Hiring at Scale | Patrick McKenzie, StarFighter | BoS USA 2015

No comments

Patrick McKenzie, StarFighter

Patrick McKenzie (or patio11 as he is better known on the Internets) gave a great, practical and actionable, talk about how to get better at hiring engineers at Business of Software 2015.

“Every hiring process sucks, but think of hiring as a funnel.

  • Outreach — Creating the eco system (100 nerds)
  • Attract — Getting people to come to you (10 nerds)
  • Assess — Do they meet the requirements? (6 nerds)
  • Close — Give them an offer (Your new employee)

Slides, Video, AMA & Transcript below

Slides from Patrick McKenzie’s talk at BoS USA here

Video

 

 

AMA

 

 

Ticket prices rise 3rd May 2017

Register for BoS EUROPE Register for BoS USA

Don't Miss a Thing - Get BoS Updates

Want us to let you know about new talk videos, speaker AMAs, Business of Software Conference updates? Join the smart people who get BoS updates. Unsubscribe anytime. We will never sell your email address.

Transcript

Patrick McKenzie, Starfighter: Howdy to everybody. My name’s Patrick McKenzie, better known as @patio11 on the internet. I’ve spoken at, I think four Business of Software’s so far. I missed last year, but I have a really good excuse for it. We were very close to a shipping date. This is the one start-up I’ve ever been involved in which is sustained 10 per cent week over week growth for years.

Mark was talking yesterday about Unicorns and about how Business of Software was more about values than valuations I was a little moved by that. I’ve always felt like this conference is sort of the weird counter-culture in software that says that you can be, you know, a normal loving parent, loving spouse and still have a wonderful career in the technology industry and I think for a lot of us that is made possible by our better halves in our relationships and by a lot of co — collaboration by our family. Always makes me very happy when I see folks like Greg’s wonderful daughter Emma and wife Rachel who came out here to support him and there’s a lot of folks supporting us back at home so can we just give a quick round of applause for them.

So, I come not before you, not as a loving, attentive father and husband and not as a, a small software business man like I’ve been the last couple of years but as that most loathsome of all creatures – I am now a technology recruiter and I want to teach you how to do your hiring a little better, such that you never have to deal with technology recruiters, because I have for the last couple of months and oh God they are scary people.

This talk is distilled out of something that I learned from my two co-founders at my new gig called Starfighter and we might talk a little bit about Starfighter later. My co-founders ran a company called Matasano, which was an application security consultancy operating out of Chicago, New York and Silicon Valley and they have a very high bar. It’s a professional services business, but their professional services involve parachuting into arbitrary companies, like Bank of America or Apple or Microsoft and doing horrible, horrible things to the best work that their engineers have ever done. So, one week they might be cracking the encryption on a PlayStation 3 on the firmware. The next week, they might be doing pen testing on the web applications that front-end the New York Stock Exchange.

So, they have a very difficult hiring bar.

They need to hire literally the best engineers in the country, who are capable of within two weeks, completely grasping a new company’s tech stack and then doing horrible, horrible things to it. And because they’re a professional services company, like Rick was saying the other day, they’re fundamentally rate limited by how quickly they can hire new engineers and get them staffed up on new projects.

So, at one point in 2010 they had about 20 engineers working for the company and they really needed upwards of 30 to deliver the projects that they had already promised to customers. So, they need to hire very, very fast for very, very good people and this would ordinarily not be that problematic except, unlike the little joke spy gig that we have up here, a real spy gig happened.

These are our friends at the Ministry for State Security over in China. They’re totally real and they’re totally vicious and in 2010, they did horrible, horrible things to Google’s internal security. And Google’s not very happy about this. In fact, when it was explained to Larry Page, we have a photo of what happened at that meeting. He had his own little stack overflow.

And Larry went to his team of very smart Google people, very smart Google PhDs and he said, I don’t care what it takes, I don’t care who the adversary is, I don’t care if it’s China, I don’t care if it’s the NSA, Google never gets hacked again. Put as many PhDs on this project as you need. You have unlimited budget. You have whatever resources you need. If anyone gets in the way of this project, put them on the phone directly with me, I will stomp on their face, but we never get hacked again.

And 2010 was a very, very good year if you were a candidate trying to get hired for a software security job because Google was throwing around a lot of money. This number is not what Google were hiring new security engineer for, that’s how much money Google makes, per minute. And when my co-founder Thomas gave someone an offer in 2010 of $120,000 a year to come work at Matasano, the person took that offer to Google recruiter and the Google recruiter said, $120,000 huh. That’s a decent offer. My opening bid is 360, but if you need more, I can get more. Just ask.

And at the same time Facebook was also a little worried about being the next target of the Ministry of State Security and Facebook started doing the same thing, also with like magic infinite money from the stock market. And the bottom, well, the bottom didn’t fall out of the hiring market, the hiring market went totally on fire, but the pipeline of candidates to Matasano completely dried up. And it was an existential threat to the company and so what happened was they got very good at hiring, hiring very well, very fast. And it ended up growing from 20 engineers to a team of about 50 engineers, in about a year after figuring this out, despite the fact that all the king’s horses and all the king’s men at Google were backed by infinite money to hire the exact same candidates.

We’re going to productise a bit of that, a bit of the info — the know-how that went into that process at Starfighter and I’m going to be a wonderful CEO and instead of getting in front of you guys and pitching Starfighter, I’m going to make it totally unnecessary to ever pay me or any other tech recruiter money again.

Do you know that an average, like recruiting — if you get their recruiter to hire a candidate for you right now, the going rate is about like $25,000? That’s a whole heck of a lot of money to give to recruiters. I think you can find better ways to use it. So in this presentation if you put these, if you put these tactics to use, I want you to get one marginal recruit this year. Save yourself $25,000 and then put it to better uses and I’ll have some examples of things you can do here.

So, I come out of a low touch SASS background and I see funnels in everything – websites, email addresses, recruiting,. It’s all a funnel to me and I like to think about the recruiting funnel in terms of these four steps:

  • There’s outreach, where you are going out to the world and creating ecosystems for yourself of people who know your company’s an option.
  • There’s the attraction step, where you get those people to not just know your company’s an option, but to actually come formally, informally to you and ask for a job.
  • There’s the assessment step, where you figure out whether those candidates meet your hiring bar.
  • You extend a portion of them offers, and then for some portion of those offers you close them, you get them to successfully be actual employees of your company.

And I think people think that hiring is much, much more difficult than it has to be, but when you break down the numbers, and these are just roughly representative numbers, you only need to become friends with a hundred geeks to make this work and you can all become friends with a hundred geeks. In fact, if you really worked at it, you could probably become friends with a hundred geeks like tonight. Just don’t go poaching all of each other’s employees, that might create a little bad blood, but — actually no, I’m going to check myself there, never use the word ‘poach’ with regards to an engineer. I can go on a little rant here. Poaching is a crime that’s committed against the king. You’re stealing his deer that are his as ordained by God and engineers are actual people who have an agency. You can’t poach an engineer. Bad me. That’s — I’ve been talking to too many recruiters in that last couple of months.

