Overview
Follow this guide to learn how to use the Demographics tab in an Audiense Insights report.
The Demographics tab provides an overview of your audience's demographic characteristics and compares them against your selected baseline. These insights help you understand who makes up your audience by analyzing information such as age, gender, location, language, and profile descriptions.
Because X (formerly Twitter) does not provide all demographic information directly, Audiense combines publicly available profile data with proprietary machine learning models and enrichment algorithms to infer additional demographic insights.
Understanding the Demographics Tab
The Demographics tab compares the characteristics of your selected audience or audience segment against your chosen baseline. Each visualization highlights how your audience differs from the comparison audience, helping you quickly identify meaningful demographic trends. The Demographics tab includes the following insights:
| Insight | Description |
|---|---|
| Gender | Distribution of audience members by gender and comparison against the baseline. |
| Country | Top countries where audience members are located, compared with the baseline. |
| City | Top cities represented within the audience and their variance from the baseline. |
| Language | Distribution of languages used by the audience compared to the baseline. |
| Bio | Most common words used in audience biographies and how they differ from the baseline. |
| Age | Distribution of audience members by age group compared to the baseline. |
| Location | Geographic distribution of the audience or selected segment. |
| Name | Most common names found within the audience and their comparison against the baseline. |
These visualizations provide a high-level understanding of who your audience is and how they differ from the broader population or another comparison audience.
How Audiense Enriches X Profile Data
While some demographic information is available from public X profiles, much of the data presented in the Demographics tab is inferred using Audiense's enrichment algorithms. Audiense combines publicly available profile information with machine learning models to estimate demographic attributes that X does not provide directly.
Examples include:
- Language is determined by analyzing a user's posts.
- Country and City are inferred from the profile location field using Audiense's location normalization algorithms, since X profile locations are entered as free text.
- Age and Gender are estimated using proprietary AI models and enrichment techniques.
To improve accuracy, Audiense analyzes multiple public signals, including:
- User names
- Profile photos
- Time zones
- Profile locations
- Posts
- Follows
- Likes
- Public demographic information
- Additional behavioral signals
These data sources work together to provide demographic estimates that help enrich audience analysis.
How Age Is Calculated Methodology
Audiense estimates age using machine learning models designed for facial detection, facial recognition, and age prediction. The system analyzes publicly available profile avatars to estimate an individual's age range. Rather than relying on a single analysis, the enrichment process continually evaluates user profiles over time to keep age estimates as accurate as possible.
Enrichment Process
As part of the ongoing enrichment workflow, Audiense regularly reviews profile avatars and attempts to estimate age when suitable images are available.
For an age estimate to be generated:
- The profile image must contain a single, recognizable face.
- Default profile images, illustrations, logos, group photos, or images containing multiple faces cannot be used.
- Whenever a user changes their profile image, a new age estimate is generated.
If the profile image remains unchanged for an extended period, Audiense automatically increments the estimated age by one year after approximately twelve months.
Note: Age estimates depend entirely on the profile image being analyzed. If users choose profile pictures that do not represent themselves—for example, celebrities, fictional characters, or family members—the estimated age may not accurately reflect the account owner.
How Gender Is Estimated Methodology
Gender estimates are generated using a name-matching system.
Audiense compares the display name from a user's X profile against a large database of first names that are statistically associated with specific genders.
Based on this comparison, the system attempts to classify each profile.
Enrichment Process
Profile names are continuously analyzed and classified into one of the following categories:
| Classification | Description |
|---|---|
| Male (M) | The profile name clearly matches a known male name. |
| Female (F) | The profile name clearly matches a known female name. |
| Undetermined (U) | The name is gender-neutral or ambiguous (for example, Alex or Jordan). |
| Unclassified | The name does not appear in the database or represents a brand, organization, emoji, or fictional character. |
This classification process also contributes to Audiense's account categorization. If a profile name can be classified as Male, Female, or Undetermined, the account is generally considered to represent an individual. If no classification can be made, the profile is more likely associated with a brand or organization. The current enrichment process enables gender classification for more than 95% of the user profiles in the Audiense system.