Cutting to the chase, they seem to depend heavily on a month long trial period, where the work is performed as a contract resource versus an employee, at a $25/hour rate.
Yup. And this is "actual work we need done" - and get done at well below market rates - of which they only hire 37% of those.
Now, certainly, people will do the trial who were able to comfortably "talk the talk" but not "walk the walk", and there's some attrition there. But for the others? Cheap labor!
While trial projects are actually useful things that need doing, I'd hazard a guess based on my experience that the majority of trial projects never actually launch. Some do. Some get passed off to the person after hire to further polish, and others go to another team to review and tweak and evaluate shipping.
I can say that the trial I did nearly three years ago took me 3 weeks to do, and I put 6h30m in the first week, 8h in the second week, and 6h45m in the third week, for a total of 21h15m.
Also, just to be clear, everyone gets the standard $25/h rate, it doesn't matter if you're a developer from San Francisco or customer support from the Czech Republic. It's something that's easier to standardize, is typically for a short period of time, and has zero correlation to final pay -- which is normally above market rates for wherever you happen to live.
I took a ~4% pay cut when I came aboard at Automattic, which was more than offset by company paid benefits (Health, Dental, Vision, Life Insurance all company paid at 100%, I just took 12 weeks paid paternity leave, 401k matching, work trips and conferences are 100% expense-able, etc), and got more than that raised back on my first year's annual salary review.
Source: I've worked for Automattic since May 2013. It's a great place to be.
At the point when I applied to Automattic, I was working for Speck (phone and laptop case manufacturer) as a Senior eCommerce Developer specializing in Magento, and had been an active contributor to the WordPress open source project for probably about two years?
That being said, please don't think you need to know PHP to work at Automattic. I mean, it's useful, sure, but JS knowledge is huge as well, and we have some folks on our Data team that just do Python. Our Mobile team works pretty much purely in the languages for Android and iOS. Elasticsearch is something we also do huge amounts of work with (as @gibrown can testify)
George was, but not everyone was. I had done very little PHP before Automattic, though I've certainly learned a lot of PHP since joining. We're also hiring for a lot of other development positions:
Their interview process seems to take multiple passes to filter for people who are willing to get lowballed. In a previous thread, Matt claimed not disclosing one's salary history demonstrated "we probably don't have the level of trust needed to work well together."
I love how he ignores the top response to his question: "If knowledge of a candidate's previous salary doesn't influence your offer, why do you ask for it and what do you do with it?"
It should be noted that per the article, this is for a single project for which the candidate has up to a month to complete. It's not a whole truth to describe the $25/hr as compensation for a "full month's work".
This is also, per the article again, undertaken while one has a job.
Not playing sides, just adding context -- I have no affiliation with the company.
I can see why companies want to do this sort of thing, since it gives them a decent look at how people really do their job, but it almost entirely rules out people with a life outside of work. I'm already working a more than full time job, and Saturdays are entirely spent looking after our son so that my wife can get some work done.
That leaves one full day off, and a few hours each evening, by which I've likely been up since 5am, done a full day's work, got our son off to bed, and cooked a meal. Even if I thought I was likely to produce anything worth submitting, there's no way I want to give up the few hours of time my wife and I get together as a couple for the sake of a glorified job interview.
I guess I could take some time off work and treat it as my day job for a week, but then I've given up a decent chunk of my leave allowance in return for a job interview, and the addition of some tax paperwork to my life because I've been doing freelance work.
Yea, I recently turned down an interview with a company that included a similar 'trial' period as part of their hiring process.
From now on, any long, drawn-out interview process gets a 'pass' from me. I've had my time wasted too many times (employer suddenly goes dark after the final interview, no communication, no responses to any attempt at contact, etc). In fact, I've given up on my most recent job search after experiencing two of those in a row (need to save my vacation days for an actual vacation at some point).
On second thought, I suppose if I were unemployed and desperate I might be willing to put up with an onerous hiring process.
Or, conversely, if you had very little commitments outside of work, time to regularly hack on side and sufficient interest in the job to say "ok, this month, my side project is going to be for someone else". But the point remains that this puts an uncommon expectation on applicants, one that maybe Automattic can afford to select for, but that would potentially cause untold problems if all employers hired this way.
