In 2016 Niraj Shah and Steve Conine, founders of online home goods retailer Wayfair, are faced with a decision about how to improve user experience on their e-commerce sites. A key driver of consumer interest and conversion to purchase in the home category is visual imagery, which has traditionally been generated not by the retail channel rather than by manufacturers. Catalog retailers, in particular, are famous for visually romancing the category. As a comprehensive offering, Wayfair hosts over seven million SKUs on its site, so it must solve for the lack of visual assets from manufacturers by generating visual assets of its own. Initially, the company embarks on a plan to produce these images using traditional photography studios and professional staffs, but costs mount and throughput is too slow. Seeking a solution, Shah and Conine assess the likelihood of using technology – specifically, high-resolution 3D modeling from low-resolution 2D images – to address the challenge. Resolving this issue is especially pressing, given that Wayfair has just launched private label lines and a host of what it calls lifestyle brands, which require support from high-end visual merchandising. As a consequence, the question of how to provision visual assets becomes a question of how fast Wayfair can grow. This case presents the factors involved in deciding what direction to take – a reliable path using traditional methods or a risky path using bleeding-edge technologies.