Vlad Antohi, a budding UX designer, was kind enough to put together the following notes on the event that he was willing to share with the MAB community:
Brian Cardarella - Founder, DockYard (Twitter, LinkedIn)
Mark Kasdorf - Founder, Intrepid Pursuits (Twitter, LinkedIn)
Jeremy Weiskotten - Co-Founder, Terrible Labs (Twitter, LinkedIn)
Matt Chisholm - Founder, Cantina (LinkedIn)
Ross Beyeler - Founder, Growth Spark (Moderator, Twitter, LinkedIn)
How big is each agency?
DockYard - 20
Intrepid - 40 @ 4 years old
Terrible Labs - 13
Cantina - 50+
How do you structure your agency?
DockYard - structured by ownership percentage. Founder has lion’s share and deals with sales and closing deals, also have an office manager and PM on business side. have teams under Creative Director and Lead Engineer.
Intrepid - Five Functional Areas, two not currently populated
- Under CTO = VP’s of Engineering, Product, and Delivery
- VP Sales & VP Marketing = not yet used, roles carved out for future
- Director of Finance (half time PM)
- Director of Operations (also handles HR and recruiting - targeting 2 hire/mo.)
Terrible Labs - Fairly flat structure, has a Biz Dev and Ops person and rest of organization is between two co-founders. One handles PM/Sales/Mentorship, the other is the Lead Engineer. They use Daily Stand Ups and have a Dev Lead.
What was the first inflection point in your growth, where you had to do something different?
DockYard - Growth to headcount of 20 has been at the cost of founder taking no salary. They have 20% margins after no salary, and grew thus far without a real plan and unclear what the next steps were. Growing organically and hiring the smartest, best people they could without necessarily making the best business decision.
Cantina - Hire the most senior people you can, because they can manage themselves. As you grow, people want more money. How do you define career growth at your company? Everyone wants to grow and you need to support them and pay them more by finding a dynamic that makes that viable - whether increasing the profit per employee, or selling higher value services.
Terrible Labs - The growth points all came from running into pain, into friction. They emerged organically during the retrospectives and daily stand ups. One example is not being able to do solo projects anymore, so now they sell a pair for larger projects. There is a process to tackling specific pain points - first one must identify them, then be honest about them, then try something and see if it works. Repeat until it works.
Intrepid - There is a big change that happens each time headcount doubles. From 1 to 2 is big - whoa, now I have a partner. From 2 to 4 is big - wow, I need to do payroll now. From 4 to 8, things like health insurance become a priority. From 8 to 16, suddenly you have to be careful about burn rate because you are spending a lot of money per month to keep the party going. Year over year, for the past four years, the company has doubled. Each time it does, 80% of founder’s job completely changes.
What processes do you use?
DockYard - Two things were most helpful. One was working with a design firm to get the idea out of the client’s head. Engineers lack that skillset. Second thing was Project Management. Hiring a PM who was at Fidelity for 10 years and ready to go on off on his own, totally transformed their process for starting projects and ensuring they are on time and budget. Also very excited by SWIFT development for iOS and the new design possibilities it holds.
Terrible Labs - They use Google Ventures model. Bill out Development as Time & Materials. Fixed bid creates contention and incentivizes clients to never be done. Doing retrospectives after each project drives process changes that make a big difference.
Intrepid - They bill Agile Development, no fixed bid, weekly rate of $6000/manweek. Terms are Net 15, if they don’t get paid they stop working. They also have a fixed fee 4-6 week Design and Discovery phase where they do problem definition, user stories, user flows, wireframe, look and feel, and a three screen prototype to help sell the project to whoever needs to fund the development. They got PM’s at 20 people, and shifted their pitch to bill the PM’s out to the client. Makes a huge difference. Locked down development process to be repeatable. No cowboys, one standard. Huge impact on quality of life and margin. In terms of estimates, founder doubles everything developer says, so that they can be sure to come in under estimate (often not by much). Just apply a straight 50% velocity, double everything the engineer says. They have a 6-10 week wait, so that filters out folks who are not serious.
What is your recruiting process like?
Cantina - Works with 5 recruiting firms, but the best hires come from referrals, because they are a known quantity. They put their project manager candidates through exercises and find that it works well. Culture fit is important, but it’s also important to cultivate diversity. If everyone likes someone, but they feel something is off - it’s important to identify what that thing that is “off” is, because it might be just what the team needs - a new perspective. You have to be pragmatic about hiring - do they have the discipline to produce the work, and can they get along with others and add value culturally.