So, you find a hundred geeks, you – a hundred of those hundred geeks you identify 10, roughly well-qualified people for your organisation. You convince six of them to go through the formal, informal interviewing process at your company. You extend some portion of the six offers and then one of them becomes your new employee. And that will save you $25,000. Let’s talk about getting in touch with your hundred geeks.

Before we do that there were some general principles.

Just like everything else in a software company, you want to track the heck out of what you’re doing here.

I know one company that has CIA style. Every time someone gets on their recruiting radar, they take a little picture of them and they put it on a board, with like lines connecting it to demonstrate friendships and little hand-written notes beneath it and I saw that and was a little creeped out by it. What I prefer to do is to throw people into a Trello list. So, I’ve got one list of everyone who’s like kind of on my radar but they aren’t actually in the recruiting process yet. And as people move into the process as we’re just like swapping emails about, getting a coffee date on the calendar and as we’re moving from, okay, we’ve had coffee dates, we’ve felt each other out a little bit, to moving then to like a formal interview, you just move them in between Trello lists.

That way you can make sure that you’re not dropping anyone on the ground and you can go back and see from — see your statistics of activity for the last month to say, okay, I know I have a funnel. I know I need to be setting up 10 coffee dates for every person I eventually hire, so am I doing the 10 coffee dates. And speaking of sustained effort, you need sustained effort and buy-in from your founders and your engineering management if you’re going to escape the curse of recruiters doing everything for you.

A rule of thumb at a lot of companies is 20 per cent. You need 20 per cent of the founder’s time devoted to hiring and you can tell if 20 per cent of the founder’s time is devoted to hiring by looking at their calendar. You should see eight hours a week blocked off for coffee dates, for interviews and for prospecting with candidates and if you don’t see eight hours a week blocked off on their calendar, they probably aren’t doing it, no matter how much they think they are. So, make that time for yourself. You need to get out there in the world and turn this crank week in and week out. Because hiring is a funnel, it’s just like a sales pipeline. If you didn’t do any prospecting calls for sales this month, there’ll be some month in the near future where you have no sales at all, right. If you don’t do any preliminary groundwork this month for hiring, there’ll be some month in the near future where you will not be able to hire anyone for love or money. So, do this sustained work now.

You also need to incorporate hiring into the day-to-day life of your company.

It can’t just be this exceptional process that happens once in a while. But, you will go to your product teams, you’ll go to your engineering teams, to your DevOp teams, to your marketing teams, to your sales teams and say hey guess what, hiring isn’t just the founder’s job. It’s your job too, particularly at the early stages when we’re just getting on the radar of good candidates. We’re going to give you the tools to do that better, but you’re going to have to actually do that. That’s part of your job.

And you need to play long ball. You’re not just hiring for a week from now, for two weeks from now. You’re not just hiring for the single open requisition you have right now. All of us are running real software companies. We’re going to be here in two years, we’re going to be here in 10 years. There will be people you talk to for the first time at this conference, who might be great candidates for you two years down the line. So, you’re going to comport your activities on a day-to-day basis, thinking that, okay, this person might not be ready for me right now, but they might be ready for me in two years. They might have friends who are ready for me in six months, so I’m going to treat them with every bit the urgency and the, the diligence with this relationship that I would for someone I was attempting to close in the next 24 hours.

Let’s talk about the outreach step in a little more detail.

The goal of the outreach is simple. You’re trying to expand your ecosystem among talented developers or talented whatever you’re trying to hire for and get more people to realise that you are an option and if I was to think of who at this conference has the easiest time doing outreach, I would probably think of Basecamp. And I know what some of you are thinking, well that’s easy to say for Basecamp. I mean they have a ridiculously successful product. They created a ridiculously successful open source project, which hundreds of thousands of developers use every year. They have a blog, blog that has been well read by hundreds of thousands of people and they’re co-founder is a rock star who sits in front of conferences and then goes racing in Europe for his spare time.

And I think that’s a cop-out, for two reasons. One of them, is that Basecamp accumulated these advantages the same way all of us do all of our work. They got up in the morning and they grinded it out and when you see this wall of advantages that they have, that’s simply the result of 10 plus years of patient execution day in and day out. On day one they had a blog that was read by nobody, just like you will. Their open source software was used by nobody on the first day, just like yours will. Their product was used by nobody just like yours was the first day and they just, continually accumulated marginal advantages. But another thing that I think you shouldn’t be demoralised. You shouldn’t be demoralised by the Basecamp example, because you don’t need a community of a hundred thousand people to, to create hiring at your company. You only need to make friends with a hundred geeks at the margin, so let’s get you closer to a hundred geeks.

Most of you are not using events well enough to, to drive hiring. It’s one of the secret weapons that can be done by any company, no matter how small on a budget of between a hundred dollars and nothing at all and it works very well. I want to give you an example from some of my friends at MakeLeaps. MakeLeaps is a company in Tokyo and they have an impossible hiring profile. Get this. They need someone who is bilingual in English and Japanese and capable of working in a multicultural environment, where the customers are Japanese mega corps with all the expectations of salary but their team are foreigners or young Japanese people who do not have the expectations of salary. They like want to have a life and in addition to that you need to either be a top notch Python hacker or a top notch Sass marketer sales person and these are skills which do not exist in the Tokyo area and should not exist in that combination. And MakeLeaps was having the problems of hiring these people.

And so MakeLeaps, when it was attempting to hire people was having the problem finding these unicorns. How do you find someone that hits all three of those qualifications? And the founders of MakeLeaps were geeks. They spent time on Hacker News, just like I spent entirely too much time on Hacker News and so they thought, hey wait, someone who is reading Hacker News is probably to first approximation a pretty good fit for us if they’re in Tokyo. It’s an English language website, so that, that implies that they’re probably bilingual. They’re tuned into the start-up culture. What if we just threw a Hacker News event? We thought, well what do we need to make that happen? Do we need to ask permission from Y Combinator? Nah, no, we’ll, we’ll just wing it.

So, they went to Hacker News and said, hey, you’re on Hacker News, we’re on Hacker News, come be on Hacker News together this evening in a bar. And they paid the bar, I think, a hundred dollars to cater for the first event and got maybe 15 geeks to show up and they did this one night a month for the next 48 months, believe it or not. After a while they outgrew the first venue, they went to a bar and said, hey, we’re able to pack a bar with 40 geeks on a Thursday night, which is a slow night for you. Can you just put someone out in front of the door, and charge everybody a cover charge and give them two drink tickets and just like that these events were actually totally free to MakeLeaps. All they had to do was announce it for the month and then show up and talk to people.

And what happens at these events is they just show up and talk to people. There’s no formal agenda, they have a little kick-off speech by the CEO. He introduces folks, welcomes new folks to the community and says, hey, what are you here for? Come up and tell everybody. And they’ll say, oh yeah, I’m a Python developer, I’m looking for a new job and when people come up and say things like that the CEO makes a special intention to come up to them later and say, hey, I hear you do Python, tell me a little bit more about what you’ve done recently. Oh really, that’s interesting. Would you like to grab coffee with my team tomorrow?

