Principles of Product Management: Personas (Proto‑personas, Behavioral Personas, Jobs‑to‑be‑Done Personas)

Personas (Proto‑personas, Behavioral Personas, Jobs‑to‑be‑Done Personas)

Personas (Proto-Personas, Behavioral Personas, Jobs-to-be-Done Personas) – Study Guide

What This Is

Personas are fictional but data-backed representations of your target users, designed to align teams on who you’re building for and why. They prevent the "elastic user" problem (where everyone assumes the user is just like them) and ensure product decisions solve real problems. Example: When Monzo redesigned its onboarding flow, it used behavioral personas to segment users into "Savers" (who cared about budgeting tools) and "Spenders" (who wanted instant card access), leading to a 20% increase in activation.

Key Terms & Frameworks

• Proto-Persona: A hypothesized persona based on internal assumptions (e.g., "Busy parents who need quick grocery delivery"). Used early in discovery before research.
• Formula: Proto-Persona = Demographic Guess + Pain Point Assumption + Hypothetical Goal

• Example: "Urban millennials (25–35) who struggle to find time to cook and want healthy meals delivered in <15 mins."

• Behavioral Persona: A persona defined by actions (e.g., "Power Users who log in daily" vs. "Churn Risks who drop off after 3 days"). Focuses on how users interact with your product.

• Framework: RFM Model (Recency, Frequency, Monetary) – Segment users by behavior, not just demographics.

• Example: Duolingo’s "Loyal Learners" (daily active) vs. "Weekend Warriors" (sporadic usage).

• Jobs-to-be-Done (JTBD) Persona: A persona defined by the job they’re "hiring" your product to do (e.g., "I need to feel prepared for my presentation" vs. "I need PowerPoint").

• Framework: JTBD Statement = When [situation], I want to [motivation], so I can [outcome].

• Example: Slack’s "Overwhelmed Manager" persona: "When my team is remote, I want to reduce email clutter, so I can focus on high-priority work."

• Demographic Persona: A persona based on who the user is (age, income, location). ⚠️ Weak alone—pair with behavioral or JTBD data.

• Example: "Suburban moms, 30–45, $75k+ household income" (too broad without why they use your product).

• Empathy Map: A tool to visualize a persona’s thoughts, feelings, pains, and gains in a 4-quadrant grid (Says, Thinks, Does, Feels).

• Steps:

  
  
  1. Fill in quotes from user interviews.
  2. Note unspoken frustrations (e.g., "I hate when apps ask for my phone number").
  3. Map actions (e.g., "Abandons cart if shipping >$5").
  4. Capture emotions (e.g., "Anxious about overspending").

• Persona Spectrum: A range of personas from narrow (e.g., "Uber Eats drivers in NYC") to broad (e.g., "Gig workers"). Helps prioritize which segments to focus on first.

• ICE Score for Personas: Prioritize personas using Impact × Confidence × Ease (same as feature prioritization, but applied to user segments).

• Example: If "Small Business Owners" score higher than "Enterprise Clients," focus on them first.

• Segmentation Matrix: A 2x2 grid to compare personas by behavior (e.g., high/low engagement) and value (e.g., high/low revenue).

• Example: Airbnb’s matrix:

  
  
  • High Engagement + High Value = "Frequent Travelers" (prioritize).
  • Low Engagement + Low Value = "One-Time Bookers" (deprioritize).

• Persona Validation: The process of testing if your persona matches real user behavior. Use qualitative (interviews) + quantitative (analytics) data.

• Metric: Persona Accuracy = % of users in segment who exhibit predicted behaviors (aim for >70%).

Step-by-Step / Process Flow

1. Start with Proto-Personas (Hypotheses)
2. Gather internal stakeholders (design, engineering, marketing) for a 30-minute brainstorm.
3. Ask: "Who do we think our users are? What problems do they have?"

4. Output: 2–3 proto-personas (e.g., "College students who need cheap textbooks").

5. Validate with Research (Kill or Confirm Hypotheses)

6. Qualitative: Conduct 5–10 user interviews per proto-persona. Ask:
   • "Walk me through the last time you [used a product like ours]."
   • "What’s the hardest part about [job they’re hiring for]?"
7. Quantitative: Analyze behavioral data (e.g., Google Analytics, Mixpanel) to spot patterns.
   • Example: If 80% of "college students" abandon checkout at the payment step, their pain point is likely payment flexibility, not price.

8. Output: Revised personas with real pain points and behaviors.

9. Map Personas to Jobs-to-be-Done (JTBD)

10. For each persona, write a JTBD statement.
    • Example: "When I’m studying for finals, I want to find affordable textbooks fast, so I can save money and time."