For at least a few people, interviewing and job hunting will probably eat into their life outside of work regardless of if the company is paying them to complete a contract project or using traditional days full of interviews on site.
That said, maybe it's supposed to select against people who are ultimately unlikely to block out the time to go through the screening and hiring process regardless of its form. It's some tens of hours either way.
I've got no problem with interviewing and job hunting taking up time outside of work, I expect that to be the case. However companies need to be respectful of that time, and see the difference between a day or two off to interview, and a month's worth of free time.
I respect that they at least pay for that time, which softens the blow somewhat, but at least for me the extra money isn't anywhere near as valuable as some free time. I know what burn out feels like, and there's no way I'd risk it for the chance of a job.
One assumes that it's not a month worth of free time, but a few work-days worth of time distributed over the span of a month. Basically, it's one thing if that trial period takes you 15-20 hrs of paid 'extra' work during the trial month, and a very different one if we are talking, say, 120 hrs of "after work" time. Now, many people will pass on the 15-20 hrs of 'contractor' work as well, but some people will consider it a reasonable deal.
The only point at which this becomes a problem is if the industry expectation shifts so that you can only find a good job by performing contract work for a month before knowing if you'll be hired or not.
As long as there are options, well... the people who prefer doing algorithmic puzzles will interview at the puzzle-interview companies, the people who prefer doing side-projects will interview at the contractor-then-employee companies, and so on and so forth...
The example project is a redesign for a single web page. With a month to let ideas percolate, that's likely to be less than 15 hours of work for a person who has made it through portfolio and resume review for a position designing web pages on the employer side and who is still interested in the job after several rounds of interviews on the employee side. To me that seems more pleasant than a couple of days in the airport and hotels to travel cattle class.
My take is that if it selects against candidates who view remote employment as just another free lance job, more the better for both sides. Remote employment is not another form of freelance work. The relationship is different.
> My take is that if it selects against candidates who view remote employment as just another free lance job, more the better for both sides. Remote employment is not another form of freelance work. The relationship is different.
I'm not really sure how you've made that leap. It doesn't seem to me that hiring people for a freelance job selects for people who don't think of remote work as a freelance job. If anything you're selecting for people able to make that commitment.
The company's goal is not contracting with free lancers. It is hiring remote employees. I suspect the natural inclination of many people would be to treat the sample work project as free lance work because that's the short term form. The long term form is not free lance work. It is remote employment.
If the company is guarding against false positives in hiring, then what they are looking for is people who are able to reframe the short term form in terms of the long term form. That is people who treat the trial as trial employment rather than as strictly a free lance proposition because people who have made that transition are a safer bet than people who at the end of a month are still thinking like free lancers.
Putting on a suit and tie for two days of travel and job interviews is just as much a mock business situation as a trial project. It's just likely to be less informative to both parties. And it's still selecting for commitment to the hiring process.
I've had multiple companies expect me to take 8+ hours of PTO for a job interview. I ended up taking an entire week to fit in all the interviews and I could easily see myself having filled in the down time in that week with ~20 hours of work while I sat on planes/airports and such.
I'm not seeing this as much different and I get paid twice for the twice. I doubt they want more than 40 hours of work out of you in one month while you work full time on a separate job.
Not saying your conclusion is necessarily accurate (though I don't disagree), but anecdotally my own experience was similar. I interviewed for a design position with Automattic 3 or 4 years ago and made it to the contract stage, and ultimately had to decline because the $25 contract gig was just so off-putting. Not to mention the super-intimidating terms and intellectual property agreement that came with the trial gig.
$25/hour as a straight invoice (subject to self-employment tax), is roughly equivalent to a $40k/year salary at a company that gives -0- benefits, at all.
I have no idea if this has any correlation to the salary they ultimately offer.
You're right though...it would seem to narrow the field down to those willing to severely undervalue their work for at least 30 days.
Yeah - say you're working already and making, say $90K as a developer already. You're going to be taxed 25-28% of this $25/hr at the federal level alone.
Money is money, but I could think of more things I could be doing than work for a highly profitable company for what might about to $15-20/hr take home pay.
"Yeah - say you're working already and making, say $90K as a developer already. You're going to be taxed 25-28% of this $25/hr at the federal level alone."
