The S-FLPR Framework
S-FLPR is ConCRG's model for how users experience assistance inside software. It maps the natural journey of learning any feature — and defines when and how ConCRG should respond.
The Five Modes
S ─── Sense (proactive detection)
│
├── F ─── Find "Where is it?"
│
├── L ─── Learn "What does it do?"
│
├── P ─── Practice "How do I use it?"
│
└── R ─── Remember "I've done this before"
SENSE
ConCRG monitors behavioral signals to detect confusion before the user asks for help. When a frustration pattern is detected, it proactively offers the right mode.
Detection signals:
- Dwell time (5+ seconds on a page without action)
- Rage clicks (3+ rapid clicks on the same element)
- Dead clicks (clicking non-interactive elements)
- Oscillation (repeated back-and-forth navigation)
FIND
The entry point for most users when they're stuck. FIND answers "where is it?" and takes them there.
What it does:
- Interprets the user's intent from natural language
- Searches the knowledge graph for the relevant page or feature
- Navigates the user directly (ghost layer highlights the path)
- Offers escalation to LEARN and PRACTICE
Example: "Where do I create an invoice?" → navigates to /billing/invoices/new
LEARN
Once users have found a feature, they want to understand it. LEARN explains how features work — at three levels of depth.
| Level | Example Question | Response |
|---|---|---|
| Quick | "What is a pipeline stage?" | Single paragraph inline |
| Conceptual | "How does renewals work?" | Expanded card with context |
| Deep | "What's the relationship between deals and invoices?" | Multi-turn conversation |
PRACTICE
The highest-rated mode in user testing. PRACTICE lets users try workflows safely — no consequences, no fear.
Two flavours:
- Guided walkthrough — CRG leads, user follows step by step
- Supervised exploration — User leads, CRG watches and helps if stuck
Example: "Walk me through creating a deal" → CRG highlights each element, explains each step, user clicks through.
REMEMBER
For returning users who have done something before but need a refresher. REMEMBER connects the current moment to past activity.
Example: "How did I export contacts last time?" → shows the steps from the user's last successful export.
The Natural Journey
Users don't think in modes — they think in tasks. The FLPR progression maps to how anyone learns a new tool:
First encounter with a feature:
FIND → LEARN → PRACTICE
Returning after time away:
FIND (or SENSE) → REMEMBER → PRACTICE (quick refresh)
Power user:
FIND → done (they know the rest)
ConCRG tracks where each user is in this journey per feature and adapts accordingly. The same question — "tell me about invoicing" — gets a different answer depending on whether the user has never visited Invoicing, has found it but never used it, or used it last month.
Adaptive Response
The command bar is the universal entry point. The response UI is the variable. ConCRG classifies intent and journey stage to produce the right response automatically — the user never selects a mode.
| User Input + Context | ConCRG Response |
|---|---|
| "Where is billing?" (first time) | FIND: location card + "Go there" button |
| "Where is billing?" (visited before) | REMEMBER: "You were last here on March 15" + "Go there" |
| "How does billing work?" (found, not learned) | LEARN: explanation card |
| "Walk me through creating an invoice" | PRACTICE: guided walkthrough |
Read more about Adaptive Response →
Design Principles
The S-FLPR framework is grounded in five design principles learned from user research:
- Timing > Content — Right help at the right moment beats the most thorough documentation
- Show, Don't Tell — Demonstrations beat written instructions
- Safe to Explore — Remove fear of consequences; users engage more deeply
- Whisper, Don't Shout — Subtle, non-intrusive presence keeps the host app primary
- Speak Their Language — Frame help around tasks and goals, not feature names