CMOs know that agentic marketing will bring whole new levels of productivity, but deploying and scaling it is another matter. It’s time to address the biggest barriers to agentic AI in marketing and explore how to implement the technology, step by step.
In the short time that agentic AI has been around, excitement for agentic use cases – from hyper-personalising campaigns to automating lead nurturing – has blossomed.
Marketing leaders have been especially drawn in. According to the Salesforce A-shaped CMO research, 85% of CMOs expect their teams to successfully implement agentic AI workflows within the next year – despite just one in 10 (13%) reporting that agentic AI has actually been deployed within the business.
That means marketing leaders are committing to a technology they have no first-hand experience with. This is a dangerous blind spot.
When asked why they have yet to adopt the tech, more than half (53%) of CMOs said the complexity of implementing agentic AI outweighs the potential benefits.
As agentic AI reshapes the business, it’s setting a new mandate for marketing leaders to become ‘A-shaped’ CMOs who bring the pillars of science and art of marketing together with the crossbar of operational excellence. An integral part of that journey is to embed agentic AI into their marketing operations.
Here are my best practices for implementing agentic AI in marketing, based on findings from the A-shaped CMO report in partnership with The Drum.
1. Demonstrate quick wins with agentic AI
A quarter of CMOs say they have never used agentic AI. Today’s marketing leaders need to get hands-on with AI agents to build a real understanding of what’s possible with the technology – and what isn’t.
The key to success is to start small and build over time. Here’s how to structure your approach to ensure short-term proof translates into long-term enterprise value.
Start by clarifying the customer problem you’re trying to solve – not just for the sake of efficiency, but to relieve a friction point in your buyer’s journey. Identify a single, repetitive workflow that would benefit from more speed, scale, or personalisation, and track your team’s baseline performance in this area.
From there, monitor ROI signals for 60-90 days to prove the use case’s value. By following an incremental rollout, you can test agentic AI’s capabilities and show stakeholders across the business the impact of the technology operating on the ground.
A phased adoption helps avoid the “pilot trap” – where brilliant projects are proven in isolation but never scaled across the enterprise. Compounding small, sequential wins proves the value of agentic marketing before investment is scaled – increasing the likelihood that your project makes it into enterprise-wide production.
| “One of the common challenges reported by CMOs is being asked to look at AI, but not being given additional budget. So one of the things that I’ve seen work is starting small and proving and quantifying that value early on. Success breeds success; it’s the incremental and marginal gains that build upon each other. Once you show that initial success, it starts to snowball. The board gets on board, and then you start to unlock more funding and support.” Azlan Raj, CMO |
2. Establish strong data foundations
Before scaling agentic AI use cases, you must audit the data foundations for your deployment. We call this assessing your AI readiness.
Understanding your data quality, governance, and security standards will guide the journey. You don’t necessarily need perfect data – very few organisations, if any, can say they have that – but you do need a connected, accessible foundation for your AI agents to use. Or at least for the use case you have in mind to test.
Without robust data governance, businesses using agentic AI risk security breaches, compliance problems, and customer distrust. And their AI agents simply won’t be able to make smart, accurate, or trustworthy decisions specific to their business context.
Follow these best practices to clean and standardise your data:
- Mask sensitive customer information – such as names, bank details, or medical records – before it’s used by an agent. All sensitive data must be protected throughout the development lifecycle to maintain its integrity and avoid exposing customer details.
- Clean, connect, and harmonise your siloed data sets in one place, ready to feed into your AI and agents. AI is only as good as the data it has access to. Centralising your data reduces that risk of fragmentation, with your model getting the full and accurate context of your business and customers.
- Implement ethical data practices by being transparent with customers when their data will be used to train AI models, ensuring data privacy compliance, and fostering a culture of safe, guardrailed AI adoption. For example, you may establish a clear opt-in/out mechanism when customers are asked for information.
The role of the human in the loop is to ruthlessly govern the data AI agents have access to. Rather than attempting to integrate all available data, focus on what will drive the most value early in deployment – and minimise the time spent cleaning and contextualising data to be used in AI workflows.
3. Map your team’s skills – and build upon them
Agentic marketing requires a skilled, AI-ready workforce to govern, orchestrate, and audit agent outputs. That’s an opportunity, not an obstacle. An opportunity for marketers to learn new, AI-ready skills, future-proof their careers, and scale AI that accelerates their work, rather than replaces it.
