top of page

Anjanesh Sharma

BEHAVIORAL DIARY STUDY OF TRADING HABITS

Project Overview

Role

Lead UX Researcher

Timeline

6 week

Methodology
Longitudinal Diary Study, Contextual Interviews, Behavioral Pattern Analysis

DOMAIN
Fintech / Trading Behavior

Business Context

  • A trading platform wanted to understand the behavioral patterns and emotional triggers that influenced trading decisions among retail investors, particularly during periods of market volatility.

  • Quantitative analytics showed what users did—trade frequency, win/loss ratios, portfolio composition—but couldn't explain why users made certain decisions, what emotional states preceded trading activity, or how external factors (news, social media, peer conversations) influenced behavior.

  • The product team was designing a suite of "smart nudges" and decision-support features to help users make more informed trading decisions. But without understanding the real-world context in which trading decisions happened, these features risked being tone-deaf or mistimed.

  • The strategic question: What are the environmental, emotional, and social triggers that precede trading decisions and when are users most receptive to decision-support interventions?

The Challenge

Trading behavior doesn't happen in isolation. It's influenced by:

  • Emotional states (fear, greed, confidence, panic)

  • External triggers (market news, social media, WhatsApp groups)

  • Temporal patterns (time of day, day of week, market conditions)

  • Social influence (friends' advice, influencer recommendations)

Traditional methods like surveys or one-time interviews capture retrospective rationalization, not real-time decision context. I needed a method that could capture behavior as it unfolded, in the moments when decisions were actually being made.

Research Approach & Thought Process

I designed a 4-week longitudinal diary study with a hybrid analog-digital capture method:

Study Design:

Sample: 18 participants across three behavioral segments:

  • Impulsive Traders (6 participants): High frequency, emotional decision-making

  • Analytical Traders (6 participants): Research-driven, patient, strategic

  • Social Traders (6 participants): Heavily influenced by tips, groups, influencers

Duration: 4 weeks of active diary logging

Capture Method:

  • Mobile diary prompts (via WhatsApp bot): 3x daily check-ins—morning (market open), midday, evening (market close)

  • Event-triggered logging: Participants instructed to log immediately after making a trade, seeing relevant news, or receiving trading advice

  • Weekly video diaries: 10-minute reflection on the week's trading decisions, captured via smartphone

Prompts Included:

  • "What triggered you to open the trading app right now?"

  • "What are you feeling about the market today? (scale + open text)"

  • "Did you make any trades today? If yes, what led to that decision?"

  • "Did you see any news/tips/advice today that influenced your thinking?"

Rationale: The three-times-daily cadence captured routine patterns. Event-triggered logging captured decision moments in real-time. Weekly video diaries allowed for reflection and synthesis.

Post-Study: Conducted 60-minute contextual interviews with all 18 participants, walking through their diary entries to probe deeper on key moments.

Key Findings

Finding 1: Trading Decisions Cluster Around Three Emotional Triggers

Analysis of 1,200+ diary entries revealed that trading activity (not just app opens, but actual buy/sell execution) clustered around three emotional trigger patterns:

1. FOMO (Fear of Missing Out) - 38% of trades
Trigger: Seeing a stock price surge or peer sharing gains
Emotional state: Anxiety, urgency, regret
Timing: Often within 15 minutes of trigger
Quote: "My friend sent a screenshot of his portfolio up 12% in one day on this penny stock. I felt like I was being left behind. I bought in without even checking the company."

2. Revenge Recovery - 29% of trades
Trigger: Recent loss
Emotional state: Frustration, determination to "win it back"
Timing: Same day or next morning after a loss
Quote: "I lost 8,000 on a bad call. I couldn't sleep. Next morning I woke up angry and determined to recover it. I took a much bigger position than I should have."

3. Validation Seeking - 21% of trades
Trigger: Market uncertainty, conflicting information
Emotional state: Doubt, need for reassurance
Timing: After consuming contradictory advice from multiple sources
Quote: "I saw three different YouTubers saying opposite things about the same stock. I was so confused. I ended up just following the one with the most subscribers because I needed someone to trust."

Insight: These triggers are behaviorally predictable. FOMO happens in response to social proof. Revenge Recovery follows losses. Validation Seeking happens when information overload creates paralysis. These are intervention moments.

Finding 2: Morning Routines Create Decision Anchors

Video diary analysis revealed that participants had highly ritualized morning routines that set the emotional and informational context for the entire trading day:

  • 74% of participants checked the same 3-5 information sources every morning in the same order

  • Common sequence: WhatsApp trading groups → Stock screener → Business news app → YouTube market analysis

  • Behavioral pattern: The first strong opinion encountered in this routine became the decision anchor for the day

One participant described it: "I have my routine. I check my WhatsApp group first thing. If the vibe is bullish, I'm in a buying mood all day. If everyone's scared, I don't trade. It's like the group sets my mood for the market."

Insight: The first 30 minutes after market open are the highest-leverage intervention window. Users are forming their "market thesis" for the day based on whatever information they encounter first. Decision-support features need to be present in that window, not later when anchors are already set.

