When you’re working on a routing model, at some point you will have to answer the question what the model should be using as a distance and/or duration measure. At this point, the inevitable discussion around using actual map data and incorporating traffic data will come up. We released a blog post with relevant questions and considerations around such a discussion and we would like to get your feedback on this. Is anything missing from our list of considerations? How did you come to a decision on what to use for your own models? What were the main reasons to use your current approach?
We appreciate your feedback and hope that this thread and the blog post can help others figure out the best approach for their models.