In a recent LinkedIn post, Dan Sherrard-Smith discusses a practical, three-step framework for integrating AI agents into content workflows, aiming to help businesses move beyond manual processes. Sherrard-Smith emphasizes a strategic approach that prioritizes process over tools, a crucial distinction in the rapidly evolving AI landscape.
The Importance of a ‘Playbook’ Over Tools
Sherrard-Smith’s core argument centers on the enduring value of a well-defined process, or ‘playbook,’ compared to the transient nature of AI tools themselves. He posits that while AI platforms may change, a robust playbook remains a valuable, long-term asset.
“The playbook is the asset you keep. The tool, e.g. Claude, is a rental. Tools will change. A good playbook should last 3+ years.”
According to Sherrard-Smith, businesses should focus on meticulously designing what they want the AI to accomplish and how, starting from their specific business problems rather than the capabilities of a particular AI tool. This foundational step, he argues, ensures that the AI integration serves a strategic purpose and is adaptable to future technological shifts.
Human Oversight and Iterative Development
The second step in Sherrard-Smith’s framework stresses the importance of maintaining human involvement, especially in the initial stages of AI integration. Rather than aiming for full automation immediately, he advocates for building the simplest possible version of the AI-driven process and ensuring a human reviews all AI-generated output.
“Don’t automate first. Build the simplest version possible. Review everything the AI produces on the first pass.”
This iterative approach, Sherrard-Smith explains, allows for refinement and correction, ensuring that the AI learns and adapts effectively within the defined workflow. He highlights that successful automation is built on continuous improvement and learning from initial outputs.
Prioritizing Momentum and Integration
Sherrard-Smith’s third step encourages a focus on achieving momentum by prioritizing implementation over absolute perfection. He advises identifying the smallest, most impactful playbook that can be made functional quickly, treating these as building blocks.
“Find the smallest useful playbook you can build. Then get it working. Think of playbooks like Lego blocks. Once the first one runs, snap the next one on.”
He references findings from The Growth Syndicate’s 2026 State of AI report, noting that while a significant majority of B2B marketers (86%) believe in AI’s benefits, a much smaller percentage (26%) feel their teams can effectively execute on it. Sherrard-Smith concludes that the bottleneck is not the availability of AI tools, but rather their seamless integration into existing operational processes. Therefore, he reiterates, a strong playbook is consistently more valuable than simply adopting a new tool.
Sherrard-Smith invites his network to share their own automation experiences in the comments, aiming to foster a collaborative learning environment within the business community.
📝 About This Content
This article is based on insights shared by Dan Sherrard-Smith on LinkedIn.
📅 Originally posted on July 6, 2026 | View original post on LinkedIn →