Finding 3: Peer Influence Overrides Platform Recommendations

Diary entries showed that when participants received conflicting information platform recommendation vs. peer advice peer advice won 68% of the time, even when participants acknowledged the platform's recommendation was "probably smarter."

Example from diary entry: "The app gave me a 'high risk' warning on this stock. My analysis also showed it was overvalued. But my brother-in-law made 40% on it last month and told me to get in. I ignored the app. Family knows me better, right?"

Insight: Trust is social, not algorithmic. Users trust people they know over systems they don't understand. Decision-support features that try to override social influence will fail. Instead, they need to integrate social context e.g., "3 of your contacts also researched this stock. Here's what they looked at."

Finding 4: Evening Reflection ≠ Morning Action

A striking pattern emerged when comparing evening video diaries (reflective) with next-morning behavior (active):

  • Evening diaries: Participants expressed regret, acknowledged emotional trading, committed to being "more disciplined tomorrow"

  • Next-morning behavior: 81% repeated the same behavioral patterns within 48 hours

Quote from video diary: "I know I need to stop revenge trading. I always lose more when I do that. Tomorrow I'm going to stick to my plan and not let emotions drive me."

Next-day diary entry from same participant: "Took another loss trying to recover yesterday's loss. I know this is stupid but I can't help it in the moment."

Insight: Self-awareness ≠ self-control.

Users know what they should do, but in-the-moment emotional states override reflective intention. Decision-support can't rely on users remembering to be disciplined. It needs to interrupt the pattern in the moment of trigger, not after the fact.

The Strategic Reframe

The most important output of this research was changing the strategic conversation.

OLD FRAME

Build educational content and tools to help users make better decisions.

NEW FRAME
Design behavioral interrupts at emotionally predictable trigger moments FOMO, Revenge Recovery, Validation Seeking and integrate social context rather than override it..

​

This reframe shifted product strategy from education (assume users need more information) to intervention (assume users need decision friction in the moment emotions spike).

Strategic Recommendations

Based on diary study patterns, I recommended four behavioral intervention strategies:

01  FOMO FRICTION (Cooling-Off Period)
When a user attempts to buy a stock that has surged >10% in the last hour after receiving social trigger (detected via app activity pattern):
Intervention: "This stock is up 12% in the last hour. Take 10 minutes to review fundamentals before buying. We'll save this for you."
Rationale: FOMO trades happen within 15 minutes of trigger—introducing mandatory delay breaks the impulsive loop.

02  REVENGE RECOVERY DETECTION (Loss Awareness)
When a user opens the app within 24 hours of a significant loss and increases position size:
Intervention: "You're increasing your position size after yesterday's loss. 73% of users who do this lose more. Are you sure?"
Rationale: Explicitly name the pattern—make the unconscious behavior conscious.

03 MORNING ANCHOR INTERVENTION (First-Touch Optimization)
Deliver platform's daily market summary and personalized insights in the first 30 minutes after market open:
Intervention: Push notification with "Your Morning Market Brief" with curated, balanced view before users consume information elsewhere
Rationale: Become part of the morning ritual—set a balanced anchor before emotional information sources.

04 SOCIAL CONTEXT INTEGRATION (Peer Transparency)
When showing stock recommendations, include anonymized social proof from user's network:
Intervention: "2 people in your network researched this stock this week. Here's what data they looked at."
Rationale: Don't fight social influence—make it transparent and data-informed.

Impact & Outcomes

Product Decisions:

  • FOMO Friction feature was prioritized for Q1 2024 roadmap and tested with 5% of users showing reduction in same-day loss rates

  • Morning Market Brief was designed and launched based on timing insights from diary study

  • Social Context Integration was added to backlog with detailed behavioral specifications

Research Methodology Adoption:

  • Diary study approach was adopted as the standard method for understanding behavioral finance patterns across the research team

  • Hybrid analog-digital capture (WhatsApp + video) became template for future longitudinal studies

Organizational Learning:

  • Product and design teams shifted language from "educate users" to "design behavioral interrupts"

  • The emotional trigger taxonomy (FOMO, Revenge Recovery, Validation Seeking) became shared vocabulary for feature discussions

Research Reflection: What Made This Study Work

Longitudinal observation was non-negotiable. 

One-time interviews would have captured what users think they do, not what they actually do. The 4-week diary captured the gap between intention (evening reflection) and action (next-morning behavior).

Hybrid capture was essential.

Text-based diary entries alone would have missed emotional nuance. Video diaries captured tone, frustration, excitement—the affective states that text couldn't convey. WhatsApp bot made logging low-friction enough that participants actually did it.

Behavioral segmentation mattered.

If I had recruited only "rational investors," I would have missed the Impulsive Trader patterns. By deliberately recruiting across behavioral types, I could identify patterns that cut across segments (e.g., morning routines) and patterns that were segment-specific (e.g., Revenge Recovery more common in Impulsive Traders).

The post-study contextual interview was the synthesis moment.

Walking through participants' own diary entries with them, asking "what were you feeling here?" or "what happened right before this trade?" that's where the emotional triggers became visible. The diary captured the data; the interview made it meaningful.

This study reinforced that behavioral research requires observing behavior over time, not asking about behavior after the fact. Retrospective rationalization is not the same as real-time decision context.

bottom of page