Mobile application development has been a promising field in Software Engineering (SE) for more than the last decade. It is growing rapidly due to an ever increasing popularity of smart phones. Due to benefits it offers, SE in general, is trending towards Agility based approaches and distributed development setups. These benefits include development cost reduction, predictability, visibility, changeability and availability of expertise in specialized domains due to offshore teams. Optimal deployment of resources has a key role in the success of a project. To this aim, accurate effort estimation is a vital activity. Effort estimation method should consider artifacts such as complexity of tasks and prior experience of team members in the project domain. Agility and distributed development are two dimensions that are currently changing the development philosophies and shaping the development team profiles. Due to lack of experience in these two dimensions and ever changing nature of mobile application domain, any effort estimation method suffers from uncertainty. Thus the accuracy of effort estimation methods is comprised. This work ranks dealing with this uncertainty as a major challenge. To address it, the uncertainty is translated to associated risk whose consideration contributes towards accuracy of effort estimation. Thus we improve on risk-agnostic effort estimation methods by classifying and including risks in effort estimation, especially those which arise due to agility and distributed development for mobile application domain. Classification helps to understand the risk types involved and enables to better estimate their impact. Other than risk, we also consider novelty level and type of a project. Another contribution of this work is inclusion of quality aspect in software development, without which effort estimation methods map to single effort values. This limitation does not allow trading-off between chosen quality levels and corresponding effort values. Enabling software development at multiple quality levels overcomes this limitation and offers liberty of objective optimizations such as maximal quality and minimal time needed for project completion.