The transition demands marketing leaders rethink how they train the workforce. Teams need the skills to orchestrate processes using AI agent workflows and prompt engineering. Marketers must develop a matrix of skills across:
- Human skills: Creative direction, stewardship of the brand voice, customer empathy
- Agent skills: Prompt engineering, performance optimisation, output critique
- Business skills: Campaign strategy, data-driven decision-making, ROI storytelling
When employees feel confident to guide AI agents, adoption grows, output accelerates, and capacity is freed to focus on more impactful work. They spend less time on manual campaign builds, execution, and reporting – and more time handcrafting the customer experience.
Agentic AI will reimagine every level of the marketing team’s structure. CMOs will need to pivot to address their own skills gaps – because nobody is truly fluent in a technology evolving at the pace of AI.
| “The CMO role needs new skills – leading hybrid teams of humans and agents, strengthening cross-functional alliances and architecting integrated and automated workflows – that unite brand and demand within one seamless growth ecosystem. A growth mindset will be critical for CMOs to rethink their role to bring together the best capabilities of people and AI, and set new standards for marketing.” Jo Pettifer, VP Field Marketing Marketing UKI, Salesforce |
4. Connect with the board
According to Gartner, CMOs are viewed as the least AI-savvy members of the C-suite. Just 15% of surveyed leaders graded them as competent with the technology, compared to 44% of CIOs and 40% of CDOs.
There’s a perception gap between the board’s vision for AI in marketing and the operational reality.
To close the gap, CMOs must develop and showcase their AI credentials, all while building strong alliances with their executive peers.
- CTOs and CIOs are critical relationships for CMOs. They’re responsible for co-designing the organisation’s AI roadmap, ensuring seamless integration with the tech stack, and realising growth with AI.
- CFOs are an entry point for CMOs to ensure the board is bought into the value of agentic marketing – and unlock future funding.
CMOs need to focus on building these relationships in the context of agentic marketing. They must progress from defensive reporting to proactive, board-level leadership that rivals their cross-departmental peers. In practice, this will demand that they translate marketing KPIs into tangible, defensible business ROI like Customer Acquisition Cost (CAC) Payback, Pipeline Contribution, and Net Revenue Retention (NRR).
Greater integration between marketing, sales, product, and finance data will also help CMOs shed the perception of being isolated leaders. Identifying use cases that touch multiple departments, often by simply putting the customer at the heart, is key to unlocking cross-functional buy-in.
Marketing leaders aren’t on an island – they’re a core business driver. Positioning as one will go a long way towards securing support for agentic marketing initiatives.
5. Build a hybrid team with advanced agentic use cases
As agentic AI redefines ways of working, CMOs must adapt to managing a blended workforce of humans and agents. Leaders will need to clearly define roles and responsibilities across the marketing function – identifying areas where human judgement adds unique value and where AI agents should operate autonomously.
Often, this involves moving humans up the value chain from manual execution to strategic direction, orchestration, and oversight of AI-generated outputs. The goal of this revised structure is to liberate human talent from repetitive, administrative tasks – such as data input or analysis – to refocus on high-value strategic and creative work. It’s about augmenting human talent.
These are the tasks that rely on uniquely human skills – empathy, ethical judgement, relationship-building – combined with agentic automation to drive productivity and personalisation at scale.
To unlock these benefits, organisations must identify the agentic marketing use cases best aligned to their needs, such as:
- Hyperpersonalised communications: Automatic generation, testing, and refinement of customer communications. Hyperpersonalisation across channels enables marketers to deliver highly targeted communications that engage specific audience segments without the heavy lift of data analysis, message testing, or copywriting.
- Automated lead nurturing: Immediate engagement, qualification, and response to pipeline opportunities. Rather than humans mapping prospect journeys, AI agents automatically progress high-value leads through the pipeline to accelerate sales cycles and improve conversion rates.
- Optimised media buying: Continuous, cross-channel tracking and real-time reallocation of digital ad spend. While marketing teams typically need to manually adjust bids to prevent budget wastage, AI agents instantly pivot strategies based on live competitor moves or fluctuating inventory levels to maximise ROAS and reduce latency.
Leading organisations are seeing massive gains from agentic marketing. Formula 1 uses Agentforce and Data 360 to unify real-time insights for 827 million global fans, fuelling a 22% boost in engagement. F1 can now deploy up to 5 million hyperpersonalised push notifications in under 90 seconds – translating agentic speed directly into massive audience value.
Agentic marketing is unlocking human ingenuity, removing operational friction, and elevating marketing leaders into the A-shaped CMOs of the future. It’s transforming how marketing works from the ground up – and the leaders who embrace this new paradigm today will lead the marketing organisations of tomorrow.
Learn how to overcome barriers to agentic marketing and become an A-shaped CMO in the full report by The Drum.
The A-Shaped Framework
Master the A-shaped Framework and lead the shift to agentic marketing.
