Hidden inefficiencies in organizational workflows are making work more challenging than necessary. According to IBM's research, 52% of employees report that their tasks have evolved over the past year due to AI's influence, yet many organizations fail to redesign work deliberately. Instead, employees adapt through workarounds, which often create hidden friction that slows down processes and limits capacity.
One major source of friction is the time spent hunting for information across multiple systems. Emily Lansdowne, AI Strategy Lead for Enterprise Transformation at IBM, highlights that highly skilled employees lose significant time gathering and verifying data, leaving less room for high-value analysis. This issue is exacerbated when ownership is distributed rather than assigned to outcomes, reducing the incentive to address underlying inefficiencies.
Another challenge arises when productivity gains in one function create additional work elsewhere. For example, increased output from one team can lead to more reviews and coordination for others. Lansdowne emphasizes the need to measure work across functions to identify these downstream effects and ensure that AI adoption focuses on quality rather than just volume.
The gap between official processes and actual workflows also contributes to friction. Workarounds, such as using spreadsheets or relying on informal knowledge, often indicate that formal processes no longer reflect reality. Observing teams in action can reveal these discrepancies and guide necessary adjustments, especially as AI agents begin to handle more execution work.