Business lawyer Rebecca Steer from Bristol Creative Industries member Charles Russell Speechlys recently delivered an event covering the latest employment, copyright, data and artificial intelligence (AI) legal updates that creative digital agencies need to know. Here’s a summary of the advice she shared.
Bristol Creative Industries members can book a free 30 minute legal advice session with Rebecca Steer. Slots are available on 23 May and 27 June.
Employment regulations
A new duty on employers to proactively take reasonable steps to prevent sexual harassment has been in force since October 2024, as part of the Worker Protection (Amendment of Equality Act 2010).
Steps you should already be taking include:
- Develop an effective anti-sexual harassment policy
- Carry out risk assessments
- Provide clear and accessible reporting channels
- Provide training
- Deal with complaints immediately
- Consider third party harassment
- Conduct staff surveys to measure confidence of an harassment-free environment
Employment Rights Bill
The government’s major changes to employment rights are expected to take effect from 2026. Rebecca advised that you should be prepared to adjust the way you recruit, contract and manage your workforce.
The changes, which are subject to the legislation being passed, include:
- New day one rights for employees covering not being unfairly dismissed, receiving statutory sick pay, claiming paternity and parental leave, and flexible working.
- Enhanced protection for new mothers with prohibition on dismissal during pregnancy or after a protected period of maternity, adoption or shared parental leave (other than redundancy).
- Ending “exploitative zero hours contracts” with employers required to make a guaranteed hours offer to a worker after the end of a set period (12 weeks is the suggestion), workers to receive reasonable notice of cancellation or a change to a shift, and compensation to be paid to the worker where a shift is cancelled, moved or curtailed at short notice.
To prepare for the changes, Rebecca recommended that you review certain situations, such as:
- Day one right to unfair dismissal: Ensure managers spot underperformers at an early stage and embed good practices for probationary period reviews.
- Day one right to flexible working: Review flexible working policies and consider justifications for refusing requests with draft questions checked by legal experts.
- Zero hours contracts: Monitor the number of hours individuals are working to be prepared for the new rules.
Data (Use and Access) Bill
The consultation period for the Data (Use and Access) Bill ended in January 2025. If accepted, it is likely to receive Royal Assent mid-2025 with enforcement likely to be in early 2026.
The key changes are:
- Smart Data: Data intermediaries can act as trusted third parties that facilitate customer data sharing by service providers (e.g. financial services) with authorised intermediaries.
- DSARs (data subject access requests) will be for reasonable and proportionate reasons and additional time can be added before formal timescales to clarify scope.
- Fines for cookie/direct marketing infringement will be increased from the current £500,000 cap to those under UK GDPR (higher of 4% or £17.5 million).
- Cookies: The Bill widens the scope for implementing cookies and similiar tracking technologies without the need for consent under certain conditions.
- Automated decision making: The Bill relaxes rules on automated decision making, potential allowing more flex to use in AI automated systems.
To prepare, Rebecca recommended the following:
- Review and update data policies and procedures: Ensure your practice align with the proposed changes, particularly if you use automated design making (such as AI), and update procedures around DSARs, data protection impact assessments and the use of legitimate interest as a lawful basis for processing.
- Review third party data sharing agreements and ensure they reflect the higher fines.
- Review cookies being used and ensure policies reflect the changes.
- Update training for staff, particularly around automated decision making.
Copyright and AI government consultation
A consultation on a text and data mining exemption in respect of AI development closed on 25 February 2025.
Key points:
- Proposes to expand the text and data mining exemption for AI training to all works freely available on the internet as well as any the user has paid a subscription to access.
- Rights holders would have the ability to “opt out” their rights and require a licence. The consultation sought views on how this would be accomplished in practice through technical standards.
To prepare, Rebecca advised the following:
- Keep out an eye on the introduction of the opt out rules if you or your clients are content creators.
- If you are designing an AI system, bake in the UK Voluntary Code of Practice for the Cyber Security of AI which covers the lifecycle of the development phases (design, development, maintenance and end of life).
- If you are working on AI systems for development in Europe, adhere to the EU AI Act.
Use of Generative AI in agencies: The intellectual property risks
Rebecca also covered the use of Generative AI and the risks to IP.
Risks include infringement of copyright, trademark and privacy rights when generating AI outputs. You may also inadvertedly include personal data in an input which is used to train the model or an output contains personal data which is not authorised for processing.
Other risks are hallucinations, bias, out of date information and lack of transparency.
To minimise the risks, Rebecca’s advice included:
- Create a policy which manages what AI tools can be used and how they can be used.
- Carry out internal employee training on responsible use of AI.
- Check outputs for accuracy.
- Avoid using personal data in AI tools unless you have carried out a data protection impact assessment (DPIA) and considered data protection obligations.
- Check if AI is permitted or restricted expressly in client contracts.
- Ensure client materials provided to use are licensed for use in AI applications and all rights are cleared for such use.
- Check any insurