Artificial intelligence is now part of the day-to-day marketing toolkit. Teams are using it to draft content, analyse customer behaviour, personalise campaigns and make sense of performance data.
The technology is developing quickly, but the question I hear most often is no longer, “How do we start using AI?” It is, “How do we turn that use into measurable business value?”
In my experience, the technology is rarely the main obstacle. The bigger challenge is fitting it into the way work gets done.
McKinsey’s 2025 global survey on AI also reflects this - While the adoption of generative AI was continuing to grow, only 21% of respondents from organisations using the technology said they had fundamentally redesigned at least some of their workflows.
That gap matters. Giving people access to AI is not the same as changing how a team works. For marketing leaders, this can be the difference between AI improving performance and becoming another tool in an already crowded technology stack.
Better tools cannot compensate for broken processes
Marketing rarely happens in one place. Campaign planning might sit in a project management platform, audience research perhaps on a dashboard, approvals in email and performance data in a reporting tool. Work also moves between internal teams, agencies and senior stakeholders, often with context being lost at every handover.
Adding AI may speed up individual tasks, but it does not automatically resolve the delays between them.
A team might use AI to create several campaign variations in minutes. That sounds like a productivity improvement, right? But if the original brief is unclear, the content goes through repeated revisions or approval sits in someone’s inbox for several days, the campaign will not reach the market any faster.
This is why I believe teams need to look at the process before focusing on the tool. Where does work slowdown? Why are decisions being revisited? What information is regularly missing? Which steps add value, and which exist because “that’s how we have always done it”?
The opportunity is not just to make one task faster. It is to improve the way work moves from beginning to end.
From task automation to connected workflows
The next phase of AI adoption in marketing needs to focus on connected workflows rather than isolated use cases.
Consider a typical campaign. It starts with audience research and then moves through strategy, briefing, content development, approvals, distribution and performance measurement. If each stage is owned by a different team or managed in a different system, valuable information can easily disappear along the way.
Research findings may not make it into the creative brief. Brand requirements may only be raised during final approval. Performance data may be captured in a dashboard but never used to inform the next campaign.
A connected workflow addresses these gaps. Audience insights are included in the brief from the start. Brand and compliance requirements are available when content is being created, rather than introduced at the end. Campaign results are reviewed and turned into practical recommendations for future work.
AI can support all of these activities, but the real value comes from connecting them through clear processes, reliable information and agreed responsibilities.
Ridhi Solanki, GenAI Project Director, Brandtech+:
One of the most common examples I see is the gap between campaign performance and future planning. A team may have valuable data showing which messages, formats and channels worked best, but those insights remain buried in a dashboard or end-of-campaign presentation. By the time the next campaign begins, a different team or agency may be involved and the brief starts from scratch.
Better process design means creating a formal step for sharing campaign learnings, giving someone responsibility for carrying them into the next brief and storing the information somewhere people can find and use it. AI can help identify patterns across campaigns, but those insights only create value if they reach the people making the next decision.
The aim is not to automate every part of marketing. It is to remove avoidable friction and repetitive work so that people can spend more time on decisions that need creativity, context and judgement.
Data, people and governance remain critical
Even the most capable AI system depends on the information and processes around it.
If customer data is fragmented, AI-enabled personalisation will also be fragmented. If brand guidance is outdated or difficult to find, generated content is more likely to miss the mark. If ownership is unclear, teams may not know who is responsible for checking, approving or publishing AI-generated work.
This is where data quality, governance and accountability become practical delivery issues rather than abstract principles.
Teams need to know what information an AI system can use, where that information comes from and who is responsible for validating the output. Human review is especially important when content includes factual claims, personal or customer information, regulatory requirements or anything that could affect the organisation’s reputation.
People remain central to the process. AI can help with research, analysis and execution, but marketers still need to understand the customer, question the output and make informed creative decisions.
That also means building AI fluency across teams. People need more than access to a tool or a one-off training session. They need to understand where AI is useful, where it is unreliable and what responsible use looks like in their role.
The best operating model is not AI replacing human judgement. It is AI taking on appropriate repetitive work while giving people more time for strategy, creativity and customer understanding.
Measure Impact, Not Just Adoption
Another issue I often see is that organisations measure AI adoption rather than its impact.
They track how many tools have been introduced, how many employees have completed training or how many use cases have been launched. These measures can show activity, but they do not tell us whether marketing performance has improved.
A better approach is to connect each AI initiative to a specific workflow problem and a measurable outcome.
Depending on the process, that could include:
· Reducing campaign turnaround time
· Cutting the number of revision cycles
· Shortening approval times
· Increasing content reuse
· Reducing duplicated research and production work
A baseline is essential. If a team does not know how long a campaign took, how many revisions it required or what it cost before AI was introduced, it will be difficult to prove that anything has improved.
Ridhi Solanki, GenAI Project Director, Brandtech+:
For example, “we are currently developing a creative solution for a large beauty brand that needs to produce high volumes of lower-complexity digital advertising assets. As part of the work, we began collecting data across the campaign process to understand where time was being lost. We found that the biggest delays were not in content production, but in briefing and approvals.
By clarifying ownership, standardising the brief, and using AI to support initial content development and quality checks, we reduced campaign turnaround time from three days to two and cut the average number of revision cycles from six to three. The key lesson was that AI alone did not create the improvement. The value came from redesigning the process first, then applying AI where it could have the greatest impact.”
Building a Marketing Operating Model for the AI Era
AI should not be treated as a standalone technology project. It is an opportunity to look again at how marketing work is designed, managed and measured.
I would start with a few practical questions:
· Where does work regularly slow down?
· Which decisions are repeatedly reopened?
· Where do teams lose information or context?
· Which tasks are repetitive but still require significant manual effort?
· Who owns each stage of the process?
· Where could automation reduce effort without weakening quality or accountability?
· How will we know whether the change has worked?
These questions help organisations prioritise real business needs instead of adopting a tool simply because it is new or widely available.
They also encourage better collaboration between marketing, technology, data, legal, compliance and business teams. AI projects rarely succeed when they are owned by one function in isolation. The process, data and expected outcome need to be understood by everyone involved.
The organisations that gain the greatest advantage from AI will not necessarily be those with the most tools. They will be the ones that connect technology to well-designed processes, reliable data, clear ownership and measurable outcomes.
AI can make marketing faster, more scalable and more responsive. But it cannot fix an unclear brief, disconnected teams or an approval process with no clear owner.
Sustainable value comes from improving how the work gets done first, and then using AI to make that improved process work better.
This article is penned by Ridhi Solanki, GenAI Project Director at Brandtech+
Disclaimer: The article features the opinion of the author and does not necessarily reflect the stance of the publication.