We’ve introduced a new feature that makes it easier to analyze data in Semly. From now on you can organize results faster, filter the most important information, and adjust views to your own way of working.

The new options are available in all key views in the Semly panel.

What has changed?

We’ve added three practical features:

Data filtering
You can narrow the view down to only those results that meet selected conditions. For example: show sources whose Share of Voice is higher than 20, or display only specific domains, brands, responses, or mentions.

Sorting results
You can now sort data by the most important parameters, including by source, Share of Voice, number of responses, or number of mentions. This way you’ll find the most important elements faster.

Managing columns
You can adjust every view to your needs. You only enable the columns that are needed for a given analysis, and you can hide the rest. This helps tidy up the screen and focus on the most important data.

Why does it matter?

Semly analyzes a large amount of information from AI responses, sources, prompts, brands, and competitors. The new options let you move from data to conclusions faster.

Thanks to filters, sorting, and column management, it’s easier to check:

• which sources have the biggest share in AI answers

• where your brand shows up most often

• which domains affect your competitors’ visibility

• which results need deeper analysis

• which data is currently the most important for your report

Who is this feature for?

The new options are especially useful for people who regularly analyze brand visibility in AI, compare themselves with competitors, check citation sources, or work with a larger number of prompts and topics.

This is another step towards more transparent, flexible, and practical AI Visibility analytics in Semly.

Where can you find the new feature?

Filtering, sorting, and column management options are available directly in report and table views in the Semly panel. Just use the icons above the table to tailor the data to your own analysis.

In Semly, we’ve launched a new feature for publishing content (articles) on a specially prepared knowledge page available in the Semly.ai infrastructure.

The option is available in the: AI Content tab → Configuration

Example AI knowledge base:

• knowledge.semly.ai

The new feature lets you build an official layer of knowledge about your brand, products and services, which can be more easily interpreted by AI models such as ChatGPT, Gemini, Claude or Google AI.

What is an AI knowledge page?

It’s a dedicated space for publishing content prepared for:

• language models (LLM)

• AI Search

• GEO (Generative Engine Optimization)

• AEO (Answer Engine Optimization)

• AI recommendation systems

The articles that are published create an organized knowledge base about:

• brand

• products

• services

• competitive advantages

• FAQ

• product use cases

• expert guides

• technical data

• company policies

• knowledge sources for AI

Thanks to this, AI models can more easily:

• recognize the brand

• understand the offer

• quote content

• recommend products

• answer customer questions based on the brand’s official data

How does it work?

Semly generates and publishes content on a dedicated brand subdomain, e.g.:

• knowledge.twojamarka.pl

The content is published in a special structure optimized for:

• indexing

• AI processing

• semantic analysis

• building brand authority

Once published, the articles become publicly available and can be analyzed by AI models and AI Search systems.

What does such a knowledge base give you?

Better understanding of the brand by AI

AI models get official and organized information about the company, products, and services.

Higher chance of recommendations

AI can more easily recommend the brand in its answers to users.

Building brand authority

You create your own source of knowledge that can be quoted and analyzed by AI models.

Better visibility in AI Search

The content can support visibility in:

• ChatGPT

• Gemini

• Claude

• Google AI Overview

• Grok

• other AI systems

Knowledge centralization

You can publish in one place:

• guides

• expert knowledge

• product data

• SEO/GEO content

• instructions

• FAQ

• industry analyses

How to activate the service?

Configuration requires connecting your own subdomain.

Step 1

Log in to your domain operator:

• OVH

• GoDaddy

• Cloudflare

• or another DNS provider

Step 2

Add an A record according to the instructions in the Semly.ai panel.

Example:

• Type: A

• Host/Subdomain: e.g. knowledge (it can also be another subdomain name - it’s up to you)

• IP value: 46.248.186.77

Step 3

After adding the record, it’s best to wait a few hours for DNS propagation and then run the verification, the VERIFY AND ACTIVATE button.

