Brand24 changelog brand24.com

New

May

We’ve added a new Sources tab in AI Visibility to help you understand where LLM citations come from and which sources have the biggest impact on your visibility.

In this view, you can explore:

• citation sources
• websites most frequently referenced by AI models
• comparison tables for those sources
• source type breakdowns
• which AI models generated the most citations

Why it matters: this data shows where your brand should appear to increase the chances of being cited by LLMs. For example, ranking/listing websites are often highly visible in AI-generated answers — if your brand isn’t present there, it may also be missing from LLM responses.

New

Get sharper answers to harder questions. Brand Assistant 2.0 is here, with stronger reasoning, richer visuals, and access to the open web.

Reasoning that adapts to your question. Brand Assistant works as an agent: it picks the right tools, pulls the right data, and adjusts as it goes. Complex questions get complete answers in one go.

Charts placed exactly where they matter. Visualizations appear inline, right next to the insight they support. Column, pie, and line charts, picked to fit the data.

Answers that reach beyond your project. Brand Assistant can now search the web and read specific pages on demand. Paste a competitor's URL, pull context from a news story, or fact-check a claim in real time.

Open Brand Assistant →

Read a full guide to learn about six ways to use AI Brand Assistant

New

May

Check how your brand performs in answers from ChatGPT, Gemini, Claude, and other AI models, alongside your Brand24 results, in a single panel.

Open the tab from your panel, enter your domain or brand, and you're set!

The first data on how ChatGPT, Perplexity, Google AI Overview, and Gemini talk about your brand will appear in your panel right away.

A new project automatically starts a 14-day AI Visbility trial: 30 prompts monitored across 4 AI models.

Read the full AI Visibility guide →

New

April

Upgraded emotion analysis is back in the panel - both as a graph and a filter.

With the new model we introduced for emotion analysis you can better observe which emotions are the main driver of a discussion about your (or competitor's) brand or product, analyse emotions around campaigns and specific hashtags or combine it with other data to provide a comprehensive report for the management.

You can find the graph in the Analysis view and the filter in every part of the panel that uses filters.

New

April

We’ve introduced a major performance upgrade for large-scale projects. Dashboards handling hundreds of thousands of mentions per month now load analysis views in seconds, even when using text filters.

What changed?

We’ve implemented changes to our databases designed specifically for fast, full-text querying. This significantly improves how quickly data is processed and filtered across the system.

Who will see the improvement?

Currently, the new system is available for a subset of largest projects, primarily the ones using text filter frequently.

We’re gradually rolling it out to more projects while ensuring system stability.

Important limitations

The new engine supports a rolling last 3 months of data. Queries that span over period older than 3 months will still be using the old search system.

Text filter behaviour change

Text search now matches from the beginning of words only:

✅ bra → matches Brand24

❌ rand → no longer matches Brand24

While results of some existing text filters may vary slightly compared the previous version, this trade-off enables much better performance.

Why it matters

This change directly improves one of the core Brand24 use cases—fast exploration and analysis of large mention datasets —making the platform significantly more responsive for power users and enterprise-scale monitoring.

April

Boost

We’ve updated how Advertising Value Equivalent (AVE) is calculated to better reflect the current realities of the online marketing landscape.

The AVE model now takes into account the rising costs of digital advertising and inflation across media markets, ensuring that calculated values better represent the real-world price of gaining similar visibility through paid campaigns.

This change results in:

• AVE values that better reflect the current cost of online exposure

• More realistic estimates aligned with today’s digital advertising market

• Improved accuracy when reporting the earned value of media coverage

No action is required from users — the update is applied automatically to new calculations. Already existing mentions and calculations are not affected.

New

Feb

Boost

Brand24 is now an official app in the OpenAI App Store. Connect it in two minutes and start getting live brand insights without leaving ChatGPT.

What this is

Brand24 is one of the first social listening platforms to become an official ChatGPT app. That means you can analyze your brand monitoring data - mentions, sentiment, influencers, media coverage - directly inside ChatGPT, using natural language.

No tab switching. No exports. No copy-pasting data into prompts.

How to get started

• Install the Brand24 app from the OpenAI App Store
• Connect your Brand24 account with one click
• Ask ChatGPT anything about your brand data

What to ask

Start with these prompts to see what's possible:

• "Summarize what's happening in my [project name] project"
• "What were the biggest spikes in mentions for [project name] this month?"
• "Who are the top influencers talking about [brand]?"
• "What's the current sentiment trend for [project name]?"
• "List news sites that covered [brand] this week"_

From there, you can combine insights with ChatGPT's capabilities — writing a media briefing based on current coverage, flagging a potential PR issue before it escalates, or identifying a competitor's vulnerability from their mention data.

Want to connect Brand24 to Claude or other AI tools?

The Brand24 ChatGPT app uses MCP (Model Context Protocol) under the hood — the same standard that works with Claude, Gemini, Cursor, and other AI agents.

See the full MCP setup guide →

Jan

Boost

We’ve improved how Advertising Value Equivalent (AVE) is calculated by increasing the role of geolocation signals.

AVE now better reflects differences between markets by accounting for geographic context more accurately. When available, country and language information is used to align AVE values more closely with local media and advertising conditions.

This change results in:

• More realistic AVE values across different regions

• Better comparability of results for international brands

• Increased accuracy when analyzing global coverage

No action is required from users - the update is applied automatically to new calculations. Already existing mentions and calculations are not affected.

Jan

Fix

We’ve identified that the latest Grok release affected our reach algorithm, causing reach values to be artificially boosted for some mentions.

✅ The issue has been fixed for all new mentions — reach is now calculated correctly.

📊 Historical data may still reflect the inflated reach caused by Grok.

How to analyze historical data without Grok impact:

To exclude Grok-related mentions from past analyses, use the Author filter and apply: {NOT}grok

This will let you review historical reach metrics without Grok influencing the results. If you have any questions or notice anything unexpected, let us know!

Jan

By January 31, Emotions and Intents will be removed:

• as filters
• as analysis features
• and from statistics within AI Topic Analysis

After this date, Emotion and Intent data will no longer be available in these areas. This change is necessary for upcoming sentiment model upgrades, that will significantly increase its quality, especially for languages other than English.