My RubyConf India 2010 Talk Highlights

Hashrocket sponsored the inaugural RubyConf India event primarily to support the nascent, yet growing community of Rubyists here, but also due to my personal love of Indian people and culture. I've been eager to get back to Bangalore for the last couple of years and this was the perfect excuse. I deliberated about what talk to deliver to this audience. At all costs I wanted it to be entertaining and provocative. In my opinion, too many Indian service companies are "doing it wrong" -- they stick to the stereotypical modus operandi of offshore providers, with waterfall processes and fixed bid. I wanted to impress on the audience that there is a better, more fun way to run your business. After opening comments by the organizers and Roy Singham, Chairman of ThoughtWorks, I delivered the second general audience talk.

I was charged up with the right kind of anxious energy and this being material that I know inside and out, it came straight from the heart. The talk was not entirely new, previous versions of it had similar material. However, I did go through and restructure the outline significantly as well as adding new material based on the last 12 months of operations at Hashrocket. Luckily I also added some elements that I knew would appeal to my audience -- slides based on the blockbuster Indian movie "3 Idiots". Little that I know how important the concept of "All Izz Well" would prove to be to myself!

On a scale of 1 to 10, I would rate myself a 9 on delivery. I was funny and engaging, making eye contact with the audience members and spending much time in storytelling mode. Unfortunately, I would have to deduct points for making the mistake of dropping the "F-bomb" too many times and as a result getting interrupted and vocally dressed-down by Tina Vinod halfway through! Somehow, although nervousness infected my voice and I started forgetting to breathe, I did pat my heart and tell myself audibly "all izz well, all izz well" and charged ahead with my talk -- the audience got a good laugh at that. Other challenges included the organizer mistakenly telling me I was out of time when I actually had at least 15 minutes left! Luckily my Keynote display presenter mode with its timer saved me. I did pick up the pace a bit, but finished in 48 minutes, just 3 minutes over my alloted time.

Reading List

In the section of the talk about Process and Practices, I mentioned that it is necessary to go "by the book" when starting out. Too many people criticize practices that they haven't properly implemented. You must learn to crawl (so-to-speak) before you can evolve your own practices. 

The list of books I recommended constitute a large part of the orthodoxy of progressive software development nowadays. I asked the audience, how many of you have read at least half of these books. Sadly, only a dozen or so hands (out of 400) shot up. I truly believe that you cannot be effective nowadays without studying these books and applying them to your day to day.

Agile Software Development, Principles, Patterns, and Practices

Agile Retrospectives: Making Good Teams Great

Apprenticeship Patterns: Guidance for the Aspiring Software Craftsman

Design Patterns: Elements of Reusable Object-Oriented Software

Domain-Driven Design: Tackling Complexity in the Heart of Software

Extreme Programming Explained: Embrace Change (2nd Edition)

Refactoring: Improving the Design of Existing Code

Secrets of Consulting: A Guide to Giving and Getting Advice Successfully

Software Craftsmanship: The New Imperative

Test Driven Development: By Example

The Principles of Product Development Flow: Second Generation Lean Product Development

User Stories Applied: For Agile Software Development

Working Effectively with Legacy Code

 After you've mastered the fundamentals, then you can begin to evolve your process and practices to actually achieve the best results for your context. I must have said 20 times during my talk that "context is king."

 

One of my favorite and most controversial pieces of advice, at least based on the feedback and comments that it generated:

"I don't defend breathing... I don't defend Agile. It just is."

I believe unequivocally that the time to be evangelizing and advocating proven Agile techniques is over. It's been at least 10 years since Extreme Programming hit the scene. We know that TDD yields excellent results. We know that pair programming yields excellent results. We won. To continue being in defensive mode against stupidly obstinate management is a waste of time. It may take a long time from now, but eventually people that don't work in smart Agile ways will fail and go away. As for you, concentrate on being excellent, perfecting your craft and building a loyal and devoted group of clients that will provide you with strong references and testimonials.

If you want to put into practice my advice, refuse to work for clients for whom success is not 100% important. This will necessarily exclude most large enterprise clients, where politics and status quo are as important as results. Those types of clients will drag you down and make you compromise your values. Startups and entrepreneurs usually are 100% focused on successful results. These stakeholders will not be worried about your technology choices or whether you want to pair program and do test-first. They will let you do your job, and the way that you sell them on using you is with successful track record, demonstrated by copious amounts of testimonials from previous clients. I stressed the importance of photos and videos to capture the spirit of your success in a vivid and emotional way. We have the technology -- use it!

 

My advice to fire clients if they need to be fired was perhaps shocking. It shouldn't be. I've only had to fire a client a few times in the history of Hashrocket. That's considering a list of now almost 60 happy clients of all shapes and sizes. The key to not having to fire clients, is to not get the wrong clients to begin with. This is known as being able to "choose your clients." It's not as hard at it seems. Sure, when you're starting out it is somewhat difficult to put this advice into effect and you will make mistakes. That's when it's important to have the confidence to cut ties and look for greener pastures.

"Never allow your clients to corrupt your values or integrity."

The worst thing that can possibly happen, more than having money difficulties, is to corrupt yourself by compromising on your values and integrity due to pressure from a bad client. I'm not saying to be inflexible! I'm saying that there are core values that will allow you to reach for excellence and achieve it. You won't know what those values are when you're starting out, but you can be informed by studying the successful results of others and internalizing that knowledge. Then when you're on a roll and doing great work for your clients, you will realize what those core values are. From that moment on, no matter how big the pressure is, you should not compromise. The more confident you are about your way of working, the easier it will be for you to sell your services and acquire quality clients. The last thing you ever want to do is to appear desperate -- at that point great clients will go away running and bad clients, those that will take the worst advantage of you will become very friendly! Never allow your clients to compromise you. Stick to your values and focus on success.

When I'm evaluating software developers to hire, I don't give much credence to resumes, cover letters, referrals or work history. The only things I care about are work ethic and cultural fit. And there's only one way to determine whether a candidate satisfies my criteria for success -- actually see them in action over the course of a week or more. Since we always pair program, it is trivial to do short contract gigs with qualified candidates and bring them in putting them on multiple real projects. Seeing them working on real code in real situations is very instructional. Again, context is king. Over the course of time, we've created a very desirable work environment at Hashrocket, which means that there is a strong demand for people to come work with us. That means I can diverge easily from traditional hiring models.

Conclusion

Indian business culture, at least the stereotypes that are propagated to the west, doesn't strike me as particularly fun. But if you're doing great work for your clients and sticking to your core values, then you're going to have a lot of success and be able to enjoy it by building a very fun work environment. My final series of slides was a photo set of various fun activities that Hashrocket folks engage in on a regular basis, from participating in user groups, to music and sporting events, games, having parties, drinking and dancing and doing crazy stunts like all jumping into a swimming pool at the same time.

I truly believe that if you're not having fun at work, you're doing it wrong. A happy, healthy environment is one that leads to the best results. However, you can't impose fun on a group by fiat, just like you can't create culture. It has to emerge from hiring happy, progressive folks that want to do the best work possible while sticking to strong core values. This is the final message of my talk and I hope that the audience found it inspiring and worth passing on...