As a contrary opinion, that its really not so bad, I'm old enough to remember job interviews where they use the interview as an excuse to take the whole team out to their favorite expensive restaurant for lunch, and I'm stressed and frankly not going to enjoy it because its an interview full of people I don't know well, whereas a day or two's pay at $25/hr works out very well to "date night with my wife at our favorite steakhouse". For people who don't like steakhouses, or wives, they can find a similar enough way to entertain themselves in compensation for time spent.
It feels like a relatively tasteful way to say "thanks for your time, now go out and have a nice evening to celebrate its all over".
I'm not sure how much time it takes, just that they said it's a month long contract.
But let's assume it's 30 hours total. I often turn down contract work because I'm booked full, so I could have worked those 30 hours at $100 per hour. This "interview" will cost me $2250--in addition to the time spend doing the actual asynchronous interviews.
I'm not willing to spend 2 grand for the chance at a job with this company. I might be willing to lose $500 or so in lost income to spend half a day interviewing.
Do companies really want to limit their potential applicant pools like this?
The reason they give (I applied, but wound up taking a different job before completing that month's trial) is that the $25/hour is a flat rate they give to anyone applying for any position they have, not just the comparatively well-paid development positions. So a CSR, a web developer, a deployment engineer, whoever -- all $25/hour.
They further elaborated that they wanted a flat rate like this because they wanted to avoid what would effectively be salary negotiations before they even knew if they wanted to hire someone.
I can sympathize with that position, though I think it might be more reasonable to do something like paying trials the lowest hourly rate given to anyone in the company doing that job. That'd provide context, while still being a vaguely appropriate payment for the work done.
We debated a 'trial' or take home work as part of our process after reading similar blogs touting their success.
Then we talked to our most talented engineers about how they would receive such a challenge and almost all said they would be put off by the notion of doing either free work or low paid work.
I myself feel like it's a chance for me to see if I even want to work at the company... but for many the mere suggestion can answer that question for them...
And probably encourage prospective employees to violate employment contracts / confidentiality agreements in the process.
Not exactly the best thing to do and/or broadcast. How do they think it is ok to have people moonlight for what is almost certainly a competitor, most likely leaking out a lot of techniques and work that they have been developing for their current employer.
If you work in a related field (e.g. digital agency, web development company, devops), check to see if you signed an employment agreement or confidentiality agreement.
Read it carefully, then imagine if you did a side project relating to what your current expertise is, for Automattic on the weekends.
Then ask yourself if it would violate your agreement.
It might not, or you might not have such an agreement, but we do and I think (although IANAL) this does not sit well with that agreement.
If you do have a conflict, they're normally willing to work with you on how the trial is structured. If I had been under such a non-compete, personally, I'd probably have just done the trial unpaid as a targeted contribution to an open source project or something. Which is very rarely going to be a conflict of interest. :)
tl;dr: they're normally willing to work with you if there's a problem.
Would have been nice if he gave an idea of how long the typical projects were. A 4 hour project at $25/hr is a lot more palatable than a 15 hour project.
It's better than nothing (lots of companies ask applicants to do this home work gratis), but seems curiously low.
The whole process is really geared towards finding people who are dying to work at Automattic. And that's fine if that's who they want to pursue. Truly high caliber people with experience are going to be put off by this process (I actually had to stop and think about whether billing at $25/hour is even worth the effort to invoice it and complicate my taxes, and it is, but not by much), but those people are probably too expensive for Automattic anyway. So it works out.
I don't know about Automattic, but in many larger companies, super-stars-with-nothing-to-prove go through different hiring experiences than most other employees. I can't imagine Google or Dropbox sat Guido Van Rossum in a room and asked him to reverse a linked list in 'whichever language he was most comfortable in'. Although I might be wrong, of course... ;)
For people who have something to prove, well, the only question is which particular strategy the company will use to ask them to prove it. As far as it goes, 20 hrs of paid labor outside of work if your contract allows it, is not the worst I can think of.
