Guest Blog from Jan Labrie: Winning Architecture

Pensioenregister

The Dutch National Architecture Forum (NAF) has a tradition of handing out annual awards. The 2011 project award was won by mijnpensioenoverzicht.nl, a site that gives a combined insight into all your pension information.

Being one of the architects this award directly confronted me with the issue of how the jury would select the winning architecture from all the contestants. Or in other words how can you compare architectures?

The jury took a clever approach and did not value the architecture in isolation. Instead they assessed the effects of the architecture. In particular the following aspects were taken into account.
Was the project successful?
Was the project realized as specified in the architecture?
Was the architecture used in the communication to management?

When the end result is right, the underlying architecture should be right, right?

This indirect approach is defendable, but wouldn’t it be possible to find characteristics of the architecture itself; characteristics that express the inherent quality of the enterprise architecture. This could be aspects like the degree to which the architecture addresses the business requirements. But also quality metrics on the data and the application architecture could be useful. In this area we might be able to reuse metrics from the research on software quality.

Things like size, complexity, dependency structures and repetition are metrics that could be useable to express the quality of architectures. It would be used in the context of formally defined viewpoints, mostly diagrams. Eventually they will become standard functionality of architecture design tools, defined for Archimate and part of the architecture deliverables.

So far for the future thoughts, back to reality

The architecture of mijnpensioenoverzicht.nl is award winning because it turned out to be simple, concise and based on principles that worked well for all involved parties. Enterprise architecture at its best!

Jan Labrie

Ronald Streekstra

Enterprise Architect. Dat zou genoeg kunnen zijn voor een introductie, maar dat voelt als te weinig. Enterprise Architect met een achtergrond in Java/J2EE en RUP. Dat voelt al beter maar is misschien wat te technisch. Er is toch meer te vertellen? Hoe ver ga je met een beschrijving. Dat heeft allemaal te maken met balans. Daar ligt voor mij de uitdaging in enterprise architectuur. Want ja, we kunnen de hele wereld beschrijven (zie Zachman of TOGAF), maar krijgen we dan een architectuur die werkbaar is en producten die bruikbaar zijn. Begrijp me goed, het raamwerk of de methode is niet fout, het gaat erom wat je met het raamwerk doet en hoe je de methode uitvoert. Mijn speciale aandacht in architectuur gaat uit naar het realiseren van samenhang. Dat kan samenhang in het IT-landschap zijn maar ook samenhang in hoe mensen met IT werken of hoe ze onderling werken. Enterprise architect met balans en een dosis pragmatisme. Dat ben ik en daar houd ik het maar bij.

2 comments

  1. Jan Labrie says:

    One of the applied principles was to collect all pension information real-time from the responsible parties. This in comparision to a solution where all data is collected and stored centrally.

    This worked really well because it made that the information is always as actual as possible. And maybe even more importantly it made that the connected pension companies did not have to ‘give away’ their valuable customer information. This increased the acceptance and cooperation level of all involved parties.

  2. Ronald Streekstra says:

    I’m always looking for principles that work, because sometimes principles backfire at you. Can you provide an example of a good one ?

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>