Category Archives: Software Development

Architect Driven Development

Architect Driven Development, or ADD (just add the H in the appropriate place for most of us), is my new methodology of choice.  In the large Enterprise context, where Agile is difficult to implement across many teams, ADD hits the gap by positioning the Architect to enforce the development methodology as well as the software architecture.  Let’s face it, as Architects we end up being the people that, at least at first, are ensuring the development methodology is followed.

Why ADD?

In the Enterprise context where there are numerous teams spread across many mini-empires, one of the few roles that can be consistent is the Architect.  In my role at Best Buy, I run a 70 person development team.  That team is broken into eight different workstreams all lead by top tier architects.  As long as the architect team is aligned on both a methodology and high level architecture goals, we get consistent results from our teams.  It helps a lot that we all believe in Agile self-managed teams as it gives the developers the freedom to do the work how the team sees fit.  But the architects carry the consistency through the development process from start to finish.

There are also a couple hundred developers working on various parts of that are not under architect control yet.  These teams are largely driven by outsourcers who have a vested interest in keeping Best Buy architects away from their work.  If we were there, they wouldn’t get to staff their $200/hr architect on the team or their 10 offshore counterparts.  The problem here is the outsourced architect is not on board with our development methodology or high level goals, and given our structure there’s little incentive for them to even care.

So as we strategized how to get better control over the dotcom platform, I put forth that as long as we controlled the architects and the process, nothing else mattered.  Inherent in that control is control of resources as well.  Since our hiring standards are tough, we only get top notch developers which makes everything else easier.  But without a strong architect voice to push a methodology, pick resources and keep the architecture vision, the process would fail.

Next time you are thinking about restructuring your development environment, stop thinking about what methodology you’re PMO is going to use and start thinking about getting strong technical architects to carry the vision and methodology.  You’ll find that the rest of the process falls into place afterwards.

Think, Do, and let the Doers Think Too

I just did my first video interview about what is trying to build and how we are trying to build it.  It was fun but I came out of it thinking about the overall strategy Best Buy has applied toward technology for the last 10 years.  In a nutshell, and this is certainly not new information, Best Buy has decided that thinking is the most important part of technology strategy.  Therefore, we only staff managers who think about business problems, and farm out the doing to the cheapest vendor we can find.  If you ever try to do anything technology related at Best Buy, you have to have a big IT vendor do it for you.

The problem with this attitude is that no value is placed on the technical solution. When technology is driving the base strategies of so many eCommerce companies, it is hard to understand why we don’t place value on technology.  Last year alone we taught a number of our vendors how to build elastic systems in the Cloud.  When our projects were over, all those people we taught went off and built elastic systems in Clouds for other companies, likely our competitors as well.

Not only Think.  Not only Do.  Think and Do.

If you only staff Think you are reliant on someone else to Do.  You are giving away you’re strategies and technologies.  Most importantly, you’re giving away your minds that have learned new technologies and new strategies.

If you only staff Do you may pursue interesting technological pursuits, but their business value may be suspect.  From a tech person standpoint, just staffing Do is very seductive.  I’ve been on plenty of teams where the implementing team knew the business requirements better than the business.

Think and Do.  Tech and business teams highly coupled and dependent on each other can produce amazing results.  One of my favorite projects was with a highly engaged business customer who took a high functioning Agile team in directions we never would have expected.  The results were superior to anything the tech team could imagine.

But the best of the best comes from Think, Do, and let the Doers think too.  This is the core of the Agile process where decisions are distributed to the people that know the system best.

Think, Do, and let the Doers think too.


What’s Wrong with the Twin Cities’ Developer Career Path

I’ve hired over 100 developers and architects throughout the years.  I’ve worked at around 20 companies, some very briefly, as both an employee and consultant.  I like fast development teams that ride the fine line between excellence and chaos.  But I’ve recently come to understand that a lot of my attitudes around being a developer in the Twin Cities are vastly different than those of our West Coast brethren.  This post is full of vast generalizations so your particular situation may not apply, bear with.