And this is approximately the size of the MakeLeaps team. Five of these unicorns were found at those Hacker News events for a grand total expenditure of under 300 dollars over 50 events. That’s $125,000 saved in recruiting fees for candidates that no Tokyo recruiter could actually find for you, because the recruiter would tell you sensibly that those candidates do not exist. By the way, one of them is the co-founder of the company now.

So, you send Devs out to all manner of meet-ups and events and they probably don’t do wonderful jobs of incorporating your hiring message into the company — into their like, speeches at those events. Partly that’s cause Dev’s are just, not sales-oriented by nature, we’re a little put off by, oh no, I’m going to pollute the purity of my Ruby talk by like, talking the company’s book for a moment.

And then partly it’s just because they don’t have that in their skills set. So you need to say, look, as a quid pro quo here we pay for your ticket to come, go out to this event, we give you the extra time for it and we’re going to help you be a little bit better – add just one slide to your presentation when you’re talking about hiring. You’re not going to give out the world’s worst call to action which is, by the way we’re hiring. You’re going to give out a better call to action and it looks something like this:

Thanks for listening, I love talking to you more about this kind of stuff. If you’d like to learn more about this topic, you can go here and get free lessons from us, or you can just send Bob at our company a message and he’ll get you started on that. By the way, we’re hiring and the thing I like most about working at this company is, fill in the blank.

That’s ,by the way, a magic thing for you. Every jobs page in the world has the same tone. I think it’s because they’re all written by either HR recruiters and they’re all universally terrible. It’s like, oh yay, dental benefits and a 401k matching plan and you’re going to change the way the world does customary communication, yeah. But, the actual employees who live day in and day out in your company, who decide to invest a significant portion of their working career and who love your company, they’re the best possible investors for it.

I guarantee you that absolutely nothing on Twilio’s job page would, would, compete in the slightest with asking Greg in two sentences, Greg what do you like about working at Twilio? He’s much more passionate, much more real, much more authentic than the jobs page can ever be. So just tell them, hey, be real and authentic. Tell what you like about our company to the, to the audience for a second. And then have an easy call to action which is, hey, since we’re hiring and I know the guy who’s doing the hiring, if you go through me, I can jump you through some of the BS at our — in our hiring process, because every hiring process ever has a little BS in it. And this can be as simple as, yeah, I put you directly in touch with the hiring manager and we skip the whole, you know, send your résumé into DevNull.

So, I want to talk a little bit about Matasano Cryptopals, which was their secret weapon

One of their secret weapons for developing the hiring process that resulted in them hiring 30 people from under the nose of Google.

Let me tell you a little bit about what their hiring process was before Cryptopals. It’s what a lot of us do. Have you guys heard the term network hiring? It’s, hey, write down a list of everyone you’ve ever worked with before, who you liked working with and everyone you went to school with before, who you think is probably a good developer and all of your buddies and all of your cousins. Okay, now write down that list and we’re going to treat this as our candidate pool. And there’s some problems with network hiring.

First, that it over-concentrates your candidate pool among birds of a feather. I mean, shocker, twenty-something white guys who graduated from Stanford tend to know other twenty-something white guys, who graduated from Stanford. And if you do network hiring, starting with a seed set of five white guys from Stanford, you end up with 20 white guys from Stanford and you wonder, why can’t I hire anyone who is not a white guy from Stanford?

Smart people in the valley run into that problem every day. Another problem with network hiring, if your hiring thesis is, I hire a new VP of engineering and the VP of engineering says, no sweat, I’ve got six guys I worked with before, they’re great. I can bring them straight into the company and they’ll start within six weeks. At 24 months from now that VP of engineering will be sitting in another office saying, no sweat, I got six guys I’ve worked with before, they can start within two weeks and suddenly your entire engineering team vanishes. And that sounds like a funny joke until it happens to your company.

So, what you should instead do.

The other thing people do instead of network hiring, is hiring by proxy for aptitude. So you go out to other institutions that are known to have like a high bar of people they allow to be affiliated with the institution, whether that’s Stanford or Google or say it’s the black hat conference on security, you say, wow, if you were good enough to pass their bar you’re probably good enough to pass mine.

Problem with this is that everybody in the world gets the same signal from Stanford and Google and black hat and everyone is trying to hire the same pool. So, you have to create your own pool, a little bit of what VCs would call proprietary deal flow, to find the candidates that the rest of the recruiters, the rest of the hiring managers in the world, aren’t looking at. Cryptopals was how they did it for Matasano. They not only found candidates that the rest of the world wasn’t looking for, they created candidates, that the rest of the world wasn’t looking for.

Here’s the overview of Cryptopals. They picked what was they thought was the most interesting part about their day-to-day work at the company, which was that the company had a side line in cryptography. If you happen to have say, a video game console which had a firmware implementation of a special cryptographic algorithm, to do DRM, you would call up Matasano and Matasano would do horrible things to break that before the hackers could and deprive you of several tens of millions of dollars in your video game’s new release windows.

And they said, this is interesting. People want to learn about cryptography, but you can’t really learn cryptography very well at a university and most companies that actually do work in it, don’t really spread that outside the company. So, they just built basically a university course on cryptography, which is seven problem sets, which each set gated by the last one and it took about two weeks of development time, just writing down problems. It was graded manually. They would give you a set of problems for the week and say, produce computer code which does X, Y and Z regarding cryptography, which we’ll teach you how to do.

You produced the computer code. You sent it them over email and then they had their smart geek who was, who was comfortable working with any language that you came up with, just evaluate whether your code worked over email. And they had people submitting everything from C++ to Java to an Excel spreadsheet which did encoding or something. To, they even had a couple of guys do brain and you know the rest of that language. But they suddenly had a hundred, you know, hundreds of people doing hundreds of solutions to cryptography problems in every language known to man.

And just the grading correspondence with these folks took about 50 per cent of one junior engineer’s time. It was totally managed in Gmail for the first couple of weeks. After that became a little unmanageable. They spent two days making a Ruby on Rails application, which would list like, email address, what level they are and the problem set and let them save their answers so they could review them later.

There were seven problem sets. Problem set eight was, you seem to have learnt a lot about cryptography and you are a good enough programmer to produce programmes, which break cryptography. Would you like to be doing that on a day-to-day basis? If so, we should talk.

Here’s an example of set number one and I apologise, it’s a little hard to read, but set number one starts out very easy. It’s just, okay, using library hash functions, take a string, reduce it to the hash. And this problem set was designed to be blown through by qualified programmers in less than 10/15 minutes. And then the difficulty curve got hard. Here is problem set six, where stuff gets real. One of the things here, I’ll just read it to you. DSA nonce recovery from repeated nonce. That’s an actual real attack which broke the PS3s firmware at the cost of hundreds of millions of dollars.

