Wednesday, June 6, 2012

Dear CR: I Need To Hire Engineer #1. Can You Help?!


My co-founders and I just raised some seed capital for our startup and now we're ready to grow our tech team! Can you help us recruit our first software developer? I've never worked with a recruiter before, but I've heard helping startups grow is what you do! We currently have 4 people, 3 founders and an awesomely fabulous office manager.



Your company is at a stage where you need every dime you've got. I'd love to help you, but there's a good chance you're not ready yet. Here's why:

  • You don't know how to value recruiting services.

    If I send you a bill for a couple thousand dollars and you haven't made a hire yet, how are you going to feel about it? My past experience in trying to help super early stage companies tells me you're at a stage where taking risks that cost several thousand dollars and don't go directly into building your product aren't a good bet for you.

  • Your needs are going to change.

    If I spend a couple weeks looking for A, you're likely going to discover that you really need B. Can you afford to spend money on something that doesn't produce results immediately? Probably not. It's the same reason you probably don't want to be spending a lot money on PR, marketing, advertising, or a big fancy sign in front of your office. Premium services can be quite effective, but getting you to the point where you really know what you want is expensive at your size.

  • You'll respect me later if I don't take your money now.

    Even if you said you'd be willing to take a risk and hire Captain Recruiter knowing it might not work out, you'll never really feel good about. I'd rather take you out to lunch, answer a few of your questions, and make a recommendation on how to do recruiting yourself on a budget. When you've reached the next stage, you'll trust me more since I didn't take advantage of you when you only had a few bucks to spend.

  • Recruiting is inherently unpredictable.

    I've stopped trying to guess what people are going to do. Even when I manage to be perfectly consistent, clients and candidates throw me curveballs all the time. For example, if I tell you that a filling a position costs $X and you can afford to spend $X, you may be thinking that I can help. But it's easy to forget that $X is the average; filling half the positions cost more than $X! The unpredictable nature of recruiting and people makes recruiting more suitable for companies with deeper pockets who can let the costs of filling a position average out over time.

It's an exciting time to be you! You've got big dreams and money in the bank that buys you the time to chase those dreams. Be wise with your money, making small bets and taking calculated risks. When you're ready to scale, I'll be here to help you build the team you need.


Job Matching Algorithms Don't Work Well

There's an article in today's Wall Street Journal called Seeking Software Fix for Job-Search Game. The article talks about the race to develop technology that matches people to jobs. This is also my first time being interviewed by a major news publication!

I like the idea of a creating a job matching algorithm. How cool would it be to have a computer automagically find you your next job?! That'd be frikkin' sweet! There's only one problem... job matching algorithms are fatally flawed.

Here's a few reasons:

  • Good algorithm, crappy data.

    The job ads written by most companies are crap. These ads are too long and don't clearly state what the company really needs. The resumes most candidates write are also crap. I look at a lot of resumes, and most of them make my eyes bleed. No algorithm can address the fact that the data being analyzed is crap.

  • Algorithms are obsolete.

    I used to get a lot of irrelevant job applications. I honestly thought job seekers don't read job ads, but it turns out they actually do. So I got better at writing job ads and drastically reduced the number of people who weren't a fit. To help sort the people who do apply, I simply ask them a few targeted questions. "I need X. Yes or no -- Can you do X?" It turns out that people generally answer truthfully when given a chance to do so.

  • Great algorithms might be unintentionally racist.

    I met a guy a couple years ago who had built a job matching algorithm. This guy was a pretty hardcore technologist and science-y type. He said he'd raised millions of dollars to develop this algorithm, but it died for a very simple reason -- it supposedly worked so well it could tell you whether a person of a particular racial background was more or less likely to get the job. As soon as race entered the picture, he couldn't sell it. No HR department in the world wants to get anywhere near something that could accurately quantify or even imply that they make racially-based hiring decisions.

Until companies, recruiters, and job seekers are able to effectively communicate about what it takes to do a job well, I fear we're all going to be stuck wasting each other's time trying on job applications that were never really going to go anywhere.

Tuesday, June 5, 2012

Recruiting Good, Affordable Engineers During A Talent Crunch

You've recruited all of your coder friends, but you still need to fill open positions on your software team. As soon as you look outside your own network of contacts, you're going to quickly learn that every company like yours is struggling to find software engineers that don't suck at a price that doesn't break the bank. You're not in a position to acqui-hire a pre-existing team of developers, hiring Pivotal Labs blows your long-term budget out of the water, and many of the applicants you're finding don't have the skills for which you are asking. Are you just totally screwed? No! It's possible to build a great software team on a budget if you're willing to work within real world constraints.

