One of the great joys of my career is getting to talk to architects, all over the world. I am absolutely sure that I have spoken to more architects than anyone else in the world. I don’t mean that as hyperbole. Hundreds of interviews, organization assessments, conferences Iasa holds, communities we start, online, in person, in big companies and small. Vendors, banks, insurance, retail, government. Peru, Australia, Malaysia, Colorado, Seattle, New York, London, Spain, Netherlands and just so many more. And I hope I don’t ever have to stop.
But I’ve noticed something over the years. We don’t seem to be learning anything as a field.
I was on three calls last week that, in my opinion, simply shouldn’t exist. In one, the leader of a large team in a Fortune 500 wanted to stabilize the ways his architects “describe and design architectures in a repeatable way”. I actually have had three of those in the last month (big orgs who are frustrated that the titled architects all describe their systems radically differently). In another, the quote was, “we don’t really know what decisions we are making, or if we are making them well, the outages suggest we are not.” And in another, “We would really just like to know how to spend our time on the ‘right’ things, you know, the valuable things”. These are in no way unique.
I have begun calling this Professional Dissociative Amnesia… Which roughly means our profession seems to experience a kind of collective social amnesia every three to five years or so and we simply start everything over again with slightly changed titles and a different set of thought leader books which end up saying the same thing. Part of the dissociation comes from the stresses architecture practices are under.
To wit… here is the list of outcomes we have been helping organizations with for the past 25 years.
- Stable and standardized design and development viewpoints. Not constricting in the sense of an inability to be flexible, but using common methods to describe architectures.
- Implementation of scoped decision improvements, especially around quality, velocity, and traceability.
- Becoming trusted advisors to both business and technical leadership while establishing a long-lasting source of authority for the architecture practice itself. (more on this soon)
- Assessing whether the architects have the skills of architects or are simply experts.
- Improved involvement in value-based outcomes. Effectively being involved with the most valuable projects.
- Reduction of risk and cost for technology initiatives.
- Driving innovation.
- Improving governance outcomes in the governance and review cycle.
For many of these, companies have had to come to us two or three times as their leadership and teams change. A new CTO, let’s revamp the whole thing! A change in the definition of modern… let’s flip it. Each hype cycle comes with a complete renewal of the definitions of architecture.
The challenge seems to come not from any factual differences. Most of these practices (what we call a group of architects within a single org) end up learning the same lessons, but in the WAY our profession ignores stored learning in favor of reinvention.
Case in point, I recently spoke at a conference where decision records were one of the top highlights of the conference on ‘modern architecture’. Now don’t get me wrong, I love decision records as tools. That is why they have been in the BTABoK and Iasa training for over 17 years. I first spoke on decision records around 22 years ago. And here is the thing: we measured 200 companies recently, and they still hadn’t improved in these techniques. Why? Why, 22 years later, are we presenting on these methods like they are brand new?
This is the heart of the collective amnesia. We seem to fixate on reinvention rather than retention.
Because there has never been a collection point that was not behind a vendor’s paywall. And never one that architects could impact for themselves. Or didn’t belong to someone’s marketing budget. Microsoft talked to me about building a knowledge base for them around 2010. Amazon built Well-Architected around the same time.
So I set out to fix this.
And the root of the fix is simply a location to store our collective discoveries and shepherd them together. You see, I don’t see the BTABoK as a framework, but instead a repository. A store for all of the techniques that worked, that we have heard from you through the years.
https://www.iasaglobal.org/browse/btabok/3.2
It is also a way to call bullshit. As I do on so many of today’s ‘innovations’. Event driven architecture is in NO way new. One of my friends wrote a book on it in 2005. DDD is also not new. I interviewed Eric in the mid 2000’s as well.
They are both awesome; they are both in the BTABoK because they are awesome. But so are viewpoints. ISO did work on that a LONG time ago. So are patterns.
So let’s come back to the goals:
Stable and standardized design and development viewpoints. Not constricting in the sense of inability to be flexible but using common methods to describe architectures.
This exists in the BTABoK. 16 viewpoints, roughly 100 concerns, 4 supported model languages (ArchiMate, UML, c4, custom)
Implementation of scoped decision improvements especially around quality, velocity and traceability.
This has existed for a LONG time. Decision records, decision cascades, decision biases, stakeholder management, and value connection at multiple scopes and with templates for all of them. All free.
And if you want to improve our methods you simply have to participate with a group of contributing architects.
Becoming trusted advisors to both business and technical leadership while establishing a long-lasting source of authority for the architecture practice itself. (more on this soon).
The methods have been fully established for both value accrual (buy in and support) as well as stakeholder management. These techniques have come from dozens of practices that have made them work. All fully documented and implementable.
Asessing whether the architects have the skills of architects or are simply experts.
The career path, job descriptions, competency model and specializations (types of architects) have not only been around and supported but are fully documented and have both training and certification support.
Improved involvement in value based outcomes. Effectively being involved with the most valuable projects.
Value management is both fully established and well understood in the BTABoK to the extent that professionals can do so. Benefits, measures, business cases, objectives, intent and capability all exist.
Reduction of risk and cost for technology initiatives.
Risk management and cost management exist both as procedural outcomes but also as methods.
Includes signficant understanding and techniques for technical debt management.
Driving innovation.
If anything our techniques are too innovative for modern architecture practices which currently veer towards governance and review.
Driving structural and quality attribute excellence.
From ATAM and PBAAM architecture assessments to quality attribute measures and alignment with both process and design tools.
A full viewpoint management technique for large practices, including systems architecture for product and engineering companies (gas, oil, defence etc)
Improving governance outcomes in the governance and review cycle.
A structured participation in overarching governance method exists and is implementable today.
So I know sometimes I sound frustrated. Heck sometimes I sound Angry… but that is only because Ive had this conversation before. Sometimes Ive had it 50 times, sometimes 100.
These answers are not only established, they are old. They are here. They are safe. I realize the thought leaders and vendors love to sell you something new. But can we start with getting the basics behind us?
We have been arguing the same thing for decades. And if we dont get a memory, the world may just pass us by.
Come get your memory here: https://www.iasaglobal.org/