Step 4

After activation and changing the article status to:

• “Published on the website, e.g. knowledge.twojamarka.pl”

the content will be automatically published under the specified subdomain.

Hybrid publication

The new feature also supports a hybrid model. Remember to publish one article on only one site.

You can:

• publish selected articles on the AI knowledge page

• publish other articles on your own website

Just change the article status to:

• “Published on my site”

and provide the publication URL.

This way you can build:

• your own blog

• an AI knowledge base

• and official content sources for AI models at the same time'

Additional SEO and GEO recommendations for the AI knowledge base

To increase the effectiveness of the knowledge base and make it easier for AI models to find new content, we recommend:

• adding a link to the knowledge base, e.g. in the footer of the main service

• adding the knowledge page to the sitemap (sitemap.xml) of the main domain

Thanks to this:

• search engine bots will find new content faster

• AI models will more easily connect the knowledge base with the brand’s official domain

• the authority and credibility of the source goes up

• indexing improves and the chance of being cited in AI answers increases

Automatic technical infrastructure for AI

Every knowledge base created in Semly has automatically generated technical elements that support SEO, GEO and AI Search:

• XML sitemaps (sitemap.xml) - updated automatically

• robots.txt file

• llms.txt file for AI models - updated automatically

• a structure optimized for indexing

• automatic article linking

• URLs prepared for semantic analysis

The llms.txt file helps AI models understand:

• what the service is

• what content it contains

• which sections are the most important

• which data is the official source of the brand's knowledge

Everything has been prepared to make it as easy as possible to:

• analysis by language models

• building visibility in AI

• content citation

• recommending brand products and services

What’s next?

This is just the first stage of the system’s development.

In the next stage, Semly.ai will enable:

• automatic article generation

• automatic publishing

• creating complete topical clusters

• continuous development of the brand’s knowledge layer without manual handling

This means you can build a constantly updated brand presence in AI pretty much in automatic mode.

Why is this important?

AI models, however, need official and well-prepared sources of knowledge.

The new Semly.ai feature helps build such sources automatically, in an organized way and in line with the requirements of modern AI Search systems.

New! AI answer source analysis

We’ve added a new Sources module in Semly.ai that shows which domains and URLs AI models use when generating answers. This way you can check not only whether your brand appears in the answers of ChatGPT, Gemini, Claude, Grok or Google AI Mode, but also which sources those answers are based on.

This is an important step towards better understanding brand visibility in AI. Language models are increasingly pointing to specific websites, stores, rankings, guides, marketplaces and industry services as information sources. The new report helps you see which of them actually influence your brand’s presence in AI answers.

How the source report works?

The module analyzes AI model answers from the selected period and shows all detected sources. For each domain, Semly shows, among other things:

• share of voice, or SOV - the percentage share of a given domain in all detected sources,

• number of answers - how many AI answers included a given source,

• number of occurrences - how many times the source appeared in total, including multiple times in a single answer,

• co-occurrence with the brand - whether your brand appeared together with a given source.

This way you can quickly see which domains dominate in a given industry and where your brand has the biggest chance to increase visibility.

Key elements

Total number of source occurrences
Shows the total number of all detected sources in AI model answers in the selected period. It’s a general picture of how intensively the models use sources in the analysed area.

Main source
Indicates the domain that appeared most often as a source in AI answers. It can be a marketplace, store, comparison site, blog, ranking, or industry service.

Biggest gap
Shows a source that often appears in answers where your brand has a low share or doesn’t appear at all. It’s a practical hint for where it’s worth looking for ways to boost visibility.

Source table
The main table lets you compare all detected domains. You can check their SOV, number of answers, number of occurrences, and whether they appeared together with the analyzed brand.

Single source details

When you open a specific source, for example a competitor’s domain, a marketplace or an industry site, Semly shows an in-depth analysis:

Source SOV share
Tells you how often a given source appears compared to all detected sources in AI answers.

