Signal Master

Signal Master is the screen for defining and managing signals captured from customer meetings and their impact on win probability. The signal list that sales representatives select when recording meeting results in the Activity Board is managed on this screen.

All users can view signals, but creating, editing, and deleting signals and impact types is restricted to admin or super_user roles only.

Location: Sidebar → Settings → Signal Master

What is a Signal?

In sales, a signal is a reaction shown by the customer during a meeting — a clue that indicates whether the deal is more or less likely to succeed.

For example:

• Customer says "The budget has been approved" → Strong Positive signal
• Customer says "Let's have an additional meeting next week" → Weak Positive signal
• Customer says "We've decided to go with a competitor's product" → Strong Negative signal
• Customer says "We'll review it" (polite formality) → No Signal

EXAWin doesn't just record these signals. Each signal is assigned a mathematical weight (Impact), and every time a signal is entered, the Bayesian engine automatically recalculates the win probability.

Signal Master manages two things:

1. Signals — Specific situations that can be captured in meetings (e.g., "Budget Approved", "Contact Changed")
2. Impact Types — Categories that determine the magnitude of a signal's influence (e.g., "Game Changer", "Strong Affirmation")

Impact Types: The Scale That Determines Signal Influence

To understand Signal Master, you first need to learn about Impact Types. Impact Types define the magnitude of influence that a signal has on win probability.

Standard Impact Types

EXAWin provides 7 system standard Impact Types:

How Impact Values Affect Win Probability

Understanding how Impact values actually work will naturally explain why these values should not be changed arbitrarily.

When a sales representative records a meeting, the following calculation occurs internally:

• If it's a positive signal, α (success weight) increases → P(Win) ↑
• If it's a negative signal, β (failure weight) increases → P(Win) ↓
• SWV (stage weight) and Impact are multiplied together to determine the influence

To put it simply:

A signal is "what happened" and Impact is "how important it is". Just like in news reporting, the same event has a completely different impact on the world depending on whether it's a front page headline (Game Changer, Impact 5.0) or a small article on page three (Weak Affirmation, Impact 0.4).

⚠️ Why You Should Not Arbitrarily Change Impact Values

Impact values are not random numbers. They are calibrated using a ratio-based design (f-coupling) linked to the company's initial setting of Prior strength (S = α + β = 10):

If this ratio is broken, the following problems occur:

If Impact is set too high — Probability surges or drops dramatically with a single meeting. For example, raising "Strong Affirmation" from 1.0 to 8.0 could cause P(Win) to jump from 25% to 80% with just one signal in the Negotiation stage. This produces predictions completely detached from reality.

If Impact is set too low — Even the most decisive signals barely move the probability. Imagine a customer saying "We'll send the purchase order" and the system only moves the probability from 28% to 29%.

💡 It is strongly recommended to maintain standard values for Impact. Once sufficient data has accumulated, EXAWin's Auto-Tuner will analyze past Won/Lost records and automatically suggest optimal values.

For detailed documentation on parameter calibration principles, see the following:

• Bayesian Application: Prior Alpha, Beta — Prior Setting Principles
• Bayesian Application: Parameter Calibration and Auto-Optimization — f-coupling Principle, EPR Guardrails, Auto-Tuner

Screen Layout

The screen is composed of a 2-panel layout:

• Left Panel — Signal list or Impact Type list (tab switching)
• Right Panel — Signal/Impact Type creation and editing form (synced with tab)

Top Header

Tab Switching

There are 2 tabs at the top of the right panel:

Switching tabs simultaneously changes both the left table and the right form.

Signals Tab

Signal Table

Clicking a table row loads the signal information into the right form for editing.

Creating Signals

After resetting the form with the Reset button at the top, enter the following fields:

Click the Save button at the top after entering to save. Newly created signals are marked as Custom.

Editing Signals

Clicking a signal in the table switches the right form to edit mode:

• An "Editing" badge is displayed at the top of the form
• Modify fields and save with the Save button
• Click the Reset button to exit edit mode and return to new signal entry mode

Signal names can be changed freely. Name changes do not affect Bayesian calculations.

Deleting Signals

Click the Delete button in edit mode to delete a signal.

• ❌ System signals cannot be deleted
• ❌ Signals that are in use in activity records cannot be deleted (referenced in past meeting records)
• ✅ If conditions are met, deletion proceeds after a confirmation popup

Deleted signals' past activity records are preserved as-is. They are only removed from the list, so rest assured that existing data will not disappear.

Impact Types Tab

Impact Type Table

Creating Impact Types

When "No Signal" is selected, the Base Value is automatically set to 0.1 and cannot be modified.

Editing Impact Types

Clicking an Impact Type in the table switches to edit mode. The UI pattern is the same as for signals.

Deleting Impact Types

• ❌ System Impact Types cannot be deleted
• ❌ Impact Types in use by signals cannot be deleted (signals are linked to this Impact Type)
• ✅ If conditions are met, deletion proceeds after a confirmation popup

Relationship Between Signals and Impact Types

Signals and Impact Types have a many-to-one (N:1) relationship. Multiple signals are linked to a single Impact Type.

For example, the "Strong Affirmation" Impact Type can have the following signals linked to it:

All these signals are reflected with the same Impact value (1.0). Signals distinguish "what happened", and Impact Types determine "how important it is".

Important Notes

• Deactivating a specific signal excludes it from the signal selection in the Activity Board.
• Changing signal names freely does not affect Bayesian calculations. However, Impact values (Base Values) are standard values calibrated according to the f-coupling principle, so arbitrary changes may distort probability calculations.
• When adding new Impact Types, it is recommended to set values within the standard scale range (0.1 – 5.0). Values outside this range will violate the EPR (Evidence-Prior Ratio) guardrails.
• Click the signal help button (❓) to view detailed descriptions about Impact.