🧠 Workshop Plan: Designing an AI IKEA Agent for Component Guidance

Objective: Improve the current IKEA Kitchen Planner by introducing an AI agent that helps users select the correct kitchen components, reducing drop-off and in-store dependency.

Start Time: 10:00 AM

Duration: ~3 hours

✅ Agenda Overview

📌 Problem Statement

Users abandon the IKEA Kitchen Planner because they are unsure about which components they need (e.g., technical parts, finishes, corner units), and they rely on in-store experts to validate and complete their configuration.

🎯 User Goal

“I want to be sure I’m ordering all the right components so that my kitchen will work and nothing is missing.”

🔍 Key Insights

• Lack of component-level guidance online
• Fear of making a costly mistake
• In-store staff takes responsibility, which builds trust
• Online experience lacks explanation, validation, and clarity

💬 “How Might We” Prompts (10:30–10:40)

Facilitator prepares 3–4 HMWs based on insights. Example:

• How might we guide users through all the necessary components without overwhelming them?
• How might we replicate the reassurance and expertise of in-store staff?
• How might we help users recognize what they’ve forgotten before they check out?

Let participants add their own HMWs on sticky notes.

🗺️ User Journey Mapping (10:40–11:10)

Use IKEA Kitchen Planner screens or a simplified flow to map the journey.

Steps to cover:

1. Choose a kitchen layout
2. Add base units and cabinets
3. Add appliances
4. Add technical/structural components
5. Attempt to validate or finish
6. Drop-off or book an appointment

Ask the team to annotate:

• 💬 Frustration points
• ❓ Unclear decision moments
• 💡 Places where AI could help (label with a lightning bolt ⚡️)

🧠 Brainwriting: AI Ideas (11:10–11:40)

Silent ideation to encourage volume and inclusivity

Instructions:

• Each person writes 3–5 ideas per round (on sticky notes)
• Pass notes to next person for building/improving
• Repeat 2–3 rounds

Focus prompts:

• What kind of AI behaviors could help?
• How might AI provide suggestions, warnings, or validation?
• What should AI say or do to build trust?

🧩 Affinity Mapping (11:40–12:00)

Group the AI ideas into categories such as:

• 🧠 Smart Suggestions (e.g., “Add this piece because you selected X”)
• 🔄 Dynamic Rules/Validation (e.g., “You’re missing a corner bracket”)
• 💬 Conversational UI (e.g., AI chatbot or step-by-step wizard)
• ✅ Confidence Indicators (e.g., green check when config is complete)

Let participants vote on clusters that feel most promising.

📊 Prioritization: Impact vs Effort (12:00–12:20)

Use a 2x2 matrix:

• X: Effort (Low to High)
• Y: Impact (Low to High)

Map stickies into:

• ✅ Quick Wins
• 🚀 Strategic Bets
• 🤔 Maybe Later
• 🧱 Too Complex

Decide on 1–2 winning concepts to explore further.

📈 Strategic Planning: Idea + Value + KPI (12:20–12:45)

For each top idea, fill in this mini-canvas:

✅ Wrap-Up & Next Steps (12:45–13:00)

• Recap the top idea(s) and owners
• Schedule next touchpoint: prototyping/feasibility
• Ask: “What’s one word or feeling you’re leaving with?”