It’s also an attack which is being applied in real time right now against the Bitcoin block chain. If you ever repeated nonce, which is a bit of random value that you have to use when you’re doing cryptographic initialisation, you will immediately get your Bitcoins stolen.

So, they took these people, who just a few weeks ago knew nothing about cryptography, who were now applying production graded tax, which the smartest engineering teams in the world lose to. And they find the people who have both the interest in doing this and the aptitude for doing this and then they’re immediately able to give them a job. And do you think these folks who are passing this level of cryptography course look like, engineering grad from Stanford, recently did six years in software security at Google? No, they look all over the map. There was literally someone who was a high school chemistry teacher in Chicago, who beat these, beat these challenges in Microsoft Excel.

They had people coming from things like a background in audio engineering where their résumés suggested no coding aptitude whatsoever. They had people coming from things like a multi-year career in IT or in systems administration, where you don’t naturally think, this is probably someone we could parachute into the New York Stock Exchange and have them have a shell operating within 20 minutes. And they gave people the ability to surprise them and they ended up being very surprised. And they hired about 30 people out of this channel over the course of a year. And of those 30 people, in three years not a single person quit the company, not a single person was fired and not a single person was hired off by their good buddy from their last job, as part of a network hiring thing.

You should clone Cryptopals because of the work. You just pick one high status fun bit of what you do on a day-to-day basis, spend just a little bit of time working that into a set of lessons. You don’t need to make a web front-end for it, you don’t need to make a complicated tooling on the back end to administer these things, just teach folks over email.

Have them submit the solutions, every time they solve one level, send them the next one. Make it as easy as possible to get started. For the longest time, it was actual — you just sent them an email and they would send you the first problem set, no questions asked.

They actually didn’t put it on the internet, which is an interesting choice for me and they said the reason they didn’t put it on the internet for the longest time, you can find them at cryptopals.com now, but for the years where they were actually actively hiring through this they didn’t have them up anywhere. They said that the existence of all the solutions online, in a way that someone could see them, demotivated people, because, when you were looking at problem set number three, having completed one and two, the fact that if I complete number three, I get to unlock the secret of four and see what they’re teaching me next, was very motivational to many of their candidates. And when they said, okay, we’re going to publish all eight problem sets on our website, people told them, no, no, don’t do that, I’m still working on problem set four and I want to earn five. I don’t want five to be spoiled for me because I can connect to it with any web browser. I think that’s an interesting realisation.

So, some folks are going to surprise you in this process. They’re going to come up with off-the-wall solutions. They’re going to over-engineer things, they’re going to solve problems faster than you thought possible, in ways that you weren’t even expecting and those folks, you feed into your hiring funnel at everything. You don’t need to say, quid pro quo if you get through level seven, we give you a job offer. It doesn’t have to be anything like that. Just use that as a signal. Any time that someone who’s creating the solution says, wow, this person is worth talking to. Then you fire that over to the hiring manager. You say this person is worth talking to for X reason and the hiring manager jumps on that.

And you don’t use these as assessment, which is an easy pitfall to fall into. This isn’t a test. It totally ruins your relationship with your audience if you make this into a test. They’ll feel like they’re being graded. They’ll have that problem where people get up to the whiteboard and despite 10 years of coding, suddenly are unable to write a four loop.

You want people to do their best on these and the only signal that you extract out of this is, we were willing to put them through the more formal part of our evaluation process. Or, you know, they’re someone we’ll keep on our radar for the next couple of years and we’ll, you know, connect with them on LinkedIn, connect with them on Twitter, see what they’re up to, but not someone who we’re going to go after this week.

CTF’s

So, CTFs are capture-the-flag events and a CTF is basically Cryptopals, but taken to the next level, where you instrument something such that can be played online by programming. And you have interesting engineers come to you, play with it and use that to both get your name out there in front of a lot of engineers and also gather signal about who’s a very good engineer and bringing them into contact with your hiring process.

Who here’s ever played a CTF like say by Stripe, two years ago? Okay, a few less than I expected. The Stripe CTF was a whole heck of a lot of fun. I’m going to recommend that you guys don’t implement CTFs for your company. And I’m not saying that selfishly because Starfighter is going to implement CTFs as their only line of business and I don’t want the competition. I’m just saying it because doing CTFs is murderously, murderously difficult to do well. Stripe is a very, very competent engineering team and I won’t tell you non-public stuff about what it took to get their CTF into production, but let’s just say, hundreds of thousands of dollars of engineering costs. And you talk about hundreds of thousands of dollars of engineering costs, like that almost makes hiring a recruiter sound cheap.

Also, keeping these up and running is just a beast. It’s going to be a full time job for people at our company. Stripe was only able to keep theirs up in production for on the order of for a week, a week and a half before they had to take it down because it was distracting people too much. If you want to hear about CTFs in a little more detail, I’ll talk about Starfighter later, but onto stuff that you should actually be doing.

Friendcatcher

So, all of you have products which are routinely exposed to lots of people. You should use the product as a Friendcatcher and there is a lot of ways to do this. They range from ways that are absolutely trivial to implement, but actually get surprising leverage for you. One way is if you just talk to your DevOps team or whoever’s in charge of like, server maintenance, ask them to put in a special non-standard header into your HTTP requests, like X-we’re-hiring and say if you’re seeing this header, we should probably talk. See here.

Every company I know that has done that has gotten multiple people to apply through that pathway. You can even gamify it for the engineers. Say, you know, curl your résumé into this endpoint and our API which isn’t exposed elsewhere. As you know, you can skip directly over the FizzBuzz test because if they can read an HTTP header and then figure out the curl command, yes, they are probably a programmer. But, you often tend to attract like, folks who are a little more interested in the technology on a day-to-day basis rather than just looking for whatever their next gig is which for a lot of companies is useful hiring signal.

Other ways you can use your product is Friendcatcher.

Who here sells to developers in some capacity? So lots of the folks. Here is a great offer I received from Keen IO the other day. They said, hey, we notice you’re getting started on Keen IO recently. It’s an API which shows analytic stuff and he said, we notice you’re not quite up and running in production with Keen yet. Would you like to pair programme with one of us? We’ll help you implement it in your, in your product and get it up and running in production.

You should steal this regardless, because it’s just a ridiculously effective sales tool if you’re selling to developers, but even if you’re not like, just bracket the sales part of it for a, for a second. You’re making time out of someone else’s work day to have one of your developer’s pair programme with them. You get to see exactly what their behaviour is in a pair programming scenario. How they typically work, what their code base is like and you get a chance to say, yeah, we’re doing really, really interesting things here. What do you guys do every day? Oh, you’re a middle American insurance company with a big freaking Java enterprise app. Well that’s cool too, but if you want to work on path breaking mobile analytics every day, I could get that ball rolling. And you know, it feels like a developer/developer conversation, because it is and you’ve given them a lot of value, because you have. And it’s much less, you know, they’re much less likely to reject it out of hand than they would be if your recruiter called them up and said, hello developer, I’ve seen your profile on GitHub and I see that you do Java and I have a front-end development position for you. Click.