Twin Cities Career Path

As a software developer in the Twin Cities, the general path is to get your foot in the door anywhere that will hire you.  Small companies, large companies, whoever will pay you to learn enterprise scale, team oriented development.  Honestly, your first 5 years as a developer probably cost the company you work for more money to have you there than your worth in business value.  Even so, you’ll add the most value at a small Agile shop where you get to do everything from development to deployment to maintaining the Cloud infrastructure.  Your learning experience at these places will set you up for your next 10 years of work.

Career Choice

Anywhere after your 5 year mark you get the choice to become a career developer.  Around here is where all the people that can’t really cut it become PMs or BAs or some other manager.  This is a fine choice as the life of development can be tough.  But the choice for developers is to continue to work as a cog in the corporate machine or to become an independent or sponsored (employee at a body shop) consultant.

Employee Route

When you go the employee route you get perceived stability in exchange for little control of your own destiny and work.  Due to the nature of financial funding of development projects, employees are often relegated to maintaining existing systems or as the token employee or two on the new development projects staffed by contractors.  You’re only on these projects because they want someone to maintain the system after all the consultants leave.

Why is this the case?

Budgets at corporations split development into two buckets, capital and expense.  Projects that can be capitalized allow the company to depreciate the cost of the development which saves them money on taxes.  Projects that are expense go right to the bottom line of the budget and are a drain on that division’s profitability. From a bean counter’s perspective, adding flexible staff to write new development projects makes the most sense as the cost of consultants is higher than the cost of full time staff.  That higher cost then gets depreciated and the ongoing maintenance expense goes to cheaper full time or offshore resources.

Why does this limits the career path in Minneapolis?

What this means to you as a developer is that, since your value is lower to the company as a pure expense, the amount company’s are willing to pay you is limited.  Thus the salary of a full time developer in the Twin Cities is capped not far north of $100,000.

Consultant Route

As a consultant, you can make significantly more with a similar amount of experience.  The rates vary greatly depending on your skill set and experience level but rates to the consultant between $80 and $120 per hour are fairly common.  Also, when a company pays this much for development resources, you are more likely to land a new green field project as the project is capitalized (and can be depreciated).

Career Results

Given the general climate of Twin Cities development, the more aggressive and generally higher quality developers go into consulting once they realize the limitations of working in the Twin Cities.  But, while the pay is better the career path is over at this point and one risks being a developer hopping from contract to contract for one’s whole career or until your skillset is no longer valued.

West Coast Difference

Having not lived there, I’m now going on conversations with the west coast engineers that I know.  In general, they’ve described the opposite situation as exists in the Twin Cities, where the vast majority of developers stay full time employees and the number of contractors is limited.  Also, that the best development talent is always scooped up by the numerous tech companies in the area at high salary plus stock options.  Plus, the culture of the startup breaking out is always there and quite tangible.  The latest story being InstagramThe Instagram writeup in the New York Times drives home the point of what is wrong with the Twin Cities.  This excerpt sums it up:

The extraordinary success of Instagram is a tale about the culture of the Bay Area tech scene, driven by a tightly woven web of entrepreneurs and investors who nurture one another’s projects with money, advice and introductions to the right people. By and large, it is a network of young men, many who attended Stanford and had the attention of the world’s biggest venture capitalists before they even left campus.

What is wrong with the culture of the Twin Cities tech scene?  These structures just plain do not exist here and that is the main culprit behind the exodus of top talent in the Twin Cities to mercenarial endeavors.  With all those smart minds going to short term financial reward, there will be no culture of risk taking, the financial availability won’t appear and the notion that ideas for the next billion dollar startup are ubiquitous will never materialize.

How we can start this in the Twin Cities will be difficult.  It involves getting the venture capitalist and angel’s around here to start taking chances on smart college kids and catching them before they hit the Twin Cities career path.  Once they are in, our culture will take over and they’ll be consultants before you know it.  And the next generation of risk takers will be lost.