Successful AI investments depend on a deep understanding and clear definition of an organization’s specific workflows, enabling targeted decisions among automation, building, buying, hiring, or waiting based on workflow complexity and market maturity. By shifting focus from technology to workflow-centric strategies, companies can avoid common AI project failures and maximize business impact through precise alignment of AI solutions with where value is truly created.
The video emphasizes that successful AI investments hinge not on the technology itself but on a deep understanding of the workflows within an organization. Many AI projects fail because companies focus too much on the AI models or vendors rather than on how work is actually done and where value is created. The speaker highlights that workflows are complex and varied, even within a single department, and lumping multiple workflows into one AI solution often leads to mediocre results. Instead, organizations should dissect their work into distinct workflows, understand their specific needs, and then decide how to invest in AI accordingly.
There are five main levers for AI investment in workflows: automate, build, buy, hire, or wait. Automation is suitable for repetitive, pattern-based tasks with clear exceptions and easy validation. Building custom AI solutions is necessary when workflows are unique, complex, and require company-specific context, but this demands clear definitions of inputs, outputs, and success criteria. Buying off-the-shelf solutions works best when the workflow aligns closely with the vendor’s offering or when purchasing AI primitives that can be integrated into broader workflows. Hiring should be targeted to fill specific talent gaps aligned with workflow needs rather than searching for elusive “unicorn” AI experts. Waiting is a strategic choice to prioritize resources on high-leverage workflows and avoid premature investments in immature areas.
The speaker stresses the importance of clarity and specificity in describing workflows before investing in AI. Many projects fail because teams cannot clearly articulate what the work entails, what good outcomes look like, and who owns the process. This lack of clarity leads to poor vendor selection, ineffective hiring, and misguided build efforts. A well-defined workflow enables better capital allocation decisions and helps executives understand where to focus AI efforts for maximum impact. The video also warns against oversimplified debates about AI versus humans, advocating instead for nuanced discussions about how AI can augment human work and where talent should be developed or hired.
An investment matrix is introduced to guide decision-making based on two axes: how specific the work is to the company and how mature the market solutions are. Common workflows with mature markets are clear candidates for buying solutions. Common workflows with immature markets call for prototyping or waiting. Company-specific workflows with some market primitives suggest buying building blocks and orchestrating them internally. Company-specific workflows in immature markets require building custom solutions. Hiring decisions should be informed by workflow clarity, focusing on roles that enable defining standards and measuring success.
Ultimately, the video calls for a shift in how organizations approach AI strategy—from a broad, technology-centric view to a workflow-centric approach. Success depends on understanding and investing in the workflows that drive value, making deliberate choices among automation, building, buying, hiring, or waiting. By doing so, companies can avoid the common pitfalls that lead to failed AI projects and instead unlock significant business impact. The key takeaway is to know your workflows intimately, describe them clearly, and align AI investments precisely to those workflows for sustainable success.