Why is it so difficult to justify resourcing costs for RPA?

I think there's still some confusion about what's RPA and how it might relate to other types of automation technologies, like AI-focused, API-specific, or Low-Code Application Platforms. Then there are some differences in terms of attended and unattended RPA. The former is often associated with personal productivity while the latter with large scale enterprise deployments. I think to overcome resistance one would need to establish goals and objectives that are easily understood and to look for opportunities for small successes to build on later.

Platform modernization, advancement and acceptance of the integration technologies — like Dell Boomi, Mulesoft, etc. — are driving people to look for a faster and foolproof solution with fewer exceptions. The scope of tasks qualifying for RPA is shrinking, resulting in projects with smaller cost-benefit analyses (CBAs). Now with the rise in salaries and licensing costs, the ROI realization will take time. 

Customers should not be thinking of maintaining in-house RPA teams; it is always better to go with a partner that can help deliver the automation objectives and goals. Having a completely outsourced or partially outsourced RPA team would help drive faster ROI.