The need for quality assurance in image digitization programs has long been recognized by the cultural heritage and digital archiving communities. To that end, detailed quality guidelines, such as those from NARA and Metamorfoze, have been produced and shared within these communities as an essential first step in driving the adoption of i) quantitative quality metrics and ii) common approaches to the calculation of such metrics. These quality metrics are fidelity-based in that they address the question: Is the digitized image an accurate representation of the original content or object? Examples of such fidelity-based metrics are spatial frequency response (SFR) to assess sharpness, opto-electronic conversion function (OECF) to assess tonescale reproduction, and flat-field standard deviation to assess noise. These measurements are produced from captured images of one or more test targets, which serve as reference input signals.While these efforts are important and necessary, they are not entirely sufficient when it comes to implementing quality assurance programs in real-world production environments. In this paper, we'll discuss the next steps that are required to make quality assurance more than an academic exercise. These steps include fully automated, real-time quality analysis that integrates seamlessly and easily into a digitization workflow; simple and convenient quality metadata management; efficient exception handling to identify and fix quality problems; and quality metrics that go beyond current fidelity-based metrics. Our company, Certifi Media, develops technology and solutions to specifically meet these real-world quality assurance needs. In this paper, we'll discuss how a practical quality assurance program can be implemented with such technology and solutions.