Comment by Harj | Hacker News Reader

Comment by Harj | Hacker News Reader

Parent Full thread

I'm not sure exactly how your process works, from that description there's three ways we differ:

(1) We don't submit resumes to companies. We give them a profile of the candidate which describes their technical skills and their work history but without any mention of specific schools or companies. For a company to move forward with a Triplebyte candidate, they have to trust in our screening process more than credentials.

(2) Companies agree to not do any technical phone screens or coding challenges with our candidates. Instead they do an initial pitch call and then move to an onsite if there's candidate interest. This saves engineers a lot of time spent in repetitive phone screens.

(3) The metric we optimize for is our onsite success rate i.e. how often does a Triplebyte candidate onsite interview result in an offer. Since our candidates don't go through the regular technical phone screens, any improvement we can make on the industry standard of a 20 - 25% onsite success rate saves the company engineering time spent doing those phone screens. Currently we're averaging 2x that rate across all our companies.

> The metric we optimize for is our onsite success rate i.e. how often does a Triplebyte candidate onsite interview result in an offer. Since our candidates don't go through the regular technical phone screens, any improvement we can make on the industry standard of a 20 - 25% onsite success rate saves the company engineering time spent doing those phone screens. Currently we're averaging 2x that rate across all our companies.

I can see where that's valuable to the companies. What value are you trying to provide to candidates? I already know how to apply to companies, pass the phone screen, and wash out of the interview. If I knew how to pass the interview, I'd still know how to apply and pass the phone screen. What is Triplebyte supposed to help with?

If you are someone who behind the veil of ignorance can pass phone screens, then you will wash out 75% of the time through normal interviews. Your typical time cost is 4x offsite process + 4x onsite process to generate one interview. If you use triplebyte, assuming their interview is equivalent to a full interview cycle, your time cost is 1x offsite process + 3x onsite process - one for them, and two onsites with companies since your success rate is doubled. If you want to get more than one offer, the time savings increase further (and faster).

My success rate isn't doubled. Per Harj, they filter for people who will pass onsite interviews. My success rate at interviews is the same whether I go through Triplebyte or not.

> If you are someone who behind the veil of ignorance can pass phone screens, then you will wash out 75% of the time through normal interviews.

Veil of ignorance, huh? I'll run down the ways I've successfully gotten a job:

- Amazon (CreateSpace), by winning a contest they hosted. Multiple times. There was also an interview for this, though not of a problem-solving nature.

- eBay (Milo), by passing their online hiring challenge. There was no in-person interview for this; rather, they had me come in and work for a day.

- NCC Group, by passing their two challenges. There were in-person interviews for this too, largely consisting of them asking me if I knew how to do things and me saying "no". (I was told afterward that the reason my interviews had gone so oddly was that I had never provided them with a resume.)

Triplebyte themselves told me that I was exceptionally strong in "academic CS" -- the first time around. When they asked me to reinterview for their benefit, they highlighted it as a weak point.

My rate of success in applying to companies that rely on an interview instead of a project or other objective demonstration is 0%, not 25%. But I feel safe in saying that my interviewing problem doesn't lie in the fact that phone screens are hiding my basic incompetence from innocent companies.

One of the things they do is data-mining so they can match up candidates with companies where they're likely to perform well onsite (since different companies emphasize different things, Triplebyte can look through data on how their previous candidates have faired and find which companies will value your personal strengths): http://blog.triplebyte.com/triplebyte-engineer-genome-projec...

Harj- thanks for the thoughtful reply. Do you disclose numbers about what % of candidates that get offers choose to take them? I feel like that's pretty important for companies when considering time they save.

Reply on news.ycombinator.com