If it can be measured, it can be improved.
Everyday we are living our lives. Immediate goals such as getting dressed, or making it to work on time, have their own motivation. Goals beyond the here and now take focus and dedication. Setting up feedback loops help - information on our progress can be motivating when we fall behind. Giving the right suggestion at the right time can be particularly effective. Part of this project is to look at identifying teachable moments and guiding the subject to the shorter path to their goals. Software/algorithms have the potential to turn streams of data into personal, relevant suggestions.
The inputs coming into the software should be diverse. Location is a personal favorite because of everything that can be deduced from time and location together. Purchases of all kinds are a very influential moment if one can be reached at the moment just before purchasing. This is especially true of purchasing food. Reminding one of the larger goals of healthy eating or staying within a budget can overcome an impulse purchase.
Inputs should be as timely, automatic and unintrusive as possible. All the data generated by someone from a cellphone and laptop computer can be cataloged and classified. Ownership and privacy are a critical concern. The benefits will be compelling enough to cautiously proceed with such collection.
Many inputs will not be available automatically, so a simple user interface will be important to collect this data. Examples are general mood, what is being considered for a purchase, or the kind of experience one wants to create in the next half day.
The cell phone or bluetooth earpiece or headset (google glass) is the avenue here. The software needs an avenue to intentionally interrupt what you're doing because the time seems right for a relevant suggestion.
If at the end of the month, by inputting data and considering suggestions, it feels like things are going better than they were before, or maybe better than they ever have been, then the system will be working.