Understanding GrowByData’s AI Sentiment Tracking

Devik Balami |
|READ 7 MIN
Understanding GrowByData AI Sentiment Tracking

A guide to Sentiment Composition & Favorability Score

What is GrowByData tracking?

GrowByData’s Compass platform monitors how AI engines like ChatGPT, Gemini, and others respond when users ask questions about your brand, products, or category. Rather than just telling you whether AI mentions you, Compass measures the tone and quality of those mentions over time.

To do this, Compass uses two distinct metrics. You’ll see both of them in the platform, and it’s important to understand what each one means because they answer different questions.

The Two Metrics at a Glance

Sentiment Composition

The breakdown

Shows how AI responses about you are distributed across positive, neutral, and negative sentiment. It’s a percentage split like a pie chart of AI opinion.

“60% positive, 30% neutral, 10% negative”

Favorability Score

The summary score

A single weighted number that captures overall sentiment in one figure. Designed so you can track trends at a glance and compare performance over time.

“Overall perception score = 65”

⚠️ Why don’t the two numbers match?

This is expected, not a bug. Because the two metrics measure sentiment in different ways, they will rarely produce the same number. Read on to understand how each is calculated.

What Does AI Really Say About Your Brand?

AI engines are already influencing buying decisions. Find out how your brand is represented.

Schedule a Straegy Session

Sentiment Composition

What it shows

Sentiment Composition tells you how AI responses are distributed across three categories: Positive, Neutral, and Negative. Each response analyzed by GBD Compass gets assigned one of these labels, and the metric shows you the percentage that falls into each bucket.

Example

Imagine Compass analyzed 100 AI responses mentioning your brand:

Sentiment # of Responses Percentage
Positive 60 60%
Neutral 30 30%
Negative  10 10%

How to interpret it

Sentiment Composition is your distribution view. Use it to understand:

  • How much of the AI conversation about you is positive vs. negative
  • Whether neutral responses are dominating (which may indicate AI engines are staying vague or non-committal)
  • How the mix shifts over time for example, if negative responses are growing even while the overall volume stays flat
  • Important: this metric does not produce a score. It only shows proportions.

Favorability Score

What it shows

The Favorability Score condenses all the sentiment data into a single number so you can quickly gauge overall perception and track changes over time. Rather than treating all sentiments equally, it applies different weights to each because a positive response is more meaningful than a neutral one, and a negative response should pull the score down.

How it’s calculated

Each sentiment category is assigned a weight:

Sentiment Weight
Positive +1.0
Neutral +0.5
Negative -1.0

Using the same 100-response example from above:

Sentiment Count Weight Result
Positive 60 × 1.0 =60
Neutral 30 × 0.5 =30
Negative 10 × -1.0 =-10

Final calculation: 60 + 15 − 10 = 65

The Favorability Score for this example is 65.

How to read the score

Score Range What It Means
Above 50 Mostly positive – AI engines are speaking favorably about you
Around 50 Mixed or balanced – positive and negative are roughly offsetting
Below 50 Mostly negative – AI engines are expressing unfavorable sentiment

Using Both Metrics Together

These two metrics are designed to complement each other. Looking at them side by side gives you a more complete picture than either one alone.

Use Sentiment Composition when you want to…

  • Understand the full distribution of AI responses
  • Shows how AI responses about you are distributed across positive, neutral,
  • See whether neutral or negative volume is growing
  • Present a detailed view to stakeholders
Use Favorability Score when you want to…

  • Get a quick read on overall AI perception
  • Track performance trends week over week or month over month
  • Compare you score against competitors
  • Share a simple headline metric with leadership

Finding These Metrics in Compass

Both metrics are available within the LLM Tracking section of the Compass platform. The screenshots below show you exactly where to navigate.

Sentiment Composition – where to find it and how to read the chart

sentiment-composition-view-in-GBD-compass-v2
Sentiment Composition view in Compass
Bar Chart

Sentiment Composition tells you how AI responses are distributed across three categories:

  • Green – Positive Sentiment.
  • Yellow – Neutral Sentiment.
  • Red – Negative Sentiment.
To Navigate report:

  1. Go to LLM
  2. Sentiment
  3. Sentiment Composition

navigate sentiment composition report

Favorability Score – where to find it and how to track changes over time

sentiment overview


How to Track changes over time:

  • Select the desired date range filter and review how the “Favorability Score” has changed for your brand as well as for other brands

Note: 

  • You can also view the breakdown of your favorability score by platforms.
To Navigate report:

  1. Go to LLM

     -> Sentiment

        -> Sentiment Composition

navigate sentiment overview report

Need help interpreting your data?

Reach out to your GrowByData account team. We’re happy to walk through your Compass dashboard, explain what your current scores mean, and help you identify opportunities to improve your AI visibility.