Analysis and synthesis

While rereading a book on information architecture, "synthesis" was listed as the counterpart to "analysis."

I'd never thought about the antonym of analysis, so it sparked interest — and one question came up.

Analysis techniques are celebrated in business books and many people care about that skill — so why do so few people talk about techniques for synthesis, and why is interest so low?

I've written a blog summarizing the essence of problem solving before. The essence of problem solving — already four years ago.

Maybe synthesis draws little interest because analysis results inductively lead to a conclusion on their own, so there's tacit knowledge that synthesis doesn't especially need technique.
But if analysis needs technique, it's natural that synthesis needs technique too.

Probably the real reason it goes unmentioned is that what people generally call "analysis" often already includes synthesis.

People often don't draw a line for where analysis ends and synthesis begins.

Strictly, though, the analysis phase aims to "break elements down to understand them," and the synthesis phase aims to "combine elements and converge toward a decision."

To make it clearer, here's a use case.

====================

Use case: Decide whether the company should enter a new field as a new business

===================

In the analysis phase, quantitative or qualitative:
・Research market size and growth rate
・Map competitors and each company's positioning
・Identify needs and issues by customer segment
・Inventory company resources (tech, people, capital, customer base)
・Estimate barriers to entry and required investment
and so on

Needed skills include information gathering, designing how to slice things apart, quantification, and so on.
The deliverable is the analysis result — a "bundle of organized information."

In the synthesis phase, you "decide based on that bundle from analysis."

What you do:
・Recognize contradictions between analysis results (attractive market but strong competitors, etc.)
・Ask alignment with strategic intent (why this domain?)
・Draw multiple entry scenarios (where to attack from)
・Compare risks and returns of each scenario even when they're not really comparable
・Make a final call including exit conditions

Skills needed:

・Setting priorities (what matters most — clarifying values)
・Accepting uncertainty and setting hypotheses (handling unknowns as unknowns)
・Naming trade-offs (taking A means losing B — make it explicit)
・Designing the time axis (by when should what be visible?)
・Storytelling (explain why this decision in a consistent form)

So the boundary of analysis/synthesis has a lot of room for interpretation and value scales. The common business phrase "from here it's just a matter of deciding" belongs to the synthesis phase.
(Interpretation and value judgment can enter analysis too, but clearly more so in synthesis.)

Again: when people discuss analysis methods, they often mean "analysis methods ≈ problem-solving methods," so synthesis is often included as part of problem solving.

Still, when you analyze or solve problems yourself, separating analysis and synthesis helps you notice "from here on is interpretation and value judgment" — and that should raise how convinced you feel about the final decision.

A bit of a dull blog this time — that's all.
To make it even duller I could write a slapdash business piece about how AI got good at synthesis too so humans are in deep trouble — but that'd be in poor taste, so I'll stop.