Stakeholder Interview

A stakeholder interview is a short, structured conversation (30-45 minutes) with an internal decision-maker — product manager, content lead, VP, or department head — to understand business goals, organizational constraints, and political realities before starting UX research. In the context of card sorting and IA work, these interviews tell you which parts of the information architecture are actually open for change and which are off-limits.

Key Takeaways

• 30-45 minutes per stakeholder. Shorter is too shallow, longer starts looping
• Interview 4-8 people who own content, control budget, or can block implementation
• The main output isn't requirements — it's constraints. You need to know what can't change before you study what should

Why Stakeholder Interviews Come First

Running a card sort without stakeholder input is a common mistake with a predictable outcome: you get clean data showing users want a different structure, then discover that restructuring is politically impossible because it crosses team boundaries nobody told you about.

Stakeholder interviews prevent that waste. They surface the organizational context that shapes which research findings are actionable. A card sort might reveal that users want to combine "Billing" and "Account Settings" into one section — but if those sections are owned by different teams with separate roadmaps, that finding needs a different implementation strategy than a simple merge.

The interviews also reveal what stakeholders believe about the current IA. Their assumptions are data. When a product manager says "users love our Resources section" but your card sort shows 35% agreement on where Resources content belongs, that gap between belief and evidence becomes a powerful conversation.

Who to Interview

Target people with three types of influence:

• Content owners: People who create, maintain, or approve content in the affected areas
• Budget holders: People who fund the project and will need to approve implementation
• Potential blockers: People whose team or domain will be affected by IA changes

For a website IA project, this usually means 4-8 people: product managers, the content or documentation lead, a marketing director, an engineering lead who'll implement changes, and possibly a support lead whose team fields navigation complaints daily.

Fewer than 4 interviews risks missing a constraint that derails the project later. More than 8 typically produces repetitive themes and delays research kickoff by weeks.

What to Ask

Skip generic questions like "What are your goals?" and get specific:

• "Which sections of the site does your team own, and which do you consider untouchable?" This surfaces sacred cows directly.
• "What do customers complain about regarding navigation or finding content?" Stakeholders who interact with customers have anecdotal evidence that corroborates (or contradicts) research findings.
• "If we could change only one thing about the site structure, what would have the most impact?" Forces prioritization and reveals what keeps them up at night.
• "What constraints should we know about — technical, political, or timeline-related?" Open-ended, but the word "political" gives people permission to name the real blockers.
• "Are any sections being actively restructured or sunsetted?" No point sorting content that's about to be removed.

Record the interviews (with permission) so you can reference exact quotes when presenting findings later. Stakeholders respond better to "You told me in our interview that X" than to an abstracted summary.

A Concrete Example

You're planning a card sort to restructure a B2B SaaS company's marketing site. During stakeholder interviews, you learn:

• The VP of Marketing insists on keeping a standalone "Resources" section (blog, whitepapers, webinars) because it's a lead generation funnel with its own KPIs
• The product team is about to launch a new product line that needs its own top-level section
• Engineering says the CMS can only support 7 top-level nav items without a redesign
• Support reports that 40% of support tickets involve users who can't find pricing information

These constraints reshape your card sort. You include "Resources" as a pre-set category in a hybrid sort instead of letting participants reorganize it freely (since it's not going to move regardless). You add cards for the upcoming product line even though it doesn't exist yet. You design the sort to target 6-7 top-level groupings, knowing that's the CMS ceiling. And you make sure pricing-related cards are well-represented so you can see where users expect to find them.

Without the interviews, you might have run an open sort that produced 12 top-level categories, buried Resources across multiple sections, ignored the new product launch, and generated a structure that's technically impossible to implement.

Turning Interview Data Into Sort Design

After completing all interviews, synthesize the findings into three lists:

1. Fixed constraints: Things that won't change regardless of research findings (team ownership boundaries, CMS limitations, regulatory requirements)
2. Assumptions to test: Things stakeholders believe but haven't validated ("Users prefer our current structure," "Nobody reads the blog")
3. Known pain points: Navigation problems stakeholders have heard about from users or support

Fixed constraints shape the sort structure. Assumptions become hypotheses your sort can confirm or reject. Known pain points tell you which cards to watch most closely in the results.

This synthesis document — one page, three lists — becomes the brief that connects your stakeholder interviews to your card sort design. It also becomes a reference point when presenting results. Stakeholders who see their own words reflected in the research plan are more likely to trust the findings.

Further Reading

• User Persona
• Competitive Analysis
• Card Sorting
• What is Card Sorting?

Frequently Asked Questions

How long should a stakeholder interview take? Plan for 30-45 minutes per stakeholder. Shorter than 30 minutes rarely gets past surface-level answers, and longer than 45 minutes produces diminishing returns as conversations start looping. If a stakeholder has a lot to say, it's better to schedule a focused follow-up than to let a single session run over an hour.

Who should you interview as stakeholders before a card sort? Interview people who own content, control budget, or can block implementation. For a website IA project, this typically includes product managers, content leads, marketing directors, and engineering leads. Aim for 4-8 stakeholders. Interviewing fewer than 4 risks missing a key constraint, while more than 8 usually surfaces repetitive themes and delays the project.

How do stakeholder interviews improve card sorting studies? Stakeholder interviews surface constraints that shape your card sort design. They reveal which content areas are politically untouchable, which sections are actively being restructured, and what business goals the IA needs to support. Without this context, you might run a card sort that produces technically sound results that no one will implement because they conflict with a business requirement you didn't know about.