disrupt how customers can interact with your brand.
Next, make sure capabilities that are present in your customer engagement platform (i.e., testing, optimization, reinforcement learning and omnichannel orchestration capabilities) are still available within your agentic AI framework. If they aren’t, personalization could be difficult.
Track What’s Working and What’s Not
KPIs will be needed to determine efficacy of agentic performance. Consider metrics like engagement, abandonment, conversion lift, cost savings or operational efficiency. Further downstream, how do agents contribute to KPIs like improved attribution, customer satisfaction, customer acquisition costs and net promoter scores?
To measure the success of agents, consider how you will institute data collection processes and feedback loops to fine-tune models over time. After that, you can adapt agentic flows and processes.
Train Teams and Rethink Collaboration
This is perhaps one of the most important tips. Prepare and equip teams to work with AI agents. Redefine roles where agents take over repetitive tasks, and have humans focus on strategy, creativity and oversight. It’s important to have a large majority of the organization (i.e., IT, marketing, service, data science and legal) involved in enablement activities around agentic. Train on change management, how to use the agentic systems, how to co-create with AI and overall ethical AI literacy.
Webinar
Sep
25
CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms
Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.
Register
Webinar
Sep
25
Call Spoofing Trends Financial Institutions Can't Afford to Ignore
Don't let robocalls sabotage your customer relationships. Learn how to secure your voice channel.
Register
Webinar
On demand
Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential
Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.
Watch Now
Webinar
On demand
CX Reality Check: What Your Customers Are Really Thinking
Register now and get the insights you need to not just meet rising expectations, but exceed them.
Watch Now
Webinar
On demand
Content Leaders Collective: What a Good CCMS Actually Looks Like
Discover how content leaders transform clunky systems into growth engines — and why your competitors are ahead.
Watch Now
Webinar
On demand
From Hype to High-Impact CX Strategies That Actually Scale
Turn buzzworthy AI and outsourcing trends into measurable CX wins with fresh 2025 data.
Watch Now
Part three of this series will take us further down the path of agentic AI. I’ll detail some quick win use cases that brands can consider using as they get started with all things agentic AI. Stay tuned!
Learn how you can join our contributor community.
