SIGNIFICANCE: Cerebral tissue oximetry is imprecise and confounded by an uncertain and variable arteriovenous volume ratio. Venous saturation is better grounded in physiology. The superior sagittal sinus (SSS) is relatively large and placed under the open fontanel on the top of the head in newborn infants. AIM: To enable the development of a dedicated near-infrared-spectroscopy-based cerebral oximeter with sufficient claims on accuracy to be tested for benefit of clinical use. APPROACH: To set up a research agenda based on the combination of dedicated, high-fidelity digital and physical phantoms. RESULTS: A seven-step path is outlined to identify an optode geometry with high sensitivity to variation in hemoglobin-oxygen saturation in the SSS, with little confounding by changes in the optical properties of the skin and scalp or brain tissue, or in the width of the subarachnoidal space, and that is robust to variations in the placement of the optode. CONCLUSION: If an oximeter that is designed after exploration of digital phantoms can produce measurements in physical phantoms with good agreement with predictions, it will contribute credibility that cannot be achieved by direct gold-standard validation in newborn human infants.