Hacker News Reader

Comment by minimaxir | Hacker News Reader

Whenever a statistical report uses averages (and no medians anywhere), I always look for a sample size, since it gives an indicator on the potential variation for the true average. (Error is inversely proportional to square root of sample size)

So I checked the methodology and found something odd.

> The salaries included reflect more than 80,000 interview requests and job offers from the past year

Why are interview requests and job offers lumped together? Wouldn't there be disproportionally more interviews by definition? As a result, it's very hard to gauge the sample size of the salary data.

View on HN

Hi my name is Jessica and I am the data scientist at Hired who pulled the numbers for this report.

The way the Hired platform works is that companies must submit an initial salary offer when they start the interview process, and so all interview requests have salaries attached to them. Of course these offers can be negotiated later, but for the vast majority of our hires on the platform, the final base salary is equal to or higher than the initial offer given at the interview stage. Therefore we feel confident that this salary data is representative (if not a slight underestimate) on 2015 salaries for software engineers.

Give me a break. I have been using Hired for almost two years and everybody knows that number is made-up bullshit. It means nothing and everybody knows it!! You're slanting the numbers to make Hired more attractive. I love Hired. The company has totally changed my life for the better, and I could not be happier using it, but this is BS, and anybody that's used the platform knows this number that accompanies interview requests is almost nonsequitur to what actually happens.

Unnecessarily rude and combative. Flagged.

Well, sometimes the truth hurts.

This is Jessica, Data Scientist at Hired again...

I'm glad you've had a great experience on Hired and that it has changed your life for the better.

I'm surprised that your experience is that the initial salary offers (at the interview request stage) are significantly different than your final salary offer, because I can tell you that this is not the experience of most people on our platform. The majority of the final salary offers are equal to or greater than the offer given at the interview request stage. On average the final salary offers are slightly higher than the interview request offers across all Software Engineering Hires on the Hired platform.

Thanks for your reply, I appreciate it (and if I came across as rude, I apologize to you personally, that was not my intent). So what I'm saying is that companies come forward with a number slightly higher than the "base salary" on your profile. I've been using Hired nearly continuously for two years and interviewing with many companies in that time. I've experimented with that "base salary" number and it is almost totally the case that companies come forward with interview requests with a number slightly higher than this one. Some companies are more honest and give a number that they are expecting to pay, but by-and-large, companies just put a number slightly higher than the "base salary" there. I'm serious about the interviews that I go to through Hired, but very selective about considering any final offers. What I've found is that the salaries offered during this process are just wildly varied for whatever reason. Some higher, some lower, some about the same. I'm not sure if talent advocates receive these offer numbers from companies, or if they're only getting some of them via Hired's communication system, but there are many channels outside of Hired that companies use to convey their offers. It may be that some or even most people using Hired use it for one auction, possibly two, and then take a job and then maybe come back in two years, and their paper offers are similar or slightly higher than the number attached to their initial interview requests. But my experience with nearly continual interviewing via Hired is that the offers put forward by companies are almost never pegged to that number that accompanies the initial interview request. Especially startups are most guilty of lowballing this number, but I'm expressing surprise that these interview request numbers were included in something that comes across as a report on actual salary information, because it is 1) not actual salary information and 2) not correlated to the various offers I have received in a long time of interviewing via Hired. Again, thanks for your reply.

Are the base salaries of each applicant shared with interested companies? Or are they used to match a company with a potential candidate?

Are you referring to the "base salary" on an applicant's profile? That number is created/edited by the applicant and visible to everyone that can see the profile. AFAICT that number has an effect on the "potential salary" number companies give to applicants, but I do not believe there is an internal matching between applicants and companies based on this number.

Hi Jessica, thanks for the clarification.

That's a fair approach, although I agree that it adds a little bit of ambiguity around the process. As noted in other threads, there can be metagaming around the salary process on the employer side, so caution is still necessary.

I also believe it's skewed by the fact that it's only using the Hired platform and not including focused head-hunting of more senior people.

The numbers are more or less the same anywhere when you look at averages, if you look at glassdoor for bay area it's likely the same.

The way Hired works, an "interview request" is sent once by an employer to a potential candidate (along with a potential salary). If the candidate accepts, then they may interview many times with the employer and eventually get a final offer. (I work at Hired)

We're only interested in the salaries of the jobs we actually take. Final offers will be negotiated upward from the interview request. Interview requests where the candidate (or company) don't proceed won't.

So why use interview requests in the data at all? The fact that some joker company offered me 90k on hired when everyone else was offering 130+ shouldn't pull the average down.

Actually median also a not good metric when power laws are involved. For example, if half of the country is living below poverty line then you being above median still doesn't mean you have comfortable lifestyle. Tech people are often told that they have top 10% of salaries in the country but when you look at most of their lifestyles (i.e. below 1%), they are living in cramped apartments, doing 2+ hour commutes, reluctantly sending kids to schools they don't like etc etc. When most of the people in country have minimum wage below $10, even median would be pulled too low and that is one reason poverty line is defined such a way that these people don't come under it.

Reply on news.ycombinator.com