Many of you do OSS development or you do API development, which is wonderful because developers are coming to you every day and saying, hey, I use your stuff and I was willing to take, either time out of my work day, or time out of my own free time to work with your things. If you’re doing that like track people like the CIA thing where you put photos on the board, every time someone does a commit to your OSS repositories, whether it’s like a documentation change or a new feature, their photo goes on the board. Cause that is someone who is saying that A, they’re interested in what you’re doing and B, they have the skills to help you. You should keep in touch with these folks.

You should reach out to anyone who does, who does any level of collaboration with your stuff, whether it’s with your API, or with your OSS projects and say, hey, we really appreciate you working with us. This is both from a sales perspective and again from a hiring perspective. It’s like, congratulations on getting your first features submitted. We’re going to merge that into core and everyone will see it. We want to say thank you, can I get your address and send you like a T-shirt or a book that’s relevant to our interests.

That allows you to deepen the relationship and get that ball rolling with them. You don’t have to immediately say, wow, you committed to our repository, great, do you want a job. But you’re playing the long ball, you’re developing a relationship with them, such that, they know who you are and when that conversation happens, three weeks from now, six months from now, two years from now, they will be favourably disposed to your company. Or when a totally different conversation happens, when one of their buddies says, yeah, my start-up went under. We’ve got two weeks of runway left, I need to find a new job. They’ll say, hey, you know who I really like. I was just working with these guys back at, your company, the other day. They seem like good folks and I know they’re hiring for API Devs. You do API, why don’t I put you in touch with them?

So, let’s talk about attracting those inbound candidates to your company.

Increasing the number of qualified candidates, who actually apply to do a job for you, whether that’s a formal application process. Although in my advanced years, I’m getting a little down on formal application processes. I like just starting a conversation, because conversations are easy to get into, easy to get out of and they feel like a whole lot less work and a lot less suck than sending in a résumé and doing a “formal application process”.

One of the highest things that they ever did at Matasano was sending out books. So, anytime they became aware of a developer, whether it was meeting someone at a meet-up, whether it was meeting someone in their day-to-day affairs, they had someone apply for a job and that person wasn’t quite at the skill level they were looking for. They said, hey, you’re interested in software security, that’s great. The best book that has ever been written about practical software security is The Tangled Web, by Michal Zalewski, who by the way had no connection with the firm whatsoever. His book is available on Amazon for 25 bucks.

And they would, they would say, it’s great that you’re interested in software security, if you want to learn a little bit more a bit about it, give me your address, I’ll send you a copy. And people would be like, what, what do you want out of it? Nothing, we don’t want anything about that, we just want to send you, send you a free book. If you want to read it and apply the lessons and come apply, you know, come back and talk to us in a few weeks when you know more about software security, that’s awesome. If you don’t, enjoy the free book and they sent out a thousand of these books.

And you think okay, 25 dollars a book, shipping and handling, that’s like five bucks in the US. That’s three thousand bucks, that’s like a tenth of what you pay a recruiter for a single — man did I do my math wrong? Yes, I did. Okay, they sent out a hundred of these books and I’m a CEO, I have a reality distortion field around me. Say what I need to, to get the deal done.

So, anyhow, whatever, they sent out some number of these books, but sending out a large number of these books is not nearly as expensive as paying a recruiter 25,000 dollars just to get one more candidate in the door and it creates knock-on value for the company in a lot of ways. You’re not just getting the one candidate who applies to you out of that, but you’re making a lot of friends. A lot of people, who, when they’re bouncing around the ecosystem for the rest of their days and someone asks them, hey, have you ever heard about Matasano. They say, yeah, Matasano, great guys, they sent me this book totally out of the blue. I think they’re really interested in software security. Who I do know there? Thomas, you should really talk to Thomas. You want people who are talking about you when you’re not in the room like that. This is a great, easy way to bribe them. By the way, books have a higher perceived value for developers than a lot of things including like company swag.

I do love getting company swag from companies. In fact, I have more start-up T-shirts than I know what to do with

And as you guys are probably aware, I wore the Twilio jacket substantially every day for a good five years. But a book says, hey, you know, it’s why books are great Christmas gifts. Hey, I’m being thoughtful. I know something you’re going to be interested in and it’s not clearly self-interested like sending you a T-shirt is. You’re not saying, hey, carry the light of my company into all your conversations with your friends, you’re just like lightly implying that.

By the way, can I tell you something about how Twilio does jackets, which I don’t think is super public knowledge, but it’s really useful? So, Twilio has like a, two-tier system for swag. Like you can get a T-shirt from them for any reason and lots of people have them, but you can only get jackets in one of two ways. Either, you get hired at Twilio and you’re given a jacket the first time you ship an application on top of the Twilio API internally. And they call you in front of the entire company in an all-hands meeting and the CEO drapes the jacket on your shoulders, knighting you, anointing you as one of the Twilio’s chosen. And that really speaks to both the Twilio employees. It’s a great thing for internally, but it also like, preserves the value of the jacket.

The way you get a jacket if you’re not inside Twilio and one of the reasons I wore it every day for five years, strutting my jacket stuff to all the people who knew what a Twilio jacket is worth, was, you have to do conspicuous above and beyond, service to the company. So I actually found a security vulnerability at one point that would allow you to totally take over the Twilio platform. And rather than blabbing about that on the Twitter site, got in touch with Twilio and they fixed it within three minutes. So, I got a jacket from that and I got four jackets like that and that was why I was wearing my Twilio jackets with pride. So, if you ever use swag to like motivate people or to maybe motivate people in your hiring funnel, it’s useful to say, this is not just a jacket, there’s a story behind this jacket. So think of how you can tell interesting stories like that to folks.

By the way, if you tell people this jacket comes with a story, then one of their friends says, that’s a cool jacket, it’s nice, it’s well made, it’s available from American Eagle for $35, they’ll say, no, no, this isn’t just a jacket, there’s a story how I earned this jacket. I’ve told like the Twilio, you know, I IO’d up their application and then told them about it and they responded very, very well and they gave me a jacket story to probably hundreds of people over the years and each time that was someone being exposed to Twilio. It’s like, hey, there’s another company I could work for.

So, I want you guys to systematically reduce the cost of interacting with you because applying for a job is a huge commitment. It’s a huge commitment both because, most hiring processes suck and your prior expectation when going into a new hiring process is, oh man, I’ve got to work up on my résumé and get that submitted in. And then I won’t know where I am for a couple of weeks. And then they’re going to put me through a gauntlet of interviews and everything is going to suck. Ugh. And because for most employees in engineering have jobs right now and so applying for a new job is a major emotional commitment for them.