I think there's quite a difference between your average no-name senior developer and someone like GVR. Of course they didn't ask him to do coding exercises; the whole world knows his work. But as a member of the population of senior developers who would have his abilities tested in an interview situation, offering me $25/hour for a significant amount of my time is a bit insulting. I understand that they aren't actually attempting to value your time at $25/hr and it's more of a show of good faith than an attempt at complete compensation, but come on, last time I did a short term contract was 9 years ago and I billed $150/hr. I haven't been in the world of contracting lately but my salary has more than doubled since then, so my feeling is that $25/hr is like saying "do fizzbuzz for me. I'll give you a dollar."
On the contrary, Max Howell, who wrote HomeBrew for Macs was asked by Google interviewers to invert a binary tree. To quote Max: "Google: 90% of our engineers use the software you wrote (Homebrew), but you can’t invert a binary tree on a whiteboard so fuck off."
That works for everyone except those who would have to declare to their employer that they're working on the side, for whom they're working and on what. Also, there are places in the US which allow an employer to claim any ideas related to your occupation that occur both on the job and off it (i.e. any code a developer writes, period.)
I'm assuming that type of thing is a no go and leave off a larger chunk of candidates than most people would be willing to admit.
I know there are places where that truly is a requirement to employment but I've twice successfully negotiated (once at a startup, once at an ad agency) those clauses out of my contract and I highly suggest others at least try to.
It resulted in a few awkward conversations with HR / legal but I stood my ground and insisted that them owning my entire life was unfair and said I was happy to have them own my output during work hours but nothing more. If they don't come around I'd consider it a red flag.
>I know there are places where that truly is a requirement to employment but I've twice successfully negotiated (once at a startup, once at an ad agency) those clauses out of my contract and I highly suggest others at least try to.
I don't do this any more. I realized after a couple of these jobs that this level of explicit disrespect written into a contract pretty much tells you all you need to know about working there before you've even joined.
Now I just pass politely and move on.
The pay at places that would try and ram that in there was never that great anyway.
> I don't do this any more. I realized after a couple of these jobs that this level of explicit disrespect written into a contract pretty much tells you all you need to know about working there before you've even joined.
Tbh, we had this come up when we purchased and I negotiated it out of my "new" hiring contract successfully.
So far everyone I have on LinkedIn at #ParentCompany has quit but my little isolated bubble trundles on largely immune to such problems as long as we hit our sales/profit numbers.
So I suspect you are correct in some cases but you can still sit in protected bubbles that have that in the contract AND are still worth working for.
Hey, wheaties, I work on engineering hiring at Automattic. For the purposes of the trial project we’re very flexible and always account for the applicant’s situation. It’s fine if their employer must retain copyright, we can always choose an open-source project for the trial :-)
I don't understand this, sorry. If the employer retains copyright, then isn't it the case that, in most situations, the applicant cannot legally license their work under an open-source license? (Because the employer owns the copyright, and may not agree to the licensing.) I don't see how the fact that it's open source makes any difference.
Most jobs already have a severely restricted hiring pool because they are location-bound. When literally the entire world is your hiring pool, I'm sure they can still find the talent they need, even if there are large chunks of people who have a problem with their process.
Sure, on an individual level for someone who has this problem, it sucks. At a personal level when people at Automattic talk about individual candidates, it probably also sucks. But at a corporate level, evaluating whether their process is meeting their business needs? This issue, though legit, doesn't smell like a deal-killer.
Sometimes yes (without prior approval). In practice this depends on the size of contribution, purpose, and relation to your current employer's industry. I've found there are two types of people: those who do and ask later or claim ignorance, and those that ask up front.
I'd guess if you are employed somewhere with these draconian rules concerning unrelated side projects (i.e. pre-approval) then you might not be the type of person they want. I know that feels a bit like employee/victim-blaming, but it also shows how important that employee values his personal software freedoms.
Any company that claims it can't find talent is artificially limiting their hiring pool, perhaps based on geographic constraints or the like. We're up to 422 Automatticians at the moment, and that's not really slowing down.
In that time I’ve reviewed a total of 251 resumes.
63 of those have gone on to an interview (25% of applications received)
That statistic stuck out to me and it looks really high. When I've gone through resumes for technical positions, less than 5% looked like something we'd want to set up an interview for. I'm guessing places like Google and Facebook interview less than 1% of their stacks of resumes. Is anyone else in that 25% acceptance range for interviewing programming positions?
