SD SEO Suite is undergoing a major change.
We’re moving beyond traditional SEO reporting and AI visibility tracking. The next stage is about giving SEO professionals the ability to simulate, interpret, analyse, compare and build strategies around Google AI Mode.
Why?
Because knowing whether your website appears in an AI-generated answer is only part of the picture.
You also need to understand why Google is using your website as a source, why competitors are being cited instead, which sources appear repeatedly, and what you can do to improve your chances of being included.
We’re building a new set of capabilities to help answer those questions.
Using Gemini 4 Argon, Gemini 3.8 Flash and Gemini 3.5 Flash-Lite, with Gemini 4 Carbon coming soon, we’ll be able to investigate AI Mode responses in much greater detail.
This is not simply another AI visibility tracker. It’s about giving SEOs the tools to understand what’s happening behind the answers and make better decisions based on the findings.
1. What’s Changing?
The biggest change is how we approach Google AI Mode.
Until now, most SEO tools have focused on tracking rankings, monitoring AI mentions, counting citations and comparing brand visibility.
These metrics are useful, but they don’t always explain what is happening.
For example:
- Why is Google using a competitor’s website instead of yours?
- Why does one source appear across several AI Mode responses?
- Which websites are being used to support particular parts of an answer?
- Are your competitors being cited for information that your website also provides?
- Which topics and questions create opportunities for your brand?
- What would happen if you changed the way a page answers a particular question?
We want to make it possible to investigate these questions directly.
The new Google AI Mode functionality will focus on five areas:
| Capability | What it means for SEO teams |
|---|---|
| Simulate | Run controlled AI Mode-style investigations around target searches and customer questions |
| Interpret | Understand the answer, supporting sources and relationships between the information |
| Analyse | Investigate why particular sources are being used and identify patterns across responses |
| Compare | Compare your website’s visibility and citations against competitors |
| Strategise | Turn the findings into practical SEO, content and digital PR opportunities |
The aim is to move from knowing what happened to understanding what to do next.
2. Why Google AI Mode Needs a Different Approach
Google AI Mode isn’t simply another set of traditional search results.
It can break a question into several related searches, explore different topics and retrieve information from multiple sources before generating an answer.
For example, someone searching for:
Which air compressor should I buy for a professional car detailing business?
might need information about:
- Compressor capacity.
- Airflow requirements.
- Tank size.
- Pressure.
- Portability.
- Noise levels.
- Different manufacturers.
- Price and availability.
- Suitability for specific detailing equipment.
A traditional keyword tracking tool might tell you where your website ranks for the original search.
But that doesn’t tell you which websites Google uses when answering the individual questions behind it.
This is where the new functionality becomes interesting.
We want SEOs to investigate the different parts of the search journey, understand which sources support the answers and compare how their own website performs against competing websites.
The opportunity isn’t just to rank for the original question. It’s to become a useful source across the wider set of questions that influence the answer.
3. Simulate Google AI Mode
The first part of the new functionality is simulation.
We’ll allow SEOs to investigate searches using different Gemini models and examine the responses they produce.
The purpose is to create a controlled environment for researching how different questions, topics and sources influence AI-generated answers.
For example, an SEO team working with an air compressor manufacturer could investigate:
- Best air compressors for car detailing.
- Portable compressors for mobile detailing businesses.
- Air compressors suitable for Tornador cleaning guns.
- Recommended compressor tank sizes for professional use.
- Quiet air compressors for indoor workshops.
- Differences between leading compressor manufacturers.
Instead of looking at each keyword independently, the team can investigate the wider topic and compare the sources appearing across the responses.
This creates a much more useful starting point for AI search research.
Important: The simulation uses Gemini models to investigate and analyse AI search behaviour. It is not a direct copy of Google’s production AI Mode, nor does it provide access to Google’s private retrieval or ranking systems. Actual Google AI Mode results can differ.
4. Understand Why Websites Are Grounded and Cited
This is where I think the biggest opportunity sits.
We don’t just want to show you a list of websites that appear in an AI-generated response.
We want to help you understand why those websites are being used.
Grounding connects an AI-generated response to information retrieved from external sources. In Google’s AI search experience, this includes retrieving relevant pages and using them to support the answer.
For SEO professionals, this creates several questions worth investigating.
Why is Google using a particular source?
Imagine that Google AI Mode is answering a question about the best air compressors for a car detailing business.
The answer might reference:
- A manufacturer’s product pages.
- An independent equipment retailer.
- A specialist automotive publication.
- A discussion on Reddit.
- A product comparison website.
The important question is what each source contributes.
One website might provide detailed technical specifications. Another might explain how a particular compressor performs in a real workshop. A third might provide product prices and availability.
The new functionality will help us investigate the sources and the information they contribute to the response.
Why are competitors being cited instead of your website?
This is one of the most important questions for a modern SEO professional.
You might already rank well in Google Search, have extensive content and publish detailed product information.
Yet a competitor could still appear more frequently in AI Mode.
We want to make it easier to compare the sources used across responses and identify where competitors have an advantage.
For example:
| What we investigate | What we want to understand |
|---|---|
| Competitor citations | Which competing websites appear most frequently? |
| Shared citations | Which sources appear in responses for both your brand and competitors? |
| Source coverage | Which topics and questions are supported by your website? |
| Missing sources | Which websites are being used that your SEO team hasn’t considered? |
| Content differences | What information do cited competitors provide that your pages don’t? |
| Citation patterns | Are particular sources repeatedly appearing across related questions? |
This should help turn AI search research into something much more actionable.
Understand shared citations
Another area we’ll be investigating is how sources overlap across different answers.