Here's an incomplete list of things you can try to find a coder on a budget:

  • Focus on a candidate's ability to grow with your needs.

    If you can't find someone who has the skills you need, try finding someone who can learn what you need. Test a candidate's ability to learn a new skill or simply look for a candidate who enjoys picking up new skills for fun. There's a reasonable chance are a candidate who knows 2 or 3 programming languages really well will be conditioned and willing to pick up whatever tools you need them to learn.

  • Ask everyone you meet if they happen to be a software engineer.

    If they are, now you're talking to a potential hire. I recently struck up a conversation with a person walking down the sidewalk in Palo Alto. It turned out he was a software engineer visiting from out of town to attend a conference. I asked if he'd be interested in checking out my client's office and 15 minutes after I met him he was interviewing with the hiring manager.

  • Be less Agile.

    I see a lot of companies become so fixated using Agile processes because they lose sight of the fact that a processes is meant to serve the team, not the other way around. Your end goal is to ship great product. If you find yourself turning away a large number of good engineers because they don't fit into your Agile paradigm, you may be building a monoculture of rules lawyers that lack perspective on alternate points of view. Good software was written by talented people before Agile came along.

  • Lower your expectations.

    Contrary to popular belief, not everyone on your team needs to be a superstar. Some of the work you need done is tedious, time consuming, and not that challenging. There are people who enjoy that type of work. Just remember that even on non-critical tasks, you still need someone who possesses the judgement necessary to do the job well.

  • Put people to work.

    At some point, you'll talk to an engineer that you really want to hire, but you can't quite seeing them being effective. Don't waste time trying to predict the whether someone would work out. Hire the person for 2 days or 2 weeks. Give them a shot of showing what they can do. Be prepared to lose a little money using this technique, not everyone will work out. The one's who do will not only be proven, they'll feel more confident, secure, and wanted in their position when they realize you took the time to actually get to know them.

  • Make more friends.

    Chances are you'll need to hire again in the future. Get out there. Meet people who do what you need and then hire them.

  • Implement a training program.

    It's natural to think that training is expensive and you can't afford it as a company. That's not true. Here's a simple example. Let's say a less experienced engineer costs you $X per year, and a more experienced engineer costs you $X+40K per year. If you plan on retaining them for 2 years, what kind of training program could implement for $80K? I don't know, but you can use it over and over again, and it's probably worth at least exploring as a viable option.

Monday, June 4, 2012

5 Questions To Ask Before Recruiting A Sys Admin

Your business has grown to the point where you need a full-time person to manage your computer systems, so decide it's time to hire a Systems Administrator. If you're being honest, the job title should probably be Person-Who-Can-Figure-Out-My-Systems-That-Weren't-Designed-To-Scale-While-I'm-Trying-To-Scale Administrator. How the heck do you find someone who can come in and run systems they've never seen before? I've probably recruited dozens of sys admins at this point, but I still hard to spot a good one.

Based on my experience, I've come up with 5 questions I like a hiring manager to ask and answer before recruiting a sys admin:
  1. Do I want a DevOps engineer?

    Some systems administration roles really fall under DevOps. A good DevOps engineer will deal with things like writing SQL queries, writing custom software when no other solution exists, working with cloud systems that are closer to programming than doing things by hand, and dealing with all software below the application layer. If you really need a programmer but end up hiring a faux-grammer with programming languages listed on their resume, this is not good.

  2. What are the 3 toughest problems our new sys admin will tackle in your environment?

    Think about what your sys admin really needs to know well, then think of scenarios you can discuss with your sys admin. By talking with someone about why you need something, you can see how they'll address the problem. For example, if you need to store your customers' credit card information, you can see how a candidate's healthy fear of that information getting out affects their approach to securing a system.

  3. In the next 3-6 months, with which projects will my sys admin need to contribute well?

    Talk about a big project you'd like to take on. Ask the candidate how they plan for change, deal w/ uncertainty, and get them to provide details about large projects they've taken on in the past. Ask them about any concerns they might have or what information they would like to see in order to make better decisions, too.

  4. Does anyone around here know how to interview a sys admin?

    If no one on your team knows how to interview a sys admin, let alone do it well, don't be afraid to ask for help. Track down an experienced person, preferably someone who knows you and specifics about your company's systems, to help screen candidates. You can also tell the sys admin you need their help in determining whether or not they're the right fit; ask the sys admin what you should be looking for and see if they can help increase your confidence in their application.

