Show HN: Fruits and vegetables in season in your zip code(harvestsignal.com) |
Show HN: Fruits and vegetables in season in your zip code(harvestsignal.com) |
I guess there's a reason we call this hunting season... deer, turkey,pheasent, quail, dove plus many non-game animals (hogs, etc) are in season in the midwest.
Some very angsty vegetarians on HN :D
One correction: 'Cantaloups' should be spelled 'Cantaloupes'.
It would be really nice to have the tags clickable (for example 'Peak', so that we can figure out what produce would be cheaper).
I'm curious where the data comes from. For my zipcode (souther New Hampshire) it said apples are in peak season. Peak Apple around here ends early to late October (depending on weather).
I'll have a look at southern New Hampshire specifically and see if anything weird is going on.
She spent three minutes on it and said “this is crap.”
When she entered our Dallas zip code, it noted that the nearest harvests for very basic veggies was over 270 miles away in the middle of no where. Meanwhile she’s growing the very same plants in our backyard.
Gardening is far from my area of expertise. But unless we’re missing something fundamental about the program, there may be some work to be done on this commendable project.
Next question: Is this public code?
I'd love to make a mobile app for it. Is it an abuse if I just pinged https://harvestsignal.com/api/v1/public/whatsfresh/[zip code]? I don't want to spam you but I doubt I'll have any traffic any time soon anyway.
Also lucky to be in a west coast state with lots of rain and lots of agriculture.
>Days left in season: 253
Artichokes, man. You just can't convince them to die.
Distribution is another matter entirely of course!
The last point is important to me. Food porn is common these days, they put out these recipes with beautiful pictures, but the recipe is so impractical I'll probably never make it. Frankly, if I never make the recipe... it's of little added value to me.
If it's not a pointless blog/article header, then the ingredients and method are split up across "tabs". So that now, when it says "melt the butter and whisk in half the flour", you need to flick back and forward multiple times for each step to remember what to do and how much to add.
or the food amounts to cookies dipped in honey dipped in butter dipped in icing sugar.
But you've got some work to do to get the data in there. I put in my Southern Indiana zip code and got 12 results back. Only one was actually anywhere near my zipcode (or even in my state). The rest were anywhere from 200 - 500 miles away. Sorry, but I don't consider Maryland to be local to Indiana.
I would suggest you work on getting data about local farms in there, and then tighten up the constraints for what is considered "local".
At the very least, in the results paragraph, change the display from "12 for your zipcode" which is just flatly false to "1 for your zipcode and 11 in the broader region" or something of that nature.
Also, a softer (lighter?) background might bring out the fruit/vegetable images a bit better.
1. There is a field to enter zipcode and empty space below.
2. I enter zipcode.
3. Empty space filled with fields for an email and name and a "we found results banner".
4. Because all the results were off my screen I figured it was nothing more than another email harvesting page.
5. Later I saw it on the front page of HN and tried it again. It still looked like an email harvesting page on my laptop. Only because I had seen the comments did I think it might be possible to scroll down.
It might be worth delivering value (results) before asking for value (a visitor's email). Fresh fruits near me is awesome. Signing up for email isn't.
Good luck.
One suggestion I have is this: it would be great if there was a way to "peek" ahead and see what other items will soon be coming into season within the next 1-2 weeks.
Not only was collecting the data time intensive, but determining the feature ingredient(s) was a challenge.
That's why my friend started Market Wagon[0] which acts as a middle-man between farmers who have crops and people who want to buy the food. The experience is like shopping on Amazon but you're buying local.
This is very location specific to NYC but I've always used this site to get local farmer's market info: http://www.whatisfresh.com/
There is also the problem that just because a certain produce item is being harvested within x miles doesn't necessarily mean that produce is available at local stores, and vice-versa.
My hope is that if I can get some early adopters, I'll be able to concentrate improving the data in a smaller number of targeted areas. This is the first time I've shown the site to anyone, so I'm hoping I can get some feedback to make the most effective improvements first.
http://msue.anr.msu.edu/news/predicted_2017_apple_harvest_da...
So "peak season" might better be interpreted as "as least one variety is in peak season", at least for now. And "peak season" is a bit nebulous due to how hard it is to normalise between different data sources, but it's something akin to "most active harvest period"
https://developers.google.com/web/fundamentals/design-and-ux...
Harvest: https://itunes.apple.com/app/harvest-select-best-produce/id3...
Seasons: http://www.seasonsapp.com/
Locavore: http://www.getlocavore.com/
Uh oh…
We keep a copy of the NC seasonal produce chart [1] hanging in the kitchen, and sometimes remember to check it.
Seems like one option for the OP would be to let users (moderators) update the site with vendors. That's a can of worms, but it could provide local information not available from online sources.
[0] http://www.ncagr.gov/markets/facilities/markets/raleigh/ [1] http://www.ncagr.gov/markets/chart.htm
The wording seems to imply that it thinks 22003 is in Georgia and the top result is the best, but it's in Virginia and the top result is the most distant one.
I'd also suggest making the on-page results a bit more prominent. My first reaction after typing a zip code was that I was required to give an e-mail address to view the results, and nearly gave up and left the page before realizing they were also displayed at the bottom. (The fact that only the "1 Result from Georgia" was above the fold didn't help, since it looked unrelated.)