If you’re thinking okay, there’s this company that’s sheltered me, that’s paid for my family’s upkeep for the last couple of years and I am seriously thinking of leaving them. Oh dear. Am I ready to leave them? I feel pretty good here. Granted my boss isn’t exactly the best and I don’t really love working with my Escavon PHP every day, but I’m not ready to make that jump yet and from your perspective is the people who want to hire them into a new position, you want them to be comfortable with talking to you. So, instead of saying, yeah, come in to do a formal application process with us, we have all the suck you can possibly want. You’re going to say, no, just come in for lunch on Thursday, meet the team. Every Thursday we do a tech talk at our company. We invite in some people from outside the company. It’s totally, you know, totally public information. We just talk about what we’re doing recently with DevOps orchestration. Does that sound interesting to you? Yeah, come on in Thursday. They say, wow, these people are cool. They’re learning new technical stuff. I broke bread with several members of their Dev team and none seemed like axe murderers, I could see myself working here.

Or, one of the clichéd things is, hey, do you want to grab coffee? But engineers being the most intelligent people in the room, have like, kind of learned over the last couple of years, that when a hiring manager says, do you want to grab coffee, that that’s like code for, okay, I’m going to buy you a coffee, but then you’re going to listen to my sales pitch. So, try to ask which are above board and honest, but which don’t like presage okay, we’re getting into a formal hiring conversation right now.

And any conversation you’re having with a developer, at any time, no matter who you are in the company, can be upgraded into a hiring conversation. If you’re meeting someone at a meet-up, and you say, hey, what are you doing recently. Oh yeah, I’m freelancing, I’m kind of between jobs, I do Ruby on Rails, my last project was X. If you think, that’s clearly someone we should be talking to, they’re like, that’s really interesting. You know I think we probably want to hear more about it, you mind if I put you in touch with, with Tom, he’s the guy who runs the engineering team over here. Might talk about it in more detail.

If you’re, you know, if you meet a speaker at an event that you host and that speaker does a really good speech, you say, wow, that was really excellent. You want to come in and talk to us about it in a little more detail later? You know, or, you could even say it like, we’re looking for people who bring exactly that kind of execution to their day-to day stuff. We’re looking for exactly people like you, do you want to talk about it in a little more detail? If not, no worries, we’re playing a long game. We’ll be ready to have that conversation in six months, in two years or never, whatever meets your needs, but we want you to know that we appreciated what you just did and we are definitely willing to take this to the next level any time you are.

Ticket prices rise 3rd May 2017

Register for BoS EUROPE Register for BoS USA

Don't Miss a Thing - Get BoS Updates

Want us to let you know about new talk videos, speaker AMAs, Business of Software Conference updates? Join the smart people who get BoS updates. Unsubscribe anytime. We will never sell your email address.

An easy offer that you can always make at the start of any informal hiring process is, since X or because X, we’re going to jump through some of the hoops. Remember people don’t know about your hiring process until they’re actually exposed to it, so they might not look like every other hiring process and every hiring process out there sucks for some reason. Might be because there’s too many interviews, it might be because there’s an intellectually insulting phone screen, it might be because your résumé is bounced among three people who have no idea of the difference between, you know, sequel server and my sequel, but some part of the hiring process is just painful for the candidate and they think that this is true.

So, you should eliminate parts of your hiring process which add no value and that are painful for the candidate, but you should definitely have one sacrificial lamb in it somewhere. Which is like, hey, because we met, we’ve already broken bread together at our office, or, because we have a pre-existing relationship, or because I see from your OSS contributions you’re really good at this, we’ll skip the formality of the phone screen, which is an easy, easy thing, cause the phone screen should only ever be up or down is this person a developer or not.

  • A, it’s an extension of value to the candidates. It’s like, wow, I’m not just going in through the usual pathway. You know, these guys have an active interest in me, whereas the rest of the world only has their standard passive level of interest in me.
  • B, if you represent it to the candidate as, yeah, you know, I’ll go to bat with my team manager and expend a little bit of my capital to make sure that you don’t have to go through the phone screen. They’ll say, oh man, I have to reciprocate with that, you’re going to bat with your manager for me. Okay, I’ll reciprocate, I’ll take that conversation that you just want to have, since it’s only a conversation and since you’re being so good to me.

This is kind of sales one-o-one, but like we learn so many other times in running a software company, sales one-o-one is ridiculously useful when you apply it to other contexts.

Job ads

They exist on the internet, if you’re doing the Cryptopals kind of thing, you’re probably going to rely less on job ads for getting candidate flow. But they are a major channeller for a lot of you guys, for a lot of companies in general, so, you should probably act on it a little bit.

Can you tell me which of these three things is most important for your job ad? Like to keep HR happy, to exhaustively list what requirements apply, or to detail in exquisite detail the benefits you offer. Yeah, like all of these, it’s a trick question. None, none of the above. The only thing your job ad has to do, is get people to talk to you. You’re eventually going to have the sales conversation between your hiring manager and the particular candidate about what actually matters to them. Not, you know, just the overall stuff that you would tell every job seeker.

So, all you need to do is to write your job ad such that out of the universe of possible engineering job ads, most of which are samey and boring and bad, your job ad sticks out.

I apologise for this being in little type, so you probably can’t see why it’s cool. This is the Matasano job ad, that they posted in the Hacker News hiring thread every month. So, the Hacker News hiring thread, has 400 companies to 800 companies all posting a job ad at the same time and they’re all forced to use just a plain text ad which is about this long. And most of them are like, oh yeah, come change the future of API monitoring and we offer standard benefits and dental and a market-leading salary, blah, blah, blah and they all say the same thing. Six hundred companies all saying the same thing. How is a candidate to possibly pick yours out of that mess? And the way that Matasano got recognised, is they’re like everybody else here is telling you, that you should build wonderful software with them. We won’t. We’re going to have you do terrible, terrible things to wonderful software.

We’re going to have the job title here. I don’t know if you can read it and I don’t have it exactly memorised, but it’s something like, not you know, senior pen tester, two years required, it’s, you’re going to be a phantom in the night that little developers stay up worried and have nightmares about. And this job ad has a lot of voice, it has intrigue. It actually tells you a few of their common assessments, but leaves you wanting more. It’s like, wait, how can I get in on this pen testing firmware for a PS3? Or jumping into the New York Stock Exchange and ending up with a shell on it? Which, by the way, in terms of like ethical dilemmas for engineering, if you had root on the New York Stock Exchange, wouldn’t you think about if for just a minute?

Yeah, anyhow, so it says, yeah we have a really, really fun job and you should talk to us more to learn about it. And they got lots and lots of inbound interest from people who are like, your job sounds awesome, it totally doesn’t sound like the other 600 jobs. Tell me a little bit more. And that’s where the conversation begins in earnest, that’s where you start talking about okay, do you have the professional background that would make you appropriate to this, or do we have to teach you up a little bit first via sending you a book, or sending you to Cryptopals? Are you going to be a good fit with our firm and can we make you an offer that meets your requirements?

Assessment

I’ve 80 per cent focused this presentation on the early parts of the hiring funnel for a reason. I think that once you get to the assessment and offer portions of the hiring funnel, that you’re probably already pretty decent at it. And honestly because it’s a funnel, you can get super leverage at the early stages of the funnel for not doing a whole heck of a lot of work, but it takes a lot of work and maybe some controversial recommendations to make your assessment much better than this right now.