  5. How does this sys admin tie into the business' business model?

    Your company's products and/or services are important. If your sys admin will be working directly on the systems that make you money, they should be able to appreciate the importance of keeping that money machine running at all critical times. For example, a company selling toys to American kids wants a sys admin who thinks about what happens to your systems between Thanksgiving and Christmas.

Monday, March 5, 2012

You Can Be Steve Jobs!

A few times a month I bump into an article written for entrepreneurs where the author tells the reader "You're not Steve Jobs." Usually the main point is that Steve Jobs was a unique phenomenon and trying to emulate him is a failing strategy. These articles annoy me, mostly because they start off with a negative premise.

I'd much rather read an article titled "You can be Steve Jobs." Can I actually become Steve Jobs? Of course not. But it's a lot more interesting to read about how to live up to my potential than to have some Internet Troll be soul-crushing!

Here's my 2 cents on emulating Steve Jobs and/or your favorite role model:
  • Focus on what inspires you. All of your role models from Steve Jobs to Superman are deeply flawed, but they all have traits worth paying attention to. Steve Jobs taught me that big things can be accomplished if you learn how to work with amazing people.

  • Study the limitations of successful people. How did Abraham Lincoln become president while suffering many political defeats and being massively depressed? Maybe it's worth studying why he got shot, too, because that's something I'd like to avoid.

Thursday, November 17, 2011

Moneyball For Recruiting

I really enjoyed the movie Moneyball, probably because I love the idea of an old and stuffy way of doing things being disrupted by a more innovative approach. In the recruiting world a lot of people are excited about applying quantitative analysis in some sort of Moneyball approach to recruiting.

In my opinion, the recruiting business is a long way off from offering a really interesting quantitative tool one can just buy off the shelf. Here are a few reasons, in no particular order:
  • Companies don't share their recruiting data in any meaningful way. Google and Facebook fight over talent all the time. Each company has it's own tips and tricks for attracting good people, but they're not going to share that talent with each other. If you can't compare the recruiting processes of each company, you can't figure out the common denominator of what does and doesn't work.

  • Applicant tracking systems like Jobvite and Taleo have access to pan-company data, but they aren't in the business of selling their customers data. My impression is that these companies lockdown their customers' data pretty tight and can't/don't/won't share much at all.

  • Recruiting databases and job boards like and are in the information hoarding and selling business. I doubt they'll share their trade secrets with you and I.

  • Staffing agencies like Robert Half and ReadyForce are in the information business, too. Good luck getting them to give you access to their secrets.

  • Social services liked LinkedIn and BranchOut are another breed of information hoarders.

  • I'm sure there are services that exist that would love to help job seekers and employers come together based on a data driven approach, but these services are virtually unknown and lack wide exposure or adoption.

What I really want to see is a vertically integrated recruiting service actively working to connect with employers and job seekers and then sharing every bit of non-secret information possible. Get a venture capitalist to fund that idea please :)

Monday, October 31, 2011

Breaking Out Of A Big Company

During a recent conversation with a friend working at a large software company, he told me that in the next couple of years he'd like to move on and join a small or medium sized company. During our conversation I was talking with him about how he should start positioning himself to appear as if he's prepared to work for a smaller company, and he didn't quite understand what I meant.

In my experience, people from larger companies aren't conditioned for life at a startup. There are 1000 little things you have to do yourself that are taken care of for you in a larger company. Take a look at this article, for example (though Captain Recruiter doesn't necessarily agree with the tone of the piece). And even if he's ready 100% to go work for a smaller organization, or he's worked at a startup before, I'd be willing to bet that a hiring manager may think it'd be too hard for him to acclimate to startup life again.

In his mind, he's already conditioned to work in smaller company without big company resources. In his job now he's forced to roll up his sleeves and deal with all sorts of things, from corporate shenanigans to deadlines to managing third party relationships. It's not really possible for me to say with certainty whether he's prepped for small company life or not, so when he asks me why I think he might not be ready, I find it hard to give him a good answer. All I can tell him are what red flags a startup might see when they look at his resume.

After realizing that there was no way to conclusively say that he is or isn't prepared for work at a small company, I recommended a different approach. My suggestion was that he call up 3-5 people that might hire someone like him and ask them what they look for in a new employee. He may as well go straight to the source.