Leaders must consider a build, buy or partner approach to implement a sustainable automation program. Given the rapid evolution of technologies in this space, this is especially critical for leaders driving change with a limited talent pool.
Build the technology using your talent bench
Access to talent that can develop and implement automation programs remains a concern for leaders. A recent EY survey of 200 senior leaders found that 56% see talent shortages as the single biggest barrier to implementing AI into business operations in 2018.
For most organizations, scarcity of automation talent will be the reality. Recruiting and retaining tech talent is challenging due to the competitive hiring environment and rapidly increasing salaries for professionals with automation expertise (e.g., machine learning, natural language generation, chatbots). Many forward-looking talent leaders have sidestepped the mature talent market and recruited raw talent to develop or hire from nontraditional backgrounds. Even so, automation talent is being hoarded and most organizations should expect to work with a scarcity of it — an eventuality that can be overcome with effective talent management and cultural programs.
Other companies have looked to upskill their existing workforce. In general, leaders should consider that upskilling technical talent is an option for incremental, not categorically new, skills. For example, statisticians may become data scientists, but data entry personnel will typically fail to transition to those roles. With the right underlying talent and skills analysis, a data-backed estimate of the expected talent lift through upskilling programs is possible. Given these challenges, many organizations turn to the contingent workforce pool and hiring temporary teams, but this brings its own trade-offs.
The cross-organization “center of excellence” (COE) model has not yet gained significant footing in automation; however, similar to analytics a decade ago, the explosion of technologies, thin vendor options and scarcity of talent will likely lead to a wave of implementations. Unlike analytics, this next wave of COEs will benefit from the learnings of previous groups.