Take the problem of predicting how much time people will spend using your product one year from now.

A macro approach is to break down all the product areas your company is investing in. Using some combination of projecting existing behavior forward and estimating the percent efficacy of each new feature in development upon usage, the numbers for each area are then summed up to get a prediction of how much time will be spent on your product.

A micro approach is to look at the different segments of individuals (perhaps someone who barely uses your product, someone who is an average user, and someone who is particularly active) and understand how your product fits into these people’s lives. Would the new features you are developing (or whatever other changes may be happening externally in the next year, whether it’s phone upgrades, more competition in your product space, etc) shift any of these individuals into the next bucket up and cause them to use your product more in their day-to-day? Could you see them giving up some of their time spent on something else in their life to your product?