Design Data-Fueled Products So They Improve Experiences

Designing data-fueled products requires adapting those principles and processes in three ways: Educate, staff, and operate design teams differently; adapt design methods to machine learning (ML) algorithms; and use the cost of making a mistake as a guide for design. Blending together your design and data teams can be challenging. Keys to the success of both teams include battling groupthink, encouraging whistleblowing, and finding and fixing problems aggressively. When weaving ML algorithms into experiences, make sure to incorporate standard design methods by following a six-step method. Firms must weigh the cost of potential mistakes against the added value of human judgement when deciding if a product should be assistive or agentive.