Here’s a controversial recommendation. How to read a résumé. I recommend reading résumés on a Linux command line. Here’s the command line for you. M-V space, résumé dot D-O-C space, slash dev, slash null, enter. And I recommended this by the way for nine geeks in a room, that’s just to leak résumés don’t read them, they corrupt your hiring process.

A résumé does not give you meaningful signal about people’s ability to do the work.

The best signal that a résumé gives you is, has this person been on the golden career path from the time they graduated to the current moment in time. And résumés are really good at evaluating for that, but they’re horrible at picking up people who can actually do the work. People who are on the golden career path are not necessarily your best candidates. They’re probably going to be your most expensive candidates, because they’ve been on the golden career path their entire lives. People who are off the golden career path, can oftentimes do the work. Maybe they meandered a little bit, maybe like that high school chemistry teacher from Chicago, who was able to reverse crypto systems in MS Excel. Maybe they used to work at Boeing and then moved to Chicago because their wife got an academic appointment and that got them off the golden career path. Doesn’t mean they can’t do the work.

Instead of relying primarily on résumés, you’re going to rely on work sample tests. Work sample test is simple, you take some portion of what your job does every day, you turn it into an assessment. You have people do the actual job because demonstrated capability to do this work is the best possible indication that someone can actually do the work. This is backed up by academic studies. This is one of those studies that like spoils the conclusion in the title. Practical, theoretical implications of 85 years of research findings, which like, spoiler alert work sample tests, they really work. We’ve known this since 1910.

By the way, the only thing that works better than a work sample test, is a combination of work sample test plus an IQ test. And if you’re hiring in the United States, you’re probably legally from using the IQ tests because of a supreme court case, something versus Duke Power back in the 1970s. So, do work sample tests, they really work.

If you’re developing one for your company, what you want to do is just create a sample application. Do a two week Rails project. Knock out one feature of it, that takes about two hours to implement. Package it all up in a VM such that you can just tell the candidate, okay, get Vagrant working on your system, here’s a quick FAQ on how to do that and then just Vagrant up. There’s an entire working system that you can, you know, view on local host Aviator or whatever. And now here’s the spec for a feature, implement it. And then you give it to the candidate. This is actually not where most of the work comes in in making a work sample test. Most of the work comes in evaluating them. You need a written rubric for your work sample tests and you need it to be in excess of levels of detail. When Matasano did it, their written rubric had a hundred points. Not total scores of a hundred out of a hundred. No, there were one hundred check boxes.

They were at the detail end of, did they test this case but with a failing test? Did they, in this particular part of the feature, did they guard against this kind of attack? Did they guard against this kind of attack? Did they guard against this kind of attack? You want to get granular signal out of your work sample tests.

One way that you can make sure your work sample tests are actually testing work and not the intangibles around a candidate is; you have the person who’s creating the test be someone other than the hiring manager and they just pass them one piece of paper which says, sub score for security 16 out of 20, comments here. Sub score for performance 20 out of 20, comments here. Sub score for blah, blah, blah comments here and then the hiring manager has to actually say, okay, is this person someone we go forward with or not. And if you’re not capturing enough information from those one page evaluations of candidates to make a hiring decision, that means something is going wrong in your process.

And you’re going to keep both the engineering artefacts that people produce and your evaluations of them on file for forever, so that six months from now when you decide, okay, this person that we hired is actually one of our best engineers. We’re going to go back to our hiring test oracle and say what conspicuous about their performance on this hiring test should’ve told us that they were going to be a great engineer. We’re going to put that in our rubric, we’re going to raise the point level. What did they fail on which actually doesn’t really seem to matter? Like if, you know, if their adherence to indentation level on a toy project wasn’t really important for us, maybe we grade that down a few points.

Sell the benefits

Also, an important thing for every hiring process that you’re ever in, even if you make the decision in a job interview, okay this candidate and I were not working out, sell the benefits of the job anyhow. Sell, you know, you should really want to do this. This is a wonderful company, our team is great, we have social license to do smart stuff all day and then go meet our families at five o’clock in the afternoon like actual employed adults. You’ll be doing, you know, we’ll expose you to known technical challenges here and you will do some of the best, best work in your career. And then if you send them an email a week later, it’s like hey, it’s not exactly the right time for us, here’s what we suggest and we’d love hearing from you in the future.

Why did you try selling them on that? Because that person is bouncing around the engineering community for the rest of their days. You want to leave them with a great impression of you and a great impression of your firm, so that their impression is not, oh man, those guys like, terminated the interview on me early. They hate my guts; I’m never going to think of them again. But when they talk to your friends, they’re like, yeah, man, Matasano, I applied there once. I didn’t get in, but, oh that’s my dream job and now I’m hoping to try again in two years. You should really apply to Matasano, they’re great folks and you will actually have people whose friends were denied, well not denied, who were not accepted into your hiring process as reasons why they tried to get a job with you.

Increasing close rates, really quick. So, Steli hasn’t actually presented yet, but you’re going to catch a little bit of Steli’s enthusiasm when he does. His rule is follow up, follow up, follow up. You want to follow up, follow up, follow up at all stages of your process.

Make decisions quickly if you can. At least tell the candidate, like on a daily basis, yeah, we’re still super interested in you. I have to do X and Y and Z and then we’re going to take this to the next level. You get ridiculous leverage out of having a designed offer letter. Costs may be a thousand bucks to get a PDF whipped up for you by a designer in your town which has the standard stuff. Yeah, you get medical and you get dental what not and they have a compelling presentation which is personalised to a particular engineer.

You should have rational expectations about market rate and I talk to a lot of developers to help correct their expectations about market rates, so they make more money. You should not dehumanise engineers, but you should not expect to buy steak for hamburger prices, which you see a lot of hiring managers try to do. If you can’t afford the going market rate for say senior engineers in your town, you need to expand your parameters. Either say okay, we’re going to hire junior engineers and we’re going to train them up internally or we’re going to make riskier hires. We’re going to say, hey, you didn’t have the golden signal on your résumé or in your application, which suggested totally 100 per cent, they’re going to work out. But we’re going to take the risk on you anyhow because you’re willing to come in at a price which is 20 per cent discount to the folks who are a hundred per cent guaranteed.

And should you lose a candidate to another firm, you congratulate them on having a new job, because that is a huge step in their life and you say, congratulations on the new job. Look, I’m going to talk to a hundred other candidates over the next couple of years and I want to make sure that I don’t waste anybody’s time. What’s the thing you liked most about the new job that we don’t have? And always treat candidates like peers, like they’re professionals. I was going to say treat them like human beings but that’s kind of vacuous. So many hiring processes are ruined by treating the candidates like they’re cargo. Like, we’re up here and they’re down here. Don’t do that. Get a deserved reputation in the industry for treating candidates like you want to be treated and that will take you very, very far.

Thanks very much. My email address is up here. Please talk to me about hiring at any time and you’ve been a wonderful audience.