11. Tool: Use a JTBD Canvas (situation, motivation, outcome, alternatives, anxieties).

12. Create Behavioral Segments

13. Use analytics to group users by actions (e.g., "Users who add to cart but don’t check out").

14. Example: Spotify’s "Discover Weekly" users vs. "Repeat Listeners" (different engagement patterns).

15. Prioritize Personas Using ICE

16. Score each persona on:
    • Impact: How much does solving their problem grow the business? (1–10)
    • Confidence: How sure are we this is a real problem? (1–10)
    • Ease: How easy is it to reach and serve them? (1–10)

17. Example: A "Power User" persona might score 9/10/7, while a "Churn Risk" scores 7/8/5.

18. Socialize Personas with the Team

19. Create a 1-page persona doc (name, photo, JTBD, pain points, behaviors, quotes).
20. Pro tip: Give personas names (e.g., "Budgeting Brenda") to make them memorable.
21. Output: Team alignment on "Who are we building for this quarter?"

Common Mistakes

• Mistake: Creating personas based only on demographics (e.g., "Women 25–34").

• Correction: Pair demographics with behaviors or JTBD. A 25-year-old woman could be a "Power User" or a "Churn Risk"—same demo, different needs.

• Mistake: Assuming one persona = one user.

• Correction: Users can switch personas (e.g., a "Casual Shopper" becomes a "Loyal Customer" after a great experience). Use persona spectrums to account for this.

• Mistake: Building personas once and never updating them.

• Correction: Revisit personas every 6–12 months. Use cohort analysis to track if behaviors change (e.g., post-pandemic, "Remote Workers" became a new segment).

• Mistake: Ignoring "negative personas" (who aren’t your users).

• Correction: Define anti-personas (e.g., "Enterprise clients" if you’re a B2C app). This prevents scope creep.

• Mistake: Making personas too broad (e.g., "Everyone who uses the internet").

• Correction: Use the 80/20 rule—focus on the 20% of users driving 80% of value.

PM Interview / Practical Insights

1. Tricky Distinction: Proto-Persona vs. Validated Persona
2. Interviewer Trap: "How do you know your persona is accurate?"

3. Answer: Proto-personas are hypotheses; validated personas are backed by data. Always say, "We’d test this with [interviews/analytics] before committing."

4. Stakeholder Pushback: "Why not just build for everyone?"

5. Answer: Use the segmentation matrix to show that "everyone" = no one. Example: "If we optimize for both ‘Power Users’ and ‘Churn Risks,’ we’ll dilute the experience for both."

6. JTBD vs. Behavioral Personas

7. Interviewer Trap: "Should we use JTBD or behavioral personas?"

8. Answer: Both. JTBD explains why users "hire" your product; behavioral personas explain how they use it. Example: "Our JTBD is ‘I need to feel organized,’ but our behavioral persona shows they use the app at 9 PM on Sundays."

9. Prioritization Question: "Which persona should we focus on first?"

10. Answer: Use ICE scoring and tie it to business goals. Example: "If our goal is retention, we’d prioritize ‘At-Risk Users’ over ‘New Users.’"

Quick Check Questions

1. Scenario: Your team wants to build a feature for "Busy Parents," but your data shows that "College Students" have higher engagement. How do you decide?

2. Answer: Prioritize "College Students" if they drive more value (e.g., revenue, retention). Use ICE scoring to compare impact vs. effort for each segment.

3. Scenario: A stakeholder says, "Our persona is ‘Small Business Owners,’ but we don’t have any data on them." What’s your next step?

4. Answer: Create a proto-persona and validate it with 5–10 interviews. Avoid building on assumptions.

5. Scenario: Your analytics show that 30% of users abandon onboarding at the "Add Payment" step. How do you update your personas?

6. Answer: Add a behavioral segment: "Payment-Averse Users" with pain points around trust/security. Then test solutions (e.g., guest checkout, multiple payment options).

Last-Minute Cram Sheet

1. Proto-Persona = Hypothesis (demographics + pain points + goals). Always validate.
2. Behavioral Persona = Defined by actions (e.g., "Daily Active Users"). Use RFM model.
3. JTBD Persona = Defined by job (e.g., "I need to feel prepared"). Use When [situation], I want to [motivation], so I can [outcome].
4. Empathy Map = 4 quadrants: Says, Thinks, Does, Feels.
5. ICE for Personas = Impact × Confidence × Ease. Prioritize high-scoring segments.
6. Segmentation Matrix = 2x2 grid (behavior vs. value). Focus on top-right quadrant.
7. Persona Validation = Qual (interviews) + Quant (analytics). Aim for >70% accuracy.
8. ⚠️ Avoid: Demographic-only personas. Pair with behaviors/JTBD.
9. ⚠️ Trap: Assuming personas are static. Revisit every 6–12 months.
10. Anti-Persona = Who isn’t your user (e.g., "Enterprise clients" for a B2C app).