Hacker News new | past | comments | ask | show | jobs | submit login
Launch HN: Weekday (YC W21) – Hire engineers vouched for by other engineers
98 points by amitsy on July 15, 2021 | hide | past | favorite | 64 comments
Hey HN. We are Amit, Chetan, Anubhav and Hari, the cofounders of Weekday (https://www.weekday.works). We are building a recruitment platform centered around recommendations. We help companies hire engineers who are vouched for by other engineers.

Discovering top talent is challenging. It's easy to check for great credentials, but that's not the same thing as on-ground reality and achievement. It works the other way, too: someone with credentials may not be a great contributor, and screening for credentials eliminates many contributors who are.

Our company is based on the insight that great contributors tend to know other great contributors and are in a position to recommend them. We believe that a network-driven approach can help to discover high-caliber candidates.

We chanced upon this idea when we were doing our previous startup where we found it excruciatingly difficult to hire our first engineer. We tried all job boards and tech hiring marketplaces but nothing worked. We asked 50 of our friends if they knew a good engineer for us. Many said that they would recommend someone, but few actually did. We realized that it was because the friction to recommend is too high. Instead, we asked those friends to do a screen-share call and scroll their LinkedIn connections and just tell us which ones they would recommend. This ended up increasing our pipeline multifold and is what gave us the idea to productize the same approach and try to make it work for a global network.

In an attempt to productise the same approach of hiring via recommendations, we have built a Chrome extension which removes the friction of coming up with names to refer. We give people a shortlist of people to choose from (based on a matching of roles we have available and your connections’ profiles) and they can recommend anyone they like out of them. They get financial rewards if the people they recommend end up getting interviewed by companies or get hired. Our business model works on a success fee model, we take 15% of annual salary from the company’s side. We pass on a percentage of that finder’s fee to the engineer who recommended them, while also paying if any of your recommended friends get interviews (as we have seen on avg companies take 20 interviews to hire 1 person).

We have experienced that in hiring, neither a fully automated nor completely human-driven approach works. Having to interact only with software can be a dehumanizing experience. A fully human-led approach leads to a lot of recruiter spam as every recruiter in their silo tries to reach out to every possible engineer. We believe our approach of hiring based on recommendations leads to more targeted matching and gives opportunities to people who otherwise would find it difficult in an overly credential-driven job market.

We would love to get your feedback on our approach and hear about the problems you have faced while hiring, looking to switch jobs or interacting with recruiters!




When I "vouch" for a friend, what happens on the friend's side? Am I opting them in to getting email(s) from weekday?

If so, that seems slightly odd that I'm going through my top 10 friends on linked-in, vouching for them so that they are now going to get some a cold email(s), but I have no insight if those 10 people are looking for jobs and want this email at all. I'm pretty much saying to my friends "Hey, I just gave your info to this random company so hopefully I'll make some money off you." (And yes, it is true that I'm "helping" them get a job, but the people I'd vouch for don't need help, in this market they can get a job very quickly)

I will admit that some aspects of what I mention above are true in the "screen-share call" example you gave above, BUT it's the fact that the screen-share alternative is clunky/painful which makes it more socially acceptable. The automation/scale of your new approach starts to feel more spammy.

When you say "No recruiter mails" on your website, what does that mean exactly? Who is the subject of that sentence and who is the object? Subject: weekday recruiter? in-house recruiters at tech companies? contingency recruiters? Object: me the voucher? the person I'm vouching for?

Note: I'm not trying to be mean or negative, I'm just trying to understand the full feedback loop, so I can be empathetic to all parties, especially my friends. :)


The recruitment industry has always had parallels to dating.

We have two parties who don't want their time wasted, but want to get together as quickly as possible if it's a good match, while maintaining their dignity and their privacy.

The industry is permanently grappling with ways to make this process scalable. The road is littered with the dead hulks of companies that felt like they had cracked the code. Yet still a huge chunk of the industry belongs to plain old middlemen - recruiters - or even to word of mouth and other age old human behaviours. That reflects the fact that so far, no-one has really cracked the code, and there is often still benefit to both parties in having someone in the middle.