Ticket prices rise 3rd May 2017

Register for BoS EUROPE Register for BoS USA

Don't Miss a Thing - Get BoS Updates

Want us to let you know about new talk videos, speaker AMAs, Business of Software Conference updates? Join the smart people who get BoS updates. Unsubscribe anytime. We will never sell your email address.

Q&A

Audience Question: For context, I’m an engineer, never done the business thing before, but I co-founded a start-up last November. I was fortunate that previous employer really invested in all this. The, you know, I was recruiting as an intern. We ran CTFs, developed CTFs, did a lot of stuff. Do you have thoughts on how to apply a lot of that to hiring on the business side as well, where I’m not as comfortable as to what I should be looking for and the gut doesn’t work and say you know, they’re good at Python?

Patrick McKenzie, Starfighter: Yikes. So, I hate giving an answer which isn’t actually useful for you, but honestly, A, I’ve never actually hired anybody. I was just, you know, doing the CEO thing and taking credit for other people’s work and B, I have never hired for a business position and I have no idea how to do it better. And the focus of our company is really hiring developers better, so, I know what works working with developers. That’s been like the focus of the last 10 years working of my career. I don’t really know how we would go put together for example a sales team. I would suggest that you get coffee later with Steli who really, like, live and breathe that side of the house.

Audience Question: So, I have a six-person engineering team, so I’m not exactly hiring at scale and I think, if you’d have talked to me a month or a year ago I would’ve said that we really have this whole process down and we’ve been able to hire for technical talent, but what we haven’t figured out how to weed out is, sort of the get shit done piece of the equation. So we’ve had two hires that haven’t worked out which is kind of a big deal when you’re that small because they were very, very smart, but didn’t get shit done and so I am curious if you have any filters for that or any way of testing for that or –

Patrick McKenzie, Starfighter: Yep, so, this is a problem for a lot of organisations and for those of you who haven’t heard Joel Spolsky’s vernacular on smart and get shit done, there are some folks who are very smart and talented in like short bursts, but just don’t have the day in and day out execution ability, where they’re given a project in March and then come May that project is actually accomplished, without someone breathing down their neck every day. One of the ways that you can kind of assess for that, is to do things like Cryptopals, which have a longer time horizon than just a single afternoon. So that allows you to get sustained engagement with someone over the course of several weeks and say okay, does this person, when no one is breathing down their neck, when there’s no pay cheque riding on the line for them getting this done, it’s just their own internal motivation. Do they like progress in this or do they say, ahh you know, I could play World of Warcraft instead, I think I’m going to do that.

Another thing you can do is, is look at their previous experience with shipping meaningful things, so look at open source software, try to get signal for what their individual contribution was at previous jobs. And then, the thing I would suggest not doing, which a lot of people might suggest in your context is to do a contract-to-perm kind of thing. The reason I suggest you all avoid that is cause the best engineers have a W2 job right now and the notion of getting into a contracting position with the pot of gold at the end of the rainbow being a W2 job again, is really a non-seller for them. So, yeah I would, maybe attempt to do a more extended engagement before you start making job offers, particularly since you have about six people at the company, and you kind of have a little bit of leisure with your hiring pace. You can, you know, if it takes a few weeks of an engagement with somebody to make a decision on whether you take them down the funnel or not, that’s probably something that’s viable at your scale. And, yeah, I guess that’s the best I have for you at the moment.

Audience Question: Hi Patrick. We’re actively trying to hire a new developer at Moraware and our main non-technical filter is – thank you – Moraware dot com, yes —

Patrick McKenzie, Starfighter: And the reason I love working at Moraware is?

Audience Question: Because I get to work at home, I work with great people, it’s awesome. My question for you is, I’m going to let you talk about Starfighter. Our main non-technical filter is geography. We want someone in the Pacific time zone, but you know, we don’t live in a big city there, we just, you know, we want them in the same time zone for working reasons. Does something like Starfighter help me focus on a specific geography because if we do something like Cryptopals, we’re going to get people all over the world and have to be nice to them even though, you know, only a small slice are going to be actual candidates. Do you understand the problem and do you have ideas for that?

Patrick McKenzie, Starfighter: So, I get the problem. I’m going to push back a little bit with you about the, what you just said, if I do something like Cryptopals, I’m going to get participation from the entire world, which is true and I’m going to have to be nice to them, which is true, and it’s a cost. I think that’s costs you should be willing to pay. Simply because in the inter-connected world we live in right now, the fact that someone is in CST and not a candidate for you, doesn’t mean that they won’t be living in PST two years from now, doesn’t mean that they’re not college buddies with five people in the Pacific time zone. So, I would say, you know, just the cost of doing business that’s crying out to be wonderful to people everywhere.

Starfighter, my quick 30 second pitch. We’re making games which engineers play by programming. We’re going to evaluate which of the engineers we think did exceptionally well during our games played by programming. We’re going to reach out to them and say, hey, are you happy in your job right now? If you’re not happy in your job, you were really freaking good at doing this big data to catch for example, catch an insider trader on a simulated stock exchange which we mocked up for you. You’re really good at this, you should be, you should be doing a job that makes you happy. What would it take to make you happy?

And then we, then we have a 30-minute conversation between one of our founders and them and they might say, you know, I’m working at a company that’s building CRUD apps every day and I feel like I’ve maxed out on my skill for CRUD apps. I want to do something more meaningful, I’ve been playing with big data in my spare time recently. And I’d say something like, that’s really interesting. You know, well you have demonstrable expertise with this. One thing you can do with big data is identify fraudulent transactions which are occurring on a FinTech company, for example, so if that sounds interesting to you I could put you on the phone with somebody at, say Stripe tomorrow, who is trying to hire for their fraud team. Would that sound interesting?

And so, for example if Moraware was a client of ours you can tell me, hey, we’re looking for people in PST who, who match X, Y and Z. And then when I’m doing my quick geek-to-geek matching algorithm in my head, I’m like, oh yeah where are you based? Oh, you’re based out of a, based out of a small town in New Oregon, that’s awesome. So, what are you looking for? Oh, I’m looking to join a good team with good engineering management, which will help support me for the next couple of years of my career. I’m like, I’ve got a — so you probably haven’t heard of, but let me tell you a little bit about this company called Moraware. They’re a great company to work for, for these reasons. Give me some ammunition. And, and if it would be interesting to you, I could put you in touch with my friend Patrick there tomorrow. Would you like to make that happen? And if they do great. If they say no, that’s not for me then no skin off anybody’s nose. And then the ultimate goal for a start, well, the ultimate economic engine at Starfighter is working since the recruiter / he recruited her, if you hire through us then you write a very big cheque like, the number suggested early on in the presentation.

Mark Littlewood: Fantastic. Patrick, thank you very much. Always an absolute pleasure and what a great way to kick off day two.

Ticket prices rise 3rd May 2017

Register for BoS EUROPE Register for BoS USA

Don't Miss a Thing - Get BoS Updates

Want us to let you know about new talk videos, speaker AMAs, Business of Software Conference updates? Join the smart people who get BoS updates. Unsubscribe anytime. We will never sell your email address.