Informal market optimisation
The following does not use much technical terminology. Hopefully, this will change in future writings as I learn more about technology.
In India, hawker markets are common and superabundant. I want to call them "informal" and refer to markets with an online presence as formal. So, if you can verify information about a seller through some publicly available source, that is a formal market.
There is a problem of information asymmetry amongst consumers. They do not know which vegetable/fruit vendor is selling at the lowest price at a given moment and might be overpaying for items that are cheaper elsewhere. In fact this is probably true of every local market; not just vegetables. Bangles, souvenirs etc; I anticipate that selling price volatility and a market's informality move with weak correlation in the same direction. Thus, being able to collate all informal market information for a time period can help a customer find the best option for them.
A secondary factor, of course, will be other costs. First, direct costs involved in item acquisition such as fuel, oxygen quality, and time. Second, opportunity costs in terms of an individual's plan of action for the purchasing period--what if they want to check out specific different stores after buying this item? Or, what if they're not sure if do/don't want to check something out, but have a vague idea?
There are several variables involved in this decision. Usually, we make these decisions ourselves based on the endogenous variables that we can account for. However, important information about the environment is not incorporated, let alone aligned with our goals. Improving this decision-making process can make markets more efficient, traffic less troublesome and a lot of things a lot better: more people can do what they want to be doing. It would be liberating.
Most pivotal question: how do we acquire this information?
Answer: I think we start at using mobile phones as nodes. Or, we create our own nodes and plant them in some area. Use satellite data. Use the internet. Use social media. Collate data from across the internet, use artificial intelligence to process it and then convey the findings in natural language.
I'm probably still within the idea bubble. I'll let it pop in a few hours/days and then write again!