Thanks for the feedback - this is very first version, and there are a lot of improvements to make.
I think you're right about the wording, reading it with fresh eyes it is very confusing, and I'll change it ASAP.
Right now, all the farm information data is pulled from aggregated USDA data so there is a ton of work left to do to get better, finer-grained results. My plan is to look at which zip codes get the most queries and the most user sign ups, and improve results for those first.
The data wrangling aspect of this is definitely one of the biggest challenges. I've spent more time than I'd care to say extracting data from PDFs and other non-parser-friendly materials with regular expressions, and, often, manually.
Have you come across [Local Harvest](https://www.localharvest.org) before? That might be a good data source, at least for what farms exist where. I don't know if they collect data on what farms actually grow.
But I empathize a lot. I've pursued a number of side projects in this space. I've worked on an online farmer's market, a plant database, and a recipe database in the hope of solving a nexus of problems around eating sustainably and sourcing food locally. The only one that ever actually made it to live was http://www.fridgetofood.com and I haven't touched that in years, so it has decayed pretty hard.
I always got hung up on the data collection and translation aspect of things. It's not an easy problem to solve.
I haven't, but a few people have mentioned LocalHarvest, so I am definitely going to look into it.
Thank you for sharing your experience, it's really interesting to hear from other people interested in this space and I haven't had a chance to do so before, so I appreciate it.
Interesting project, not clear on the model for commercialization if any. Here in China we are looking at seasonality as a potential differentiator for our own food offering - http://infinite-food.com/
I think there are a few challenges with offering this as a commercial service. Firstly, the data. You stated you have a USDA data feed, that's great but is going to come away limited by its sources and I'd hazard a guess is largely going to be picking up sources that are the larger and more established versus smaller, more organic/seasonal farmers.
Second, the idea of locality. Right now I am typing this in Hong Kong and I can tell you basically nothing is produced here. So the locality model is 100% out of the question for many high density urban centers (which increasingly are where most people in the world live, usually in Asia).
Another issue from a conceptual standpoint is the high distance a lot of food travels in the US. Because of this reason the locality model is going to break down and the majority of the more established, larger-scale retail operators are sourcing in bulk from reliable supplies at distance rather than nearby vendors. What good is telling someone there are great local vegetables if they can't find them to buy them? Also, a half-price shipment may just have arrived from Honduras or Mexico.
Finally there is the reality of hydroponics. A great deal of some crops (eg. tomatoes, leafy greens, all baby greens) are produced in artificial growth conditions which are generally not seasonal or have exceptionally extended seasons.
If I were you I would think about the matches between data sources, potential demand and commercialization and see if you can't get in to the distribution ecosystem somehow (eg. 'supply chain transparecy') in more innovative places like Detroit where (according to media) a lot of this sort of thing is breaking down right now and there are probably good options for partnerships and cheap scaling.
Lots of people are trying to apply blockchain to agtech but the value add seems weak to me. If you went at the same space with a centralized model and a significant, customer-driven value add (short term/last minute deals for restaurants as per the recently fast growing and relatively new segment of hotels/travel, etc.) I do think this could have legs. But it seems there's a long way to go in finding and validating a model.
One way to look at it would be to have price and/or consumption data as your value add, another would be deal making, another would be qualitative assessment and supply chain transparency (some form of auditing/tracking). Another perspective to look at it from would be the players: restaurateurs, caterers, individual consumers, supermarket chains, farmers markets, etc. Many startups (eg. last mile food delivery) have done well partnering with restaurateurs who ware always under pressure and looking for an edge and immediate cashflow. There's surely opportunities in there, find them!
I think you're right about the wording, reading it with fresh eyes it is very confusing, and I'll change it ASAP.
Right now, all the farm information data is pulled from aggregated USDA data so there is a ton of work left to do to get better, finer-grained results. My plan is to look at which zip codes get the most queries and the most user sign ups, and improve results for those first.
The data wrangling aspect of this is definitely one of the biggest challenges. I've spent more time than I'd care to say extracting data from PDFs and other non-parser-friendly materials with regular expressions, and, often, manually.