This is probably because the resumes that the writer receives have already been pre-screened by the CEO.
I tend to think that this pre-screening likely means that they are underinclusive—if anything—in moving candidates through the process. That is probably OK for a company like Automattic, which gets tons of resumes. But for a startup that has less name recognition, it could be more of a problem. Very interest read, overall!
Note: This is after a pre-screen from Matt Mullenweg (Automattic's founder & CEO). Matt only passes on the applications that look good to him. Hence the high ratio.
The asynchronous text interview seems less expensive in both time and money for Automattic. Perhaps that lets them get away with a higher interview rate?
and still Automattic is mostly a one-trick pony, it's 100% WordPress all the way down. Even when Matt had a chance to have a second program with bbPress and fix many of the nightmares with wordpress, he doubled-back and had them rewrite it into an insanely bloated plugin for wordpress instead.
Eh, bbPress used to be on BackPress, which was a relatively unmaintained fork of the functional core of WordPress -- rather than having it run on unmaintained software or duplicating effort by pulling code across, it was changed to a plugin -- which is now much easier to set up and share users and content with your main site. It was a good change.
Also, while WordPress.com is WordPress some of the way down, some of our other stuff is anything but. We've been migrating to much more API-driven functionality, and the current WordPress.com dashboard (codenamed Calypso) is purely Javascript, based in React, and fully open source. You can download and run it yourself if you like -- https://github.com/Automattic/wp-calypso
And yes, as noted, Simperium/Simplenote, Cloudup, PollDaddy, Gravatar, others -- these aren't WordPress.
Right, when you add functionality, speed may go down. Okay. It's a worthwhile tradeoff, largely because very few websites are 'just a forum' anymore, and forums aren't distinct parts of a website, they're fully integrated into the design, and by having it be a plugin it can be more integrated throughout the full site, sharing data and widgets and such back and forth.
Hey CK! Long time, no see! And thanks for all of your contributions to the original bbPress ecosystem.
There's always going to be a special place in my heart for bbPress 1.x, and BackPress. These two projects were the most optimized, as far as all of the sister-projects go, and still to this day power large swaths of WordPress.org.
I don't think it's fair to place all of the blame on Matt, though I understand how and why that's easy to do, especially when your opinion has not been the popular vote either time.
The choice to move bbPress to BackPress was, I think, an obvious one in 2007/8. BackPress was the future, and until WordPress became hugely successful and backwards compatibility became paramount, WordPress was on course to get the same heart transplant. It wasn't until it became clearly obvious that breakage would occur and most everyone agreed that the BackPress dependency exponentially increased the barrier to entry that the shift away from BackPress happened, and bbPress lost its legs again.
Consider, though, that bbPress is WordPress's moon in this scenario. It took the BackPress bullet so that WordPress didn't have to, and for that I think bbPress (and Sam Bauers' work at the time) is hugely responsible for the current success of the WordPress platform, even if it isn't ever championed as such. bbPress showed everyone first hand how much work it was just to get back to square one again. Why bother doing the same to WordPress?
In 2010, it was very clear that WordPress is the hub of our ecosystem, and everything else is a spoke. The decision to move bbPress to be a WordPress plugin, at that time, is another obvious one. The overlapping systems (I.E. BackPress) are abandonware, and there's no sense in trying to Frankenstein everything together anymore. Rather, the decision was to rebuild bbPress with everything we had learned since 2005, to take advantage of everything new that WordPress got and BackPress did not.
* Is bbPress the plugin bloated?
I don't think any more-so than a stand-alone version would be if it were trying to keep up with WordPress's current functionalities, or other forum software.
* Is bbPress the plugin confusing to users looking for forums?
I don't think so, since the problem it's solving now is specifically "forums for WordPress" rather than "forums so fast you'll freak."
* Could bbPress be something great had BackPress or pluginizing it never occurred?
No, to be quite blunt. If demand were there or if it was really as awesome as we remember it, anyone could have forked old bbPress to resurrect it, but it didn't (and likely will never) happen, because porting WordPress's improvements would take a lifetime.
* What's my personal stance on bbPress?
