Tom Dalton :: Doer of Good

Here's the toughest job interview question I have ever been asked.

Question: Tell me about a moral decision you faced where you came down on the wrong side.

Answer: (In real life, I paused here for a pretty long time.)

I've been interviewing programmers, and I thought it would be handy to write down some of the questions that have yielded the most interesting and helpful answers. Here's my Top Five Questions to ask someone who's applying for a PHP programming job:

1. How did you learn PHP?

If the answer is something like, "I took a class at college," we've got a problem. If the answer is, "I took three classes at college," it's even worse. The person who learned it working on a contract project, late at night, reading a hundred different websites and books -- that's the one who's going to be able to learn the next language they'll need. Or learn how to use the crazy API that one of our partners releases. Or learn how that stupid old legacy code works.

What websites did they use? What books? It doesn't even matter what the answers are, really, as long as they have answers.

2. How would you write a web-based tool that would give me all the cities within a specified distance of a certain zip code or city? (Then...) How would you make it better? What would version 2.0 be like?

I want to see what data structures they're thinking about. What technologies are the comfortable with? Do they consider login and user managment? How would this tool use cookies? Can they think of ways to get the source data they'll need? Do they consider that many users may be international?

Just for fun, if anyone mentions XML, I'd ask some very pointed questions about that. It's a buzzword, and people tend to throw it in because they think it sounds good. I want to know what concrete benefits XML (or any other specific technology) would bring to this project.

3. What was your least favorite job?

This is a good one to let people talk on and on about. For some reason, programmers just don't know when to stop. And pretty quick, many people will reveal details that really show what they're like to work with. If a candidate tells me, for instance, that he hated that one company was always sending around memos with ridiculous policies and requiring everyone to read them -- I've got to worry a bit.

4. What's the coolest project you've ever worked on?

They'll try to turn everything into this, so make sure it was genuinely cool. I like to ask technical follow-up questions to this one, to see if they really understand every aspect of it.

5. What other companies are you applying at?

There is a wrong answer to this question. That is any restaurant, retail store, or just 'none.' The answer to this question should be a list of my competitors, or at least some local, related companies. Where I happen to be right now, there are tons of comparable technology companies, and some significantly bigger. If somebody's applying here and not there, I have to wonder why. Have they already been rejected there? Did they just stumble across us because we posted in right place? A serious programmer looking for a job ought to be applying at a host of companies in the area.

I view this question as a good indicator of a candidate's individual drive and work ethic. If they're not working hard now, why would they work hard once they have a secure job and comfortable place?

I'd be interested to know what anyone else thinks of these questions. Comments are open!

This is a question I think more people should ask, when conducting job interviews. The answer to this shows several things about the candidate:

1. What do they *really* love to do? What will they excel at?
2. Are they enthusiastic about work?
3. How do they view their successes? Was it the result of great teamwork, individual effort, lucky circumstances, or something else?

The whole point of a job interview is to figure out if a person will fit in a position or company. The work they love to do is what they will tend to do, so this question really addresses that core challenge.

Lecture over, here's my stab at an answer:

One ofwhat I would consider my "greatest successes" was the development of Search Engine Optimization services for 10x Marketing. We knew we needed to offer something -- we even had a few clients paying us for SEO services -- but we had no formal, scalable solution. Our CEO asked me if I would become an "SEO guru" and see what we could do there.

I knew nothing about SEO. Over the next few weeks and months, I found some books and a lot of websites about what I soon learned was a rapidly growing and very young industry. I visited the websites of every SEO company I could find, from the mom-and-pop shops to the New York ad agencies with red-carpet client lists. I heard many ideas, many contradictory and almost all unsubstantiated.

In the midst of the chaos lay principles, however. The dissertation of Sergei and Brin, outlining a primitive vision of Page Rank. And a core of ideas that fit our goal of developing a consistent, scalable, honest approach to SEO.

I evaluated tactics both for their short-term effectiveness and their long-term viability. I eliminated anything that wouldn't work if we had a Google employee sitting in the room with us. I tested a ton of ideas, measuring the success of each and weighing the impact of various tactics.

Finally, we had the guts of a process that would work. But it would take work to run. Months of work, in fact, for a single client.

So I turned to another set of books and websites, and I started teaching myself Java. (The language of choice, if only because it had the most readily adaptable set of example programs that I could work with!) If we could have afforded a programmer, I'm sure we would have ended up with much 'cleaner' code than what I produced. But the little program I created was finished after a month of development, and when we ran it -- it worked. Optify could do in two days what would have taken a person two months to do manually.

Two years later, we're still using that program. I've had to adjust some of the formulas and features, and I've expanded it to support many different languages as we've gotten clients from overseas. I had to build a distributed proxy server system when we started overtaxing our own servers and resources. As search engines have changed, I've modified it to match and to support new engines that have come online.

And our company has grown fantastically. SEO has become our primary service. Referrals from clients is our single best source of new leads. And our clients -- who typically come to us skeptical, having been burned two or three times by other SEO companies -- are growing and growing.

I'm not a programmer, and I'm not an SEO guy. I'm just a guy who does his best at whatever needs to be done. When that happened to be SEO and programming, I did a really good job.

Now I'm working on Conversion Rate Enhancement. I'm excited to see where that leads!

I've had a hard time answering this question in job interviews, because I try to be honest in job interviews, even when I know that's not what they really want. What they want is the *mostly* completely honest answer:

"I want a challenging, really hard job with enough flexibility for me to learn and accomplish great things."

That's true. But I have a more concrete idea of what my ideal job would be.

I want to work with electronics and the Internet -- I want to integrate broadly disparate systems in innovative ways. I want to deal with security and usability. Honestly, I view marketing as a usability problem in many ways.

I want to live overseas and travel, and I want to be able to research and create effective solutions to impossible problems. I'd like a job where people have problems in their foreign offices and need somebody to come in for a day or a week and assess the problem, then create something on the spot or go back home and build or buy or find something new. And then come back and install it, train everybody, and then support it.

My skills are in Internet technology. I've created prototype systems to simulate Internet connections on local machines or pass stegonographic messages through eBay. I've faked headers and forged connections between systems that couldn't talk to each other and I've created macros and programs to convert data and streams to match crazy specifications. Some of the things I've put together saved weeks or even months of human work. (One system in particular has enabled work that would have taken years of effort -- and might have been profitable even at that level of effort.)

Most of these things were done very, very quickly. I work best under pressure and in bleak situations. If we've got months and millions of dollars, we can find people to do it better than me. :o) But if it needs to be done yesterday and there's no team of developers sitting around to do it, I can typically find a way.

I'm also good at communicating with people -- execs, salespeople, tech people, whatever. I can teach and train -- I've taught at BYU and in a host of other settings. I can understand people's needs well enough to find solutions that actually work for them. (A trait not common among developers, sadly.)

So, what kind of job fits some of that?

Monster doesn't have many that seemed to fit the bill.

(And, as always, I'm mightily arrogant, I know.)