What it is
Strategic objectives speak the language of business; systems speak the language of architecture characteristics, the famous ilities: availability, performance, resilience, scalability, security and friends. The second AARM stage translates one into the other: for each objective confirmed in the previous stage, which characteristics must the platform have to enable it?
The session’s output is a prioritized list with the top 3 characteristics (out of a maximum of 7), plus the implicit characteristics every serious system carries. This prioritization is what turns "we want quality" into objective criteria for choosing patterns, technologies and where to invest the team’s effort.
"You can never escape one danger without running into another. Prudence consists in recognizing the different dangers and accepting the least bad as good." Niccolò Machiavelli, The Art of War. Prioritizing characteristics is, at heart, determining trade-offs.
How to facilitate the session
-
Derive candidate characteristics from each objective
Walk through the 3–5 strategic objectives and ask: what does the platform need to be to enable this? Growing market share tends to call for scalability; reducing error rates calls for testability.
-
Make the implicit characteristics explicit
Feasibility, security, testability and maintainability rarely show up on the business’s list, and they are not optional for that. Record them on a separate track so they don’t compete for space with the priorities.
-
Prioritize the top 3 (7 at most)
The hierarchy matters more than the list. The first three characteristics win any technical tie from then on; from the eighth onwards, it’s noise.
-
Determine the trade-offs
Every prioritized characteristic charges a price on another: consistency versus availability, simplicity versus flexibility. Name what is being sacrificed: deciding without naming the cost is not deciding.
Characteristics catalog
The starting menu for the session. No architecture pattern maximizes them all. As Mark Richards’ classic table shows, each pattern scores differently on each characteristic. The answer to "which pattern is best?" remains: it depends on your top 3.
AARM Toolkit
The complete kit of fillable templates for the 4 stages: strategy canvas, ilities matrix, risk-storming board, architectural story, ADR and backlog.
Common mistakes
- Prioritizing everything. The Swedish warship Vasa sank on its maiden voyage by trying to be the most armed, most ornate and most imposing all at once. An architecture that promises every ility delivers the same fate.
- Forgetting the implicit ones. Security and testability don’t show up in the business pitch, but their absence shows up in the news. Separate track, always visible.
- Choosing the technology before the characteristics. Deciding "we’re going microservices" before prioritizing the ilities inverts the method: technology becomes dogma instead of consequence.