Passive, behavior data is always the best source to get a true view into what individuals are doing (vs. what they say they are doing). BagIQ takes passive shopping and restaurant data and turns it into nutrition data. Once the user can see the actual nutritional values of the foods being bought (and presumably consumed) then suggestions on better choices are available.
For any system like this, the quality of the detection algorithms and database is a key element to its success. The trial described in the article didn’t seem to push the boundaries of what is possible with BagIQ, but the initial results are very favorable. If it works it might make its way onto the PCPs’ prescribed app list.