Brand occurrences
Shows how many times your brand appeared in AI answers together with the analyzed source.

Other source occurrences
Shows which other domains appeared in the same context.

Model ranking
Lets you check which AI models most often used the analyzed source. This way you can see the differences between ChatGPT, Gemini, Claude, Grok or Google AI Mode.

Ranking of other sources
Shows the pages that most often appeared together with the analyzed domain. This helps you understand in what kind of informational environment a given source appears.

Prompts
The prompts table shows specific queries in which the analyzed source was detected. You can also see there whether your brand appeared and which other sources were used in the same answer.

What value does this feature provide?

The new module lets you better understand why AI recommends some brands more often than others. If models regularly use certain sources, it’s worth knowing whether your brand is present there, whether your competitors have an advantage, and which places can have the biggest impact on improving visibility.

With source analysis you can:

• find the domains that have the biggest impact on AI answers,

• check where your competitors show up more often,

• spot gaps in your brand visibility,

• plan content, PR, SEO and GEO activities in places that are actually used by AI models,

• monitor whether your brand starts to appear next to important industry sources.

Źródła is another step in the development of Semly as a platform for monitoring and boosting brand visibility in AI. We no longer just show whether a brand was mentioned. We also show where models get their information from and where you need to be present to increase your chances of being recommended in AI answers.

What’s next?

The Źródła module will keep evolving. Soon we plan to expand it with things like competitor analysis, which will let you check even more precisely which sources AI models use when they recommend competing brands, where competitors have an advantage, and where your brand can grow its visibility. This will make the sources report an even more practical tool for planning GEO, SEO, content marketing, and PR activities.

At Semly, we are continuously expanding the Recommendations section by adding new tips and ready-to-implement solutions that help companies increase their visibility in AI chats.

These are not generic suggestions. They are specific recommendations designed to help businesses turn analysis into action faster and more effectively improve their brand presence in responses generated by models such as ChatGPT, Gemini, and other AI systems.

Why does this matter?

Because simply knowing that your visibility in AI is only the beginning. The real value comes from knowing exactly what to improve, how to implement it, and which actions are most likely to deliver the fastest results. That is why we are constantly adding new recommendations that support rapid growth and help companies make better use of the AI channel’s potential.

By implementing our recommendations, you significantly increase your real chances of being recommended in AI-generated answers.

We’ve introduced one of the most anticipated features in Semly.ai - Topics for prompts. It’s a step that changes the way you analyze visibility in AI from the level of single queries to the level of strategies and business areas.

Visibility in AI stops being a collection of single queries and starts being a game about entire areas of user intent. That’s why in Semly.ai we’re introducing Topics for prompts - a feature that lets you organize data, group prompts, and analyze results on the level of real business segments, not single phrases. Thanks to this, you get a fuller picture of your presence in AI models and can make decisions based on where you’re really winning and where you’re losing visibility.

Example:

You have an e-commerce store and you’re monitoring 100 prompts. Instead of analyzing each one separately, you create topics: “transactional”, “comparative”, “educational”. Suddenly you see that:

• you have high visibility in educational topics

• but you almost don’t exist in transactional prompts

The conclusion is instant:

the problem isn’t a lack of visibility in general, but a lack of presence where there’s purchase intent. With this, you can precisely adjust your GEO strategy and AI-focused content.

What's new?

1. Grouping prompts into topics

You can assign prompts to specific topics (e.g. “sales”, “marketing”).