DockYard - The first year, the founder just hired his friends and people who came to him through his running the Boston Ruby Meetup. They focus on high quality deliverables and work to show off, and the portfolio attracts talented folks to them. They also differentiate by focusing on Ember.JS, which is a niche framework. They have two of the core team contributors in their company, run a conference on it, and the founder is writing a book about it. Being an early adopter of a niche technology that has legs is a good indicator that someone is smart and would be a good fit. They play down the Rails part, because it may have “jumped the shark” to the point where the folks getting into it now are not the smart, passionate people DockYard is seeking, but rather opportunists who heard it’s a good way to make a buck.
Intrepid - Half the company has 0-2 years of work experience. They heavily leverage their Apprentice program, running two classes of 10-20 people each. Pay them half salary, do 6 months of training via 4 weeks of classes, then 6-10 months of a capstone project, then 3 months of billable work - charged to the client as half rate or free, until they get to an acceptable velocity. For senior people, they write down a problem and forward it through their networks via email, who forward it along and it amplifies. This “relationship based” recruiting allowed them to grow by 15 people this year and brings good talent. The “magic ratio” is:
- 1 Senior Architect (has solved 80% of problems that come up before, 20+ years experience)
- 3 Lead Developers (able to solve 80% of problems that come up, 3-8 years experience)
- 9 Junior Developers (competent, can take down story points, need guidance at times, 1-2 years experience)
Whenever projects go sideways, it’s typically because this ratio is skewed towards too many junior developers off on their own without guidance.
Terrible Labs - No recruiters. Need to have a personal relationship first, they spend more time with each other than with their families. They look for people who want to keep learning and growing. There is a glut of eager junior candidates who are teachable and can even teach them some things. Stay away from people who think they know it all, because the technology is advancing so rapidly that it’s a sure sign of stagnation and arrogance.
What about hiring or firing friends?
DockYard - Founder had to fire almost all of the friends he hired. It’s difficult to make the relationship transition. “If something doesn’t feel right, it’s probably not going to fix itself.” Stay away from hiring folks with previously existing relationships.
What are your thoughts on turning down business that’s not a good fit?
DockYard - They were just moving into enterprise, and the founder knew the terms and the rate he wanted. The deal was pivotal - if they did not get it, the company would fail. But the founder was not willing to compromise his terms because he knew it was necessary. The Biz Dev/Account Manager was not able to get the correct terms, so they rejected the deal, and then the founder stayed up until 3am thinking about it until, in a flash of insight, he figured out how to sell it. Came in the next day, fired the Biz Dev guy, resold the client on the new terms and got the business. Saved the company.
Terrible Labs - They make an effort to filter the bad clients out early. A couple of bad projects they ended up doing, they probably could’ve sniffed them out. “Your business is as much about what you don’t do as what you do.”
Intrepid - They have four factors, any three of which must be present in order for them to take the business. 1) Profitability (do they have budget, are the margins there), 2) Strategically Relevant (will they learn something new, is it a good portfolio piece), 3) Enjoyment (is it fun or new for the team), and 4) Red Flags (will it be high maintenance). Willing to skate on any one, as long as the other three are there. If it’s not a fit, they don’t say “No” as much as they just say they are too busy, and humbly introduce them to someone else.
How do you niche down and differentiate yourself?
Intrepid - We do native iOS and Android. We don’t do Phonegap or Web, just very clear. Cross-industry, ~75 apps over 4 years, all within those parameters.
DockYard - Selling niche services, recommended “Win without pitching” Manifesto. Specializing in a niche creates automatic leverage. Rails consulting gets commoditized, harder to dictate terms.
Cantina - Stay broad, technology changing fast, hire the right folks to keep up with trends. Full service, outcomes-focused.
Who is client-facing at your company?
Cantina - Everyone is client facing. Sales and services “own” accounts, but dev/design are involved and all participate in weekly Friday project team meetings. Each client has an executive sponsor. Delivery group has 5 product manages and 3 sales people (out of 50+)
DockYard - Founder protects Engineers and Designers from clients. Enterprise clients LOVE to have meetings. Lead engineers function as point people in terms of communication.
Terrible Labs - Everyone is client facing, and they have a non-billable PM that acts as “bad cop” when tough situations arise so Engineers don’t have to. Founders try to be involved and avoid “Seagull Style Management” = swoop in, shit on everything, fly away.
Intrepid - 70% of communication goes through PM, and they encourage clients to participate in sprint planning meetings and retrospectives. Clients can even contact engineers and designers directly and can participate in Daily Scrum Standups.