Content area
The purpose of this action research was to address the challenges associated with vetting artificial intelligence (AI) project ideations having non-financial motives for their strategic appropriateness. Similarly, reviewed literature reveals that organizations struggle to define evaluative criterium concerning not-for-profit portfolio considerations. Furthermore, research concludes that competent PPM is insufficient in addressing the unique nuances of AI. Both analogous and responsive to these proliferating outcomes, the researcher capitalizes on a PPM mature, nonfinancial portfolio client aspiring to grow its negligible AI assets by 20%. The client would charter a participatory pilot study mandating several deliverables. Prescriptive to the not-for-profit evaluative problem, the researcher constructs a scoring model to discriminately screen AI ideations for portfolio inclusion. Consistent with appreciating the distinctive nuances germane to AI, the action research sponsors mandated a representative AI portfolio sample (Inventory of Work). 140 PPM practitioners, participating via field guided discussion (FGD), found the scoring model flexible and portable enough to screen AI initiatives, across diverse industries. However, the participant’s insisted on and participated in the construction of frameworks to test the model’s efficacy as a qualifying tool for implementation. Similarly, to facilitate reporting demands after PPM implementation, the researcher quantified reporting format preferences via the Most Preferred Reporting Construct (MPRC) given 172 practitioners using frequency distribution. Most definitively, project managers preferred within category component stratifications by 64%, strongly suggesting a preference for trade-off constructs. Conversely, 56% of project team members preferred component rankings across the portfolio, just slightly more than peer rankings. Similarly, 55% of project sponsors preferred ideation rankings across the portfolio more than within category stratifications. Lastly, standard deviations were calculated within subgroups to measure the dispersion in the reporting design preferences (Rayat, 2018).