For example, imagine running 100 related questions around a product category.
A particular industry publication might appear in 35 responses. A competitor might appear in 22. Your website might appear in 12.
These figures would help identify which sources repeatedly appear across the research set.
We can then investigate the questions, topics and information associated with those citations.
This is particularly useful when comparing brands competing for visibility across the same category.
5. Multiple Gemini Models, Different Ways to Investigate AI Search
We’re not building this around one model.
Different models have different strengths, costs and response characteristics. Using more than one gives us greater flexibility when investigating large sets of questions.
The initial model selection includes:
| Model | Intended role |
|---|---|
| Gemini 4 Argon | More detailed analysis, complex investigations and interpreting relationships between sources |
| Gemini 3.8 Flash | Faster investigations and processing larger sets of questions |
| Gemini 3.5 Flash-Lite | Lightweight analysis and more efficient processing |
| Gemini 4 Carbon | Coming soon |
The idea is to use the appropriate model for the task rather than relying on one model for everything.
For example, a large-scale investigation might use Flash-Lite for initial processing, Flash for broader analysis and Argon for more complex interpretation.
Gemini 4 Argon and the Flash models also support Google’s grounding capabilities, which makes them useful for working with search-backed information and source references.
Carbon is part of the next stage of our model development work.
6. What Will SEO Professionals Be Able to Do?
The objective is to make Google AI Mode research more practical.
Here are some of the workflows we’re working towards.
1. Investigate a category
Start with a category, product or service.
Run a collection of relevant questions and investigate the sources appearing across the responses.
This helps establish which websites are being used to support answers about that category.
2. Compare your brand against competitors
Select your website and competing brands.
Compare the sources appearing across the same questions and identify where competitors have stronger visibility.
The objective is to understand the differences, not simply produce another share-of-voice chart.
3. Identify missing content opportunities
Find questions where competitors are cited but your website isn’t.
Investigate what those competing pages provide, whether your website already covers the subject and what information could make your content more useful.
This can help identify opportunities for:
- New content.
- Existing page improvements.
- Product information.
- Supporting guides.
- Original research.
- Digital PR.
- Industry-specific resources.
4. Find websites worth building relationships with
If particular publications, industry websites or specialist resources repeatedly appear in AI Mode responses, they may be worth investigating from a digital PR perspective.
The objective is to identify websites that matter within the topics your business wants to lead.
5. Build a strategy around the findings
Ultimately, the research needs to lead somewhere.
We want SEO professionals to use the findings to decide:
- Which pages need improving.
- Which topics deserve more attention.
- Which competitors need further investigation.
- Which websites are worth approaching.
- Which content gaps should be prioritised.
- Where the biggest opportunities exist.
The point isn’t to generate hundreds of AI responses and leave you with another spreadsheet.
It’s to help you make better SEO decisions.
7. How This Fits Into SD SEO Suite
This is a major step in the direction we’re taking with SD SEO Suite.
The platform already brings together traditional SEO data and AI visibility research.
That includes:
- Google keyword research.
- Google SERP analysis.
- Rank tracking.
- Google AI Overview citation research.
- AI visibility tracking across ChatGPT, Gemini, Perplexity and Claude.
- Competitor research.
- Backlink data.
- Website audits.
The Google AI Mode work adds another layer: the ability to investigate how Google’s AI search experience uses sources and how those sources compare across different questions.
I don’t want SD SEO Suite to become another tool that gives you a visibility score and leaves you to work out what to do with it.
I want it to help SEO teams investigate the market, understand competitors, identify opportunities and make decisions.
Data is useful. Understanding what the data means is much more valuable.
8. What’s Coming Next?
This is a substantial change, so we’re not treating it as a small feature release.
There are several areas we’re working on:
- Google AI Mode simulation: investigate searches and compare responses using different Gemini models.
- Source analysis: understand which websites and pages are being used to support answers.
- Citation comparison: compare your website’s citations against competitors across the same research set.
- Grounding analysis: investigate why particular sources are being used and what information they contribute.
- Category research: identify patterns across related questions rather than analysing isolated searches.
- Strategic recommendations: turn the findings into practical SEO, content and digital PR opportunities.
- Gemini 4 Carbon: introduce the model when it becomes available for our intended workflows.
We’ll continue developing these capabilities and improving the way the information is presented.
9. The Bigger Picture
SEO is changing, but I don’t think the answer is to abandon everything we’ve been doing.
We still need to understand search demand, rankings, technical SEO, content quality, competitors and the websites that influence buying decisions.
What’s changing is the amount of information we need to consider when measuring search visibility.
A website can rank well in traditional Google Search but have limited visibility in AI-generated answers.
A competitor can have fewer rankings but appear repeatedly as a source across important questions.
An industry publication might influence several answers without ranking particularly well for the original keyword.
These are things SEO teams need to investigate.
And that’s the direction we’re taking with SD SEO Suite.
We’re building software that helps SEOs move beyond monitoring visibility and towards understanding how search engines and AI systems use information.
The goal is simple: give SEO professionals better data, better insights and better ways to compete.
10. About Szymaniak Digital
SD SEO Suite is developed by Szymaniak Digital, a UK-based Enterprise AI SEO Consultancy.
We work with businesses operating in competitive markets, helping them improve their visibility across traditional search engines and AI platforms.
The software is built around the same challenges we deal with in consultancy work: understanding competitors, measuring search performance, identifying opportunities and turning data into decisions.
We’re continuing to develop SD SEO Suite based on the needs of SEO professionals and the changes taking place across search.
We’re not interested in building another dashboard full of numbers. We want to build software that helps you understand what’s happening and decide what to do next.
