I’m getting ready to launch my project (getanyplant.com) but I’m not sure if it’s really ready. It’s a site to help users explore and discover plants. It aggregates plant offerings and info from across online stores. I think the site looks good and works well but my concern is that it's too simple and that I should spend more time adding to it before launching. Current state is:
Plant database: about 6,000 unique across ~20 stores
Plantdetails: all plants have summary, fun facts, structured data
Images: about 50% have images, 50% don’t
Search: text search on name and various filters and categories (carnivorous, fern, etc.)
Wishlist: save plants to wishlist and add notes on saved plants
And that’s it. Some of the things that I would do with more time are
Images: Find more sources of images I can legally use
Monetization: no way to make money atm
SEO: Invest in doing SEO / content marketing
Stores: more integrations
Features: lots of ideas for different features
I’ve spent a long time on this so I’m eager to get something out. But I also don’t want to release too early and find that I don’t get any traction. What do you all think?
@mekon I think it's awesome - you should definitely launch!
Don't put a lot of pressure on the "launch". You should be doing many launches as you add new features and improve it. The main thing is that it takes time to build traffic so just launch and trust the process
@dbg Thanks for the encouragement! I think you're right. One reason I'm hesitant is that I'm also using this "launch" to more or less validate my idea. If it doesn't gain any traction it might be time to move on to something else. But I think I've delayed long enough.
@searching4the1 Thanks for your perspective, I hadn’t thought about it like that.
Also your site is so sick! I was into photography a couple years ago and I would have loved something like this. Might reach out after launch with a couple questions about how you’re driving traffic and whatnot if you’re open to it. Nice work.
@mekon Yeah, businesses are systems at the end of the day.
The idea of "validation" is really toxic, because it implies several untruths.
First, as a beginner it implies that if something "fails to validate" that you should give up or move on to something else. It implies that you need to have a "correct idea."
When the correct position is that with enough tinkering, you can make just about anything work to some degree.
Think back to when you were learning to program. Imagine if you had to give up the project and start on something else if you got an error! That would be a ridiculous way to learn. You didn't need to give up, just change how you approach different parts of your program until it worked well enough to satisfy you.
@madgirlsec Source data is being scraped from different sources - primarily plant shops, wikipedia, etc. I have a partially-manual partially-ml approach for labeling products as plants (using recognized scientific names for the most part) and using ML to extract info and summaries from that data to include on my site
@mekon Launch your website and let it develop naturally. Its simplicity is a key asset, and feedback from actual users will guide any future improvements
@mekon I think it looks pretty cool! But I am not able to see pictures of some plants like 'Philodendron gloriosum'.
I also have a question about integrating stores data / plant data in general. Do you go through each plant type and go through the stores one by one or do you have a pythonscript for that? How do you do this? I want to make some data engineering projects so I can put it on my resume before graduation this may, and I wonder how you continuously get the data. Thanks!
@tammy1957 That's one of the things I want to improve. Even though I'm bringing in plants and availability info from a bunch of stores, I don't use their images unless I get permission from them. Hoping to get more stores to agree or find additional sources of images. I'm using wikipedia as a source right now but there is room to improve that integration. As a result a lot of plants have no images.
I use python scripts for collecting the data and updating the product price and availability info! Web scraping is a lot of fun in my opinion and a good choice for a personal project.
@tammy1957 I'm doing a niche price comparison site right now. My plan was to manually keep track of all the site's prices. Because i have no idea how to automate that. But if you were to make a pythonscript or whatever for it for your resume, I would also be your best friend...
@mekon I think it looks great! Really nice work. You could probably use a SERP API and query building logic to grab the first image from google images to get the other images. What techstack did you use to build this?
@tyaandria Thank you! I appreciate that. I work as a data scientist so most of this was new. My main issue with images is finding ones that I can use (ie. CC licenses or getting permission) but I think that might be a filter on google images. Thanks for the suggestion, I'll definitely look into it!
Tech stack:
Front-end: React with Next.js. Deployed on vercel. Used Mantine for some components