All of the questions that you ask are totally valid and can be viewed as just part of the dance of bringing job and talent together.

My personal belief is that the solution is out there, it's just quite complex (human are complex). It probably involves:

- karma of some kind (randos can't arrive and start pushing their friends/colleagues in front of employers without restriction)

- rate limits (if you've put forward 20 people, maybe you need to slow down until some of them have been "processed")

- candidate care limits (employers probably can't access more "candidates" until they have courteously dispatched any existing ones by hiring them or providing a formal rejection, ideally with feedback).

- saving face (graceful ways for candidates to be told their salary is out of whack, for them to push back on referrers who are spamming them out too widely, for employers to make candidate feel they were a good second place, as opposed to a failure, etc. etc., all the social lubrication that makes the world go around.

In short, the whole area of recruitment will always be fraught and fought over because there's no much damn money to be made.

But no one IMO will meaningfully "win" in this area until they deliver a platform that has deep, rich set of human-oriented behaviours and functionality that really dig in deep to what it means to be a candidate, an employer, a referrer, and treat everyone with courtesy and (yes) financial reward as required.

Just as StackOverflow became successful because it catered to the exact question and answer communication patterns that are suited to programmers seeking help, some recruitment platform will succeed because it caters to the communication patterns that are associated with gigs finding talent and talent finding gigs.

Source: many years spent building corporate recruitment systems.


This is awesome! There are so many gems here that even if pick up 50% of it and deliver on it, we would be golden. This is probably the most nuanced advise I have received on the sector/our approach and I would have spoken to 100+ VCs and other people doing recruitment. HN is great


Good luck to you. I realized I wrote "no much money" to be made, whereas I meant the opposite, so go get it.


This is a really interesting and thoughtful read, thank you.

I'm curious why solutions that satisfy those bulleted items you mention don't exist? Whether for dating or opportunity-talent-matchmaking.

My hunch is that there's some perverse incentives that keep us stuck in a local maxima. I don't use any of the dating apps, but I would think it's fair to say that the popular apps and networks are an extremely soul-sucking, relentless grind.... but yet, they stay popular, and innovative alternatives aren't really taking off en masse. Why?

I guess here are a few thoughts:

- This reflects the absolutely massive asymmetries and inequalities that infect everything. It sometimes seems like there is no "average" anymore - everything is bi-modal. There are the FAANG total comps, and the total comps of everyone else - with a huge gap between them. Massive wealth gap for people have have been invested in real estate and/or stocks for a while, and those that have not been. And in dating: The massive gap in desirability between young or attractive women - and everyone else (for example, it's not like an attractive 25-year-old woman gets twice as many messages as, say a 25 year old male.. she'll get literally a thousand times more)

- In a globalized, massive, and mostly anonymous marketplace, there is little incentive to constrain oneself to cooperate. And this includes preferring free-for-all networks over ones that "enforce" cooperation. For recruiters: I suppose spamming endlessly, while not pleasant, is probably a local maxima. Being a cooperative agent (reducing spam, investing more in each possible connection) just reduces their chances of getting a catch. I'd imagine a similar things for males in a dating market.


Personally I NEVER want my vouching for or receiving of vouching to be automated - by definition that's not vouching. It never can be.

I can never trust anything but the person with whom I have a history.

I would never drop an endorsement and stand behind it without that history and assurance that I was right.

There are some companies that I would never hire an ex-employee from unless I had two people I've know for a while vouching for them - it's their reputation and our history together that justifying my trust in their endorsement and that can ONLY be an interpersonal relationship as its basis. Anything else is just some random acquaintance or stranger making a claim.

It's ONLY the personal history that enables vouching to work. And it's the EXISTING relationship that is more important than the welfare or benefit of ANY vouchee!

Some things simple DO NOT NEED to be automated and probably SHOULD NOT be.


> My personal belief is that the solution is out there, it's just quite complex (human are complex). It probably involves: [...]

I'd say, anything that helps connect people in genuine relationships without getting in the way.

... which is not easy, since relationships aren't easy. Many of the features you mention (karma, rate limits, care limits) seem to be trying to limit the downsides of failing to cultivate genuine relationships.