I loved 0.9. I love 1.2. I love 2.x. They're all great, and fun to work on, and rewarding, and incredibly satisfying to tinker with. I will always enjoy the thought of a BackPress powered internet vs. a WordPress one, but that's not where we are today. 10 years from now, when a RESTful WordPress has sites talking back & forth to each other, it may be possible to do the BackPress heart transplant we couldn't do 10 years prior, but that's highly unlikely.
Automattic has a funny way of taking the long way around and having it work out. In many ways, the current incarnation of WordPress.com is very much what BuddyPress was in 2008. Central activity stream, Notifications, Settings, Followers, etc... It's no surprise, then, that a company lead by Matt and projects lead by Matt should operate similarly. The long-long-games for WordPress, bbPress, BuddyPress, and Automattic separately as a company (and Jetpack), are incredibly optimistic and collectively are gunning to be an operating system for the World Wide Web; the connective tissue that keeps people communicating and congregating, and if bbPress needed to take a few hits along the way to make all of that possible, I'm okay with that, because I think it still has a bright future ahead of it, and will continue to be the forum software of choice for people already using WordPress.
I know it's unlikely I'll ever sway you to put as much passion into bbPress 2.x as you did 0.x, but I'd personally love it if you did. You had dozens of plugins and had contributed hugely in the beginning, and whether or not you feel it, I for one consider your early commitment and enjoyment of the bbPress platform to be incredibly inspiring.
> Note: Replying to applicants takes top priority over every other responsibility I have.
Although I don't care for his example rejection letter, I like that they actually respond to applicants whether they interview them or not. It's nice to get feedback that your CV was actually read and reviewed by a human.
Aside from trial payrate, the overall process seems really good. I'd like to see experiments in this method in other companies. Maybe even in large companies at the branch or store level with manager taking place of CEO in pre-approvals.
The more interesting part, though, is questions to gain the most insight. There are forums where experts in marketing, site optimization, etc detail the results of their experiments to show what had real value and what didn't. They become sets of tactics one can throw at new, but similar, problems. If it doesn't exist already, hiring people should create such a site for coming up with assessment questions and tactics. Curated by good moderators and industry veterans if possible. Thoughts? Existing examples?
> At any point in time, Matt knows better than anyone else which roles in the company have the greatest need for additional resources.
I'm skeptical of this statement. It seems to rest on many unreliable assumptions (e.g. at a busy company, it seems that a team lead may know better than the CEO when his team needs additional resources to deliver).
Other than that, I love their approach to interviewing: casual, asynchronous communication.
When people are actively seeking a new role, they may not be the best version of themselves. They may be stressed, feeling frustrated, dealing with a life event, and attempting to conduct a job search on the side. This gives such candidates a better chance to be able to put a better foot forward IMO.
> I'm skeptical of this statement. It seems to rest on many unreliable assumptions (e.g. at a busy company, it seems that a team lead may know better than the CEO when his team needs additional resources to deliver).
As a team lead at Automattic, pretty much all of us have regular chats with Matt -- some on bi-weekly group hangouts, others casual private conversations in Slack -- but we all communicate our personnel needs and concerns to him, and sometimes the solution isn't "Hire a new person for X squad" -- it may be take someone looking for a team switch and shift them around internally.
Having actually experienced this in practice, Matt really is stellar at making sure we have the people we need.
> Once I fire off a contract, I’ll post a comment on our internal hiring P2 to give HR a heads up that a trial contract will be incoming shortly.
> NOTE: P2 is a WordPress theme that makes threaded discussions incredibly simple. We use P2 for a large portion of our internal communication.
> I’ll then create a new private WordPress.com blog and invite the applicant to be an editor. This is where I’ll post the project brief. This is also where I’ll move the conversation for the remainder of the trial.
I understand WordPress is their business, but I feel that using WordPress for internal chat and interview board is kind of weird.
We also use Slack extensively, but P2s are a pretty solid choice for us (we used to have the unofficial slogan "more p2s than people!") -- it makes onboarding really easy, as the full history is easily searchable and you don't need to forward emails to new people or the like. It also defaults communication to open, which is super nice.
I interviewed, made it to the paid project stage and got cut loose. Since then I've contributed to a number of open source projects and built some great things in production. Would now be a good time to apply again?