Thanks to that:

• you organize hundreds of prompts into logical groups

• you analyze results not one by one, but in entire segments

2. Visibility reports for topics

The biggest change:

you can analyze visibility per topic, not just per prompt

That means:

• you see in which areas AI “understands” your brand

• you identify gaps (e.g. lack of visibility in “shopping” prompts)

• you make decisions on the GEO/AEO strategy level

3. Prompt search

We added quick search:

• you’ll find prompts by phrase

• you’ll assign them to topics in no time

• you scale work on large datasets

4. Grouped prompts view

You can switch the view

• classic (list of prompts)

• grouped by topics

Effect:

• better data overview

• faster segment analysis

5. Topic management (full control)

The “Topic management” option lets you:

• add and remove topics

• edit names

• assign colors (visual distinction of segments)

• maintain consistent classification

6. Assigning topics – in bulk or one by one

You can:

• assign a topic to a single prompt

• assign in bulk to many prompts at once by selecting them

Additionally:

• you’ll add a topic directly from the prompts list

• or from the topic management view

7. Topics in main reports

Topics have also shown up in dashboards:

• filtering data by topic

• trend analysis of visibility for segments

• better linking of data with the real business

Why this matters (from a GEO / AEO perspective)

So far the analysis looked like this:

• single prompt → single result

Now:

• topic = user intent = market segment

This changes everything:

1. Intent analysis, not keyword analysis

AI doesn’t work like Google

→ it works on contexts and intents

Topics let you map:

• “I want to buy”

• “I want to compare”

• “I’m looking for a solution”

2. Strategic decisions instead of chaos

Instead of:

• 100 prompts = chaos

You have:

• 5–10 topics = clear directions

3. Better content and GEO planning

You see:

• where you’re visible

• where you’re not

You can:

• create content for specific topics

• optimize feeds and data for segments

4. Scaling and analysis

For bigger projects this is crucial:

• hundreds / thousands of prompts

• many markets

• different AI models

Without topics → no control

With topics → full data structure

Summary

The new Topics feature:

• organizes data

• enables segment analysis

• lets you make GEO/AEO decisions at the strategy level

• significantly speeds up work on large prompt sets

This is the foundation for:

• better reporting

• better decisions

• greater visibility in AI

Starting today in Semly.ai we’re launching a new feature: generating expert articles designed specifically for visibility in LLM models, such as ChatGPT, Google Gemini, Perplexity, Grok, or Claude.

This tool was created for brands that want to intentionally fill AI models’ knowledge gaps and increase the number of recommendations in the topics that matter most to them.

What makes the new AI articles stand out?

1. Generated by over 20 AI agents

The articles are created with the help of specialized agents whose job is to build the semantic context needed by LLM models. The goal is to maximize the chances that your brand shows up in recommendation answers.

2. Filling models’ knowledge gaps

The content is designed to fill in missing information about the brand, products, and services—exactly in the places where models today don’t recommend your company.

3. Strengthening reputation in the AI ecosystem

A consistent, expert article structure builds brand credibility in the eyes of language models, which translates into more frequent and more precise recommendations.

4. Structure based on research and testing

The layout of the articles was developed based on numerous visibility tests in LLM to achieve the best possible results in the shortest time.

5. Option to edit before publishing

Every generated article can be freely edited, adjusted to the brand’s tone of voice, and supplemented with data, case studies, or internal linking before publication.

Limits in plans

• Premium – 5 articles per month

• Ultra – 10 articles per month

• Mini – service unavailable

In the Ultra plan, you can increase the number of articles; to do this, please contact the Semly team.

The limit resets automatically on the 1st day of each month.

How to use the feature effectively?

To get the best results:

• Choose prompts for which there are no recommendations yet, but that are strategically important for the brand.

• For one prompt, we recommend preparing 2–3 articles.

• Publish at most 1 article per day, so up to 30 per month, to keep a natural rhythm of indexing and content distribution.

How to generate an article?

1. Go to the “Content for AI”

2. Click “Generate content”

3. Choose a prompt

4. The agent will suggest a few topics – pick one

5. Wait a few to several minutes

After the article is generated, you can:

• edit it directly in the system

• copy it as HTML

• copy it as plain text

• copy it as Markdown

What’s next?

Soon we’ll make it possible to publish articles on a specially prepared page in the brand’s domain. We’ll share the details in a separate announcement. This is another step towards consciously managing visibility in AI and building an edge in a new customer acquisition channel.