Where is the Emotion in Enterprise Architecture?
Duncan Robson
Enterprise Architect | Data and AI Innovator | Retail Expert | University Lecturer | Distinguished Architect with the Open Group | MSc FBCS CITP
I’ve been working as an Enterprise Architect (EA) for decades, and for many years I’ve had a niggling sense that something fundamental is missing from how we do Enterprise Architecture. Having recently had the time to think about this, I believe I’ve finally found the answer, inspired by Rory Sutherland, Vice Chairman of Ogilvy UK
Rory is an expert in behavioural insights, and generously shares his thoughts via presentations, podcasts, and his brilliant book Alchemy - go search him out. Listening to Rory gave me the lightbulb moment, the piece that is missing from Enterprise Architecture is how do our solutions make people feel?
While we rigorously design solutions around technical requirements (functional and non-functional), we overlook Emotional Requirements, how people perceive the usefulness of the solutions we create. Our solutions may be speedy and resilient, and have great financial savings calculated on a spreadsheet, but if no one uses them because they don't trust them or they make them unhappy, then they going to fail.
"If you look at problems psychologically rather than technologically you might come up with different answers.” - Rory Sutherland.
Interestingly we do often discuss and capture Emotional Requirements at the start of projects, for example using Design Thinking techniques such as Empathy Maps, Journey Maps, Hopes and Fears etc, and we often consider emotions when presenting our ideas, for example using the excellent advice from Jeremy Connell-Waite on how to tell Better Stories.
However by the time we have got to the design stage, these insights have been ‘translated’ in to our scientific, non-contentious and measurable set of technical requirements, and often their meaning has been lost.
I’d like a way to capture Emotional Requirements from the start, and use them throughout our architectural processes in the same way we do with Functional and Non-Functional Requirements. How great would it be if when asked to change to a design to save costs, you could link it back to your Emotional Requirements and say ‘Well we can do that, but you are you going to annoy all of staff due to these reasons, and therefore the project will fail’
Emotional Requirements are also going to be very important when it comes to building and deploying AI solutions. The use of AI in the work place is very emotive, and adoption is often linked to how do people feel about using it, and more fundamentally do they trust it.
领英推荐
So how can we add Emotional Requirements in to our ways of working and frameworks? I don’t have an answer yet, a lot more thinking and discussion with community is needed, however one idea I have had is to use the SCARF model created by Dr. David Rock
The model has five key areas which we could include along with our Functional and Non-Functional Requirements;
What do you think? - I am on to something? If so how do you think we could incorporate Emotional Requirements into our frameworks and processes such as TOGAF?
It would be great to hear your thoughts, and if you would like to follow me on LinkedIn please click here.
ERP Implementation Enterprise and Solution Architect Member of Scottish Tech Army AI Performance Coach
2 周Duncan Robson nice article. For years I've use an adapted version of Barclays personas to help put teams and people into the shoes of those their EA work and project will impact. Whilst this didn't cover indepth feelings it did allow EAs and SAs to perhaps get some insight in how their work and decisions will make people work and therefore feel. Knowing you can never truly understand unless you ask and involve. https://businessdisabilityforum.org.uk/resource/technology-toolkit/barclays-diverse-personas-1/
Senior Program Architect, Advisory Services, at Salesforce.com
1 个月May be some kind of collective “emotional index” of the system users could be a KPI to be measured as part of NFR’s. This is on the assumption that happy users would make happy customers and hence lead to increased sales/ROI etc. My 2 cents worth is probably too simplistic as measuring individual’s emotions would be nuanced. Without being able to baseline and measure the impact of user emotions would not go very well far.