Very legit point! That is a concern that some of our prospective users do share. What they do in that case is vouch for ones who they know are looking out or can benefit from vouching (eg. folks who struggle at DS/algo interviews etc).

On the "no recruiter mails", I just realised that it might be confusing. What I meant that just because you vouch for someone, it doesn't mean that they will get bombarded with spammy recruiter mails. I think that means object is the person you are voucher for and subject is contingency and in-house recruiters at tech companies.


Congratulations on the launch!

A headsup in case you are not using official LinkedIn API, LinkedIn is nuclear on browser extensions operating on its site and would even ban the users using the extension[1]. They actively scan for browser extensions and regularly update their blacklist.

[1] https://www.linkedin.com/help/linkedin/answer/56347/prohibit...


As much as I'd like to give this a go, my findings so far is that recommendations from anyone you don't fully trust already are as broken and prone to gaming as any other indicator in the job market.

As (almost) always, it's so ubiquitous that the German language even invented a word for that fact: "wegloben". Loosely translated as: praising someone away.

Why is it broken? Because while it's certainly possible that someone would recommend the best people they've ever worked with in the rare case of a sudden bankruptcy, chances are usually far, far higher that people try to recommend the weaker ones they had to let go in times of hardship to others, simply because they want to be nice and soften the blow.

Been there, done that. Multiple times.

If there's _really_ someone excellent, you move heaven and earth to keep them around in your own team.


It is an excellent point and very similar to the experience we had seen when we first started with this approach. We have seen that once financial incentives and limits on no of people whom you can recommend, were introduced, people were not just praising anybody and everybody.

Another interesting phenomenon we see is that people don't generally recommend people from their current emplpoyer, they generally do from previous ones or their college friends.


Let's say that you think highly of Ivan, who works at FooCo as a C++ developer.

When Ivan starts recommending Paula for her C++ skills, and Paula used to work at Bar, LLC where Ivan used to work, you can trust that to a certain extent.

When Ivan recommends Istvan for his Haskell skills, and they've never worked together, Ivan's recommendation might be worth less (he's doing a favor for a friend) or worth more (Ivan is recommending the leader of a local Haskell User Group.)

When Ivan recommend Maura for C++ skills, and Maura works at FooCo, that could mean:

Ivan sees Maura as a threat or an irritation and would like her to go elsewhere - low value

Ivan is getting ready to leave FooCo and is pre-emptively arranging exits for people he likes - medium value

Ivan is getting ready to leave FooCo and is trying to do as much damage as possible on the way out - high value

Ivan sees that FooCo doesn't have a place for Maura to grow and wants her to thrive - high value

People are complex.


> "wegloben". Loosely translated as: praising someone away.

Wow, that's fabulous.


When I've made recommendations in the past it's usually because I understand the needs of both sides. It sounds like once I as a scout refer my top 10, you prospect them for all of your companies, not just one. Is that right? Also curious about pjam15's question about whether the people you recommend know that we were the ones that recommended them. Also, how do you know if my (or any other scout) has recommendations that are any good?


Yeah, that's how most current recommendations work (when one understands the needs of both sides). But we feel that it puts too much friction to recommend.

"It sounds like once I as a scout refer my top 10, you prospect them for all of your companies, not just one. Is that right?" - yes

"Also curious about pjam15's question about whether the people you recommend know that we were the ones that recommended them" - you can choose to make yourself anonymous selectively. for example, for people you know very well, you could choose to have your name mentioned. for acquaintances, you could choose to be anonymous

"Also, how do you know if my (or any other scout) has recommendations that are any good?" - Well, this is a little bit tricky. We currently just pass on the information while internally maintaining a score of how many of your recommendations have received how much interest from companies and what the interview experience was.


Cool, on the first point, that's fair. I've also been at startups where they ask us to just list all the interesting people we know and would like to work with and then they just take that info and prospect them independently; sounds like you're just productizing that process which is great.

On the last point, interesting, I guess you will have to take a lot of time at the beginning to do your own screening of all the referred people to make sure they meet your standards before you decide a scout is any good and needs less supervision than others.