I'll experiment with moving the email form to under the results, or putting it behind a button. I was trying to make it low friction, but I think I might have gone a bit far.
I'd be willing to pay a small monthly fee for this data being curated for me when you get the database a bit more fleshed out.
I have a suggestion which would make it very useful for me: hints on when to plant. Since you have the timing for "peak season" and (I assume) the growing times to get there, hopefully it would be easy to work backwards to arrive at planting times.
As a new gardener in Southern California, I could use some help planning my planting schedule. I can get away with planting a lot of things whenever I feel like it, but sticking with the ideal times would help optimize my yields.
At the time, I searched online, and found sites like LocalHarvest, Farmstand, RipeNear.me, LocalDirt, etc., each usually with a little different spin (some focus on specific food, some one farmer's markets, some on actual farms, and some on produce stands).
So you can see the challenges - data, use case, implementation, monetization, etc.
If it's from various feeds, they aren't standardized or consistent. They could update at different frequencies. They might omit huge swaths of geography.
If it's from the users themselves (i.e. you "check-in" when you find a local produce stand selling fruit or beef jerky or something), you need MASSIVE adoption and participation, and you'll probably have a 10-to-1 ratio of people that want to consume vs contribute. And how long is the data useful? Maybe they bought awesome watermelons but 15 minutes later, the last few sold out. Can you get people at the stands to update their listings?
What if you convince producers (farmer's markets, farmers) that it's worth their marketing dollars to get listed (and highlighted)? Does it follow the spirit of the app you envisioned?
And do you combine approaches? Multiple ways of getting data in, keeping it updated, etc? I'm curious where you'll take this, so create an "About/News" section of your site (or Twitter) and keep posting!
Edit: Also, I realize your approach "just what fruit is in season" varies from the above variations, and so the challenges and solutions do as well. I kind of went off on tangents there! Hopefully it gives you ideas anyway.
1) Some vendors of said produce. If I can get grapes that were grown 5 miles away, awesome! Now where can I buy them?
2) A filter that lets me determine the maximum distance allowed. I ran a search based on a Northern Virginia zipcode and it recommended me some Bell Peppers from Georgia...609 miles away. We're not exactly "local" at this point. Simply having a filter option to stop it from showing me anything farther than 50 miles away would be useful.
As soon as my wife heard about it - voice of joy!
Definitely solution to existing problem + tons of potentials for business (recipes, partnerships with stores, organics, deliveries)..
Couple of suggestions:
I signed up for the email, but I'd likely pay more attention to a monthly email as personally I get so many "weekly" email that I setup a filter consciously or not.
Second, `being` is misspelled in your post-signup message: "confirmation link to being receiving fresh produce notifications once a week." Might even be good to reword it a bit .:-)
My wife tends to give me a grocery list and includes things that - when i get to the store, are overpriced/look like crap as they're significantly out of season anywhere remotely close.
I visited the local farmer's market yesterday. Every 3rd stall had persimmons... Didn't see any in my Santa Clara County list.
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.
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 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!
Just my 2c.
> We found 9 results near your zip code...
> 3 Results from Michigan: (my county, my county, neighboring county...sure!)
> 6 Results from other states: Wisconsin, New York, Virginia, Maryland, New Jersey, New Jersey.
That's 3 results with 6 that are hundreds of miles away.
But yeah, I like it, but "NEEDS MORE INPUT"! :)
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.
So yes, every bit helps.
What gets me though is why Google suddenly felt the need to rank these sites over more established recipe sites that literally just gave you the recipe. I've been in digital marketing a long time and seen a few "waves" of major SEO updates come and go. When there seems to be a significant shift in quality for a category of results, Google does tend to eventually take action. So I wonder if we're going to see a correction at some point that stops favoring those sorts of aggregator blog posts and "long-form recipes."
The content being created is not valuable, and is clearly being done solely for rankings at this point.
If any Googlers are reading this, I'd love any additional insight as to why these sites continue to do well.
It works on a few sites I know are popular plus any that use the standard div with "itemtype" set to "http[s]://schema.org/Recipe", which seems to handle about 85% of the long-winded blogs out there. It extracts the recipe div and displays it nicely at the top of the page, dimming the site behind it. It has a button at the bottom of the recipe to dismiss the popup. Most sites include a handy print link or button at the top as well so you can quickly click through to that formatting as well (i.e. if you actually print recipes or use the Paprika recipe manager bookmarklet).
You can install the CRX directly: https://raw.githubusercontent.com/sean-public/RecipeFilter/m...
I will work on it some more to have proper controls, icon, work on more sites, etc then get it into the Chrome Web Store for easy installation.
Thanks for the final push!
Honestly, this is the sort of thing that once Amazon adds a Whole Foods API for ordering, you could add an unobtrusive "order these ingredients for pickup at Whole Foods" link or something at the bottom and hopefully make some commission off of it.
Ping me when it's built and I'll help you promote it if you're interested (I'm a marketer).
I have a clean site in a related space (https://keto.fm) and multiple subscribers have commented along the lines of: "this looked like a spam site, but once we browsed through it, it had rrally good content!"