Think in systems. If the desired outcome is more honest and transparent behavior from job posters, the way to get there is public data and analytics to suss out signal of undercurrent intent internally. It’s an observability platform at its core.
Consider partnering with the hiring.cafe folks (search HN for relevant threads) if you want to achieve scale wrt feedback capture from applicants to drive analytics.
I’d be willing to expand my scope if there is interest by any group or entity to support such a project - as in, expand my pool of people actively seeking and applying for jobs (in my case about 10 per week, all within reasonable % of qualification) to explore if data reveals any patterns. It’s a nuanced thing. Basically poking a hole in the “meritocracy” narrative is potentially able to expose other bias at play (age, gender, ethnicity) with a willingness to put such findings into a well researched article / study.
Something like, when a company make a job posting they must put some cash in escrow for each candidate they are currently interacting with, and if they ghost you by not responding according to a timely schedule, you get the escrowed money.
Similarly when the job posting has received applications, that have not been rejected in less than a week, they are either autorejected, or they transition to escrow state where they must be handled or else cost money.
It would also have the merit of preventing companies to cast too big a net wasting everybody's time, just to collect data.
Wonder whether it's something that could emerge by people not applying to companies not offering ghost-protection protocol. Or whether name-shaming companies lacking basic courtesy is sufficient.
It could also maybe create more problems that it would solve, because it would add some intermediary in the recruitment process preventing peer to peer contracts. Which brings potentially many problems because it centralize applications for different companies but it also add levers to control bad practices.
Not sure how it could be regulated given international context.
No communication after 3 rounds, including with very senior people. Last interview was a month ago. Sent two requests for updates. Nothing.
If this helps eradicate the issue I'm for it.
The goal is to bring more transparency to the process, so if someone else had received a reply, I could have known sooner and moved on.
There's also a "Report Job" button in place for flagging invalid job postings, which you could add a note with the reason for reporting.
However, there is a caveat, since this is community-driven so like another commenter mentioned, there could be bad actors. One way to solve that in the future is to have community moderators / some kind of verification.
Oct 2023, 130 comments: https://news.ycombinator.com/item?id=37792507
So if I were in your shoes, I'd decide what problem I really want to solve, and build an app laser-focused on a solution for that problem. Bring everything else in later, once you know you have PMF for whichever core problem you choose.
I actually started this because I wanted to track my own applications but also wanted to solve the problem of making the process more transparent for applicants. Just like the "static HTML" part, a 2 weeks rule is a general guideline, but surely we can do better than that. Especially when there is no recruiters' contact after applying, going beyond a simple yes/no answer by providing community-driven data (historic and current) would help a lot (as seen in the links I cited). Thanks again for the insights!
When we submit our application, usually we also submit a resume. Personally, I do add UTM code (e.g. utm_source) to my links in my resume so I know if it was visited through analytic tools like Mixpanel. However, that's assuming that the person viewing the resume clicks on it.
> including if an LLM looks at it first.
Wow, do you have any links / examples of this kind of tracking or have you seen that done anywhere? If it is effective, I think it would be really useful!
If the noise floor threshold is too high, the signal you are looking for can't be differentiated.
Its not so cut and dry.
You can check out the GitHub repo too if you're interested!
There's a lot of time wasted individually by candidates vetting companies. For instance I've had to resort to using OSINT techniques and public records to weed out fake jobs because my cold interview->conversion rate is abyssal (1:1000+).
I've seen companies and positions being advertised where when you look up the secretary of state info, its either been shut down/suspended, or the directors/owners listed are dead people (i.e. semi-recent obituaries +- 1 year). The positions posted are always newer than the deceased person. I've also seen a lot of business info posted resolving to a UPS store address. Definitely presents as red flags when looking.
It would be useful to be able to crowdsource a lot of the due dilligence that must be done individually that could easily be automated as well as see the average turnaround time before rejection, and identify job postings that continuously roll (identical and reposted every 3-6 months)
Do you automate the whole process? That's really a lot of applications.
> It would be useful to be able to crowdsource a lot of the due dilligence that must be done individually that could easily be automated as well as see the average turnaround time before rejection, and identify job postings that continuously roll (identical and reposted every 3-6 months)
Currently the platform does this! On the company page, you can see job postings split into "Opened" and "Closed" statuses, as well as when people applied and the turnaround time.
I'm also looking into automated approaches like a browser extension which could be useful if more people uses it.
I'm basically looking for red flags, so inconsistency in website information vs. business and government records. This includes Secretary of State information which may need fuzzy matching, but includes Agent of Service, Address, and Company Leadership, as well as renewal dates, business status, renewal status. Legitimate businesses are not inconsistent here.
I look at the website for how they get customers, and I verify their contact information is functional (usually through a VOIP number). A business without a phone line is usually not a business. Having a listed phone in the reassigned numbers database is a pretty big red flag as well.
I'll look for social media business accounts, county court records, coroner reports, etc. blackbookonline.info and judyrecords.com have been my goto for awhile. The latter search engine was broken for awhile though because they provided too much info and were sued. Aside from that sometimes I get hits on OpenPayrolls.
Companies that hide salary/company name, and postings that were posted without disclosure by recruiters are pretty big red flags as well. I'm sure there's plenty else but that's what I've been doing.
Was the Michael Bazzell book you're referring to "Open Source Intelligence Techniques: Resources for Searching and Analyzing Online Information" or the "OSINT Techniques: Resources for Uncovering Online Information" (both around $44 on Amazon)?
> OSINT Techniques...
Yes that's the one. Its in its 10th edition now I believe, pubbed in 2023. ISBN 9798366360401
Thanks for the help!