2 other things:

1) It says only software engineers can apply. I sort of understand but wonder if you'd consider making exceptions - I was early on at Gigster (another YC co but doing software development as a service with a network of freelance engineers) and some similar co's and so although I'm not an engineer, I have a ton of great ones in my network.

2) It says you're not available in my geographic area. Where are you recruiting from?


Great. On 2, we are currently live in US and India so are available in that geography. We couldn't do much about recommendations coming from other geographies (as we don't have many partner companies there) so have put that limit. On 1, yeah, we do make exceptions there. we have not yet decided if we want to keep it a hard limit but we are sure that we would want recommendations from people that would have definitely worked directly with software engineers.


Ah, my fault on point 2, I'm on vacay in Uganda and didn't turn on my VPN but usually in the U.S. Sounds good, very cool idea and I'll be signing up shortly


Awesome! Looking forward to have you as a user :)


Usually when someone asks me for a recommendation, I ask them what kind of skills they are looking for and their preferred contact info, and then if I have a friend who might be interested I send my friend the contact info of the person hiring.

So maybe you might want to consider flipping your process. Make it so I can see what skills the companies are looking for and then make it easy to send a message to my friend with their contact info, in a way where you can trace the recommendation back to me.


Interesting. So instead of recommending your friends, you would be recommending best-suited companies to your friends? I have seen a few of my friends do that when they are helping someone with a job switch.


Exactly. Then I don't have to give up my friend's contact info without permission.


Thats good point.. what was the name of that company that shared the candidate assessment data with companies without their permission?


I do exactly this. I get pings from time to time for a position that is a good fit for one of my friends, but does not interest me personally.

I'm far enough in my career that I can be picky and I don't mind the job search taking 3-6 months of mostly passive search.


I like this idea. But does it still just funnel people into the "1hr Leetcode interview + day long whiteboard session" style hiring process?

This is the problem I have with things like Triplebyte et. al. You make me jump through all of these hoops that are supposed to "prequalify" me for a position, and yet I'm then just dumped into the exact same process as if I'd submitted my resume to a job posting.


If this doesn't let you skip the whiteboard interview, it sounds like just one more step in the terrible process of interviewing for engineering jobs.


Thanks for that feedback. Yeah, that's one of the major aspects we are focusing on (we are not 100% there). Especially since we work with early-age startups and founders directly in most cases we theoretically can do that. It might take sometime to build that trust though


Is there no danger of collusion among reviewers?

Can negative reviews come back to haunt the reviewer? Or are users only asked for positive reviews (i.e. silence being the most negative review possible)?


Yes, there is that danger but we have not seen that happening as of yet.

Users are only asked for positive reviews


If you are successful, then next I suggest you will be following in the steps of vendors offering online reference checks, e.g xref, checkster.

You will need all sorts of fraud detection, e.g multiple reviewers all coming from the same ip, and your customers will pay for that security.


What steps do you take to prevent bias or systemic discrimination (intentional or not) from impacting recommendations?


Recommendations and referrals strike me as one of the ways that groups with an unintentional diversity problem end up persisting it. The composition of the group gets carried forward as people vouch for each other.

The summary mentions "scroll[ing] their LinkedIn connections", but relying too heavily on this sort of thing puts people with no connections at a disadvantage, and actually makes it more difficult for a cohort trying to address a diversity problem to do so.


I am not sure If I get it right but hiring based on the social media recommendation model (likes/upvotes/recommend) will always amplify the same group of people who are well connected and have good networking skills. I am interested to know how newcomers will be able to get in.


I would love to try out your platform for my current hiring needs. Good luck with the launch!

One UI feedback for your website - On your contact form for companies, text isn't visible when the form field is not in focus. The text colour matches too closely with the field background.

Edit - Also, on submitting the form, the success/ confirmation text did not appear in the visible window. I had to scroll up to see it


Thanks for the feedback. We will fix it right away :)


The business model of Weekday will have issues with opt-in and compliance on markets like the EU where their approach will breach GDPR by storing personal data for people who have not opted into this.

I have received an email from Weekday to my personal email address that I do not share anywhere in public.

