Grats! A few suggestions from the guy who ran WorkZoo.com - also a job scraper, one of Time Mag's top 50 sites of 2005 - sold to Jobster in the same year:
-Republish the jobs in a directory by location on your own domain. Big SEO win.
-Scrape the big boards but link to them (as you already are linking to the source). They won't mind and when we stopped scraping Careerbuilder they called us and asked why.
-Try solving the location problem by getting a list of the 5000 biggest cities in the US and parsing the content for those cities. Filter out common english words from the list of 5000 cities though. Make sure you store lat/lon once you've tagged a job with a city so you can do radius search.
-Forget categories for the UI. Users can just use search to find what they want. If you need categories for your directory then come up with a list of 100 search queries and use that to generate 100 "categories".
-This really needs to be a search engine so shove it into sphinx or a similar fulltext engine. I used to use swish-e which sucked so badly - I envy the options you have for fulltext these days.
-Not really sure why you're limiting yourself to startups, but if this is just a fun project or a way to contribute to the community then that's the way to go. If you want to earn $$ then I'd index everything or perhaps the most lucrative sectors.
If you're serious about this be sure to chat to Dave McClure who used to run marketing for SimplyHired and take a look at simplyhired.com and indeed.com for biz model and implementation ideas. I'm also happy to be an advisor if this turns into a business.
I think the whole point is that some people want to go work for only startups. By creating a site where you know those are the only jobs, you attract that niche. The market for general developer job boards seems very saturated.
-Republish the jobs in a directory by location on your own domain. Big SEO win.
-Scrape the big boards but link to them (as you already are linking to the source). They won't mind and when we stopped scraping Careerbuilder they called us and asked why.
-Try solving the location problem by getting a list of the 5000 biggest cities in the US and parsing the content for those cities. Filter out common english words from the list of 5000 cities though. Make sure you store lat/lon once you've tagged a job with a city so you can do radius search.
-Forget categories for the UI. Users can just use search to find what they want. If you need categories for your directory then come up with a list of 100 search queries and use that to generate 100 "categories".
-This really needs to be a search engine so shove it into sphinx or a similar fulltext engine. I used to use swish-e which sucked so badly - I envy the options you have for fulltext these days.
-Not really sure why you're limiting yourself to startups, but if this is just a fun project or a way to contribute to the community then that's the way to go. If you want to earn $$ then I'd index everything or perhaps the most lucrative sectors.
If you're serious about this be sure to chat to Dave McClure who used to run marketing for SimplyHired and take a look at simplyhired.com and indeed.com for biz model and implementation ideas. I'm also happy to be an advisor if this turns into a business.
Best of luck!!