Synthesis & Iteration

Research without synthesis is just a pile of notes. Synthesis is the process of transforming raw observations into patterns, patterns into insights, and insights into actions. It is where research becomes valuable.

The Synthesis Process

Synthesis typically follows five steps:

1. Download — get everything out of your head and into a shared space
2. Organize — group related observations together
3. Identify patterns — find themes that appear across multiple participants
4. Generate insights — interpret what the patterns mean
5. Prioritize — decide what to act on first

Affinity Mapping

Affinity mapping (also called affinity diagramming) is the most widely used synthesis technique. It works for any qualitative data — interviews, usability tests, diary studies, or open-ended survey responses.

How to Do It

1. Write observations on sticky notes — one observation per note. Include the participant identifier (P1, P2, etc.) so you can trace it back:

   • "P3: Didn't notice the save button because it blended with the background"
   • "P5: Expected auto-save and was confused when changes were lost"
   • "P1: Clicked the X thinking it would save and close, but it discarded changes"

2. Spread all notes on a wall or digital board — tools like Miro, FigJam, or physical sticky notes work well

3. Group silently — each team member moves notes into clusters based on similarity, without talking. This prevents one person from dominating.

4. Name the groups — once clusters form, give each group a descriptive label. "Save/discard confusion" or "Users expect auto-save"

5. Review and refine — discuss the groups as a team. Merge, split, or rename as needed.

The power of affinity mapping is that patterns emerge from the data rather than from your assumptions. You might discover clusters you never anticipated.

Finding Patterns

Not all patterns are equal. Look for these types:

Frequency Patterns

"7 out of 10 participants struggled with the same step." High-frequency issues are the most important to fix. Track how many participants experienced each issue:

Behavioral Patterns

"Users consistently tried X before trying Y." These reveal mental models — how users expect things to work:

• Users searched before browsing categories (they expect search to be primary)
• Users scrolled past the CTA and looked for pricing first (price is a prerequisite for action)
• Users opened the menu looking for account settings (they group settings under their profile)

Emotional Patterns

"Participants expressed frustration at the same moment." Emotions reveal severity:

• Mild confusion at step 2 (cosmetic issue)
• Visible frustration at step 4 (major issue)
• Gave up entirely at step 6 (critical issue)

Contradiction Patterns

"Half the participants wanted A, the other half wanted B." Contradictions often reveal different user segments with different needs. Dig deeper to understand what differentiates the groups.

Writing Insights

An insight is not an observation. Observations describe what happened. Insights interpret why it matters and what to do about it:

Insight format: "[Who] needs/expects [what] because [why], which means we should [implication]."

Example: "First-time users expect the onboarding flow to match the marketing page's promises because that is what set their expectations, which means we should align the first-run experience with the landing page messaging."

Prioritization

You will always find more issues than you can fix. Use a prioritization framework to decide what to tackle first.

Impact vs Effort Matrix

Plot issues on a 2x2 grid:

Severity + Frequency

Combine the severity rating (1-4) with the frequency (how many users affected):

• Critical severity + high frequency = fix immediately
• Major severity + high frequency = fix soon
• Minor severity + low frequency = backlog
• Cosmetic + low frequency = ignore for now

Presenting Findings

Research that lives in a document no one reads is wasted research. Present findings in a way that drives action:

Structure Your Presentation

1. Executive summary — 3-5 key findings in one slide
2. Method overview — who you tested, how, and when (one slide)
3. Key findings — one finding per slide with evidence (video clips, quotes, data)
4. Prioritized recommendations — what to change, in what order
5. Next steps — what research comes next

Use Video Clips

A 30-second clip of a user struggling is more persuasive than any chart. Trim clips to the critical moment and provide context before playing them. Stakeholders who watch a real user fail become advocates for fixing the problem.

Write for Skimmers

Most stakeholders will not read a 20-page report. Make findings scannable:

• Bold the key takeaway in each section
• Use bullet points, not paragraphs
• Include a one-page summary at the top
• Put detailed methodology in an appendix

Closing the Loop

Research is only complete when changes ship and you measure their impact:

1. Share findings — present to the team within one week of completing sessions
2. Track recommendations — create tickets for each recommendation and assign owners
3. Verify fixes — run a follow-up usability test on the redesigned experience
4. Measure impact — use analytics or A/B tests to confirm the improvement

This creates a continuous cycle: research informs design, design ships, data measures impact, and new questions emerge for the next round of research. The cycle never truly ends — and that is the point.