When I asked the team where they have bought/obtained my data, I was told:

“We do not buy data at all. We get information about our candidates from individuals using our chrome extension and recommending candidates to us via LinkedIn. Therefore, your information has come through a LinkedIn recommendation.”

While the approach is creative growth hacking, storing my data (name, email and who knows what else) by using a Chrome extension to harvest data I thought would be private does not seem ethical. It’s not something I am happy about either, and it’s definitely something that is not GDPR compliant.


Which geographical regions is this open to? I am in Canada, but got "It's great to see you here but unfortunately we are just not in your geographical region yet."


Can you check now? It should be available. We do have 1 canadian company who is hiring for engineers in both US and Canada. We had not opened it till now but have just done that.


Nope, it did not work. Will try again at home through.


US and India (I just asked a similar q)


I wish we lived in a world where no one has to vouch; trustless. I know - not possible right? But in my opinion it's an instant fix to racial, sex, any other type of discrimination isn't it. It would close the gender pay gap. It would close systemic racism. So tell me why is it so impossible? Well I know some challenges. Eg a standardized test is impossible. I'd love to start a company that offers blond interviews somehow. Virtual Avatars?


I'd subscribe (as a company) if you didn't have the 10k yearly fees. I view recruiting as point in time events (we are a startup) rather than a continuous process.


We also have a different plan where it's completely success-based. You pay only if you hire any of our vouched engineers. We charge 15% of annual base salary in that case though.


Thanks!


How do you seed & grow the network? How do you identify and vet the engineers on whose recommendations you rely?


Great question. It is something which we ourselves asked quite a bit. Currently we don't vet the person who recommends. We leave that decision on to the companies we work with. We have found that credentials of the person who recommends ends up being important to the companies. We also score the recommender if any of their recommendations ends up getting interviewed. Wrt seeding, it's currently with a combination of personal networks and launches within niche communities. Regarding growing, we have seen that a bunch of people who are recommended end up giving recommendations as well


I.e. Good ole boy's club.


Interesting take. But when looking for a job sometimes your 'network' is a good place too look.

I personally have worked with many people. Some are on my 'just hire them, dont worry about it list'. There are others on my 'do not hire them ever they will destroy any group they end up in' list. There are also 'yeah they will do ok, but meh' list. I personally put myself in the 3rd list.

I also am no longer interested in sharing that info for free to companies that are trying to break into the same space as linkedin.

Hiring is broken big time. But this does not look like it fixes much. The real issue in many companies is a lack of understanding from HR what the manager needs. Even sometimes the manager does not even really know. Or HR has an ideal they want and think there are other qualities that 'fit the form'. From someone who is hiring how would this system cut out the cruft of 'ok I have a random resume sent in how do I know it is any good?' No amount of internet points helps me say 'yeah that is a good one'.


Yeah. That does happen to be an inherent problem with a recommendation/referral based system. We do see that people recommend engineers very similar to them in profile/demographic. We try to solve it by seeding with a diverse base of scouts.


If I recommend connections on LinkedIn will my name be mentioned in outreach?


You can choose to remove your name from the outreach if you want to


Anyone else unable to get the extension to load? I can install it, when I click on in on LinkedIn, I just see a white drawer. No content loads. Tried disabling ad block as well.


Sorry for the bad experience, the extension needs 3rd party cookies to be enabled in chrome.. You can go to chrome's settings and select either "allow all cookies" or "block 3rd party cookies in incognito" and it should start working normally


Regardless of how wonderful this product may, or may not, be:

"Install chrome extension to APPLY"

No thanks.


Sorry about that. I understand that it's a big friction point. We would be piloting a web app as well soon to solve for that :)


cronyism institutionalized?


You should really add a FAQ section on the website. Quite a few people on in the comments here are asking the same questions.


Yes, sure realised that. Thank you for the suggestion, will add one soon.


Nepotism as a service!


Love this idea. Congrats on launching!


Thank you so much Leonard for the support and wishes :)


I'd given this a shot. Nice way to make some money, ha. (Got about a $1500 in Indian rupees which was neat.)


Best wishes the Memer team!


Thank you so much! Yeah, Memer was our previous startup where we figured out the extent of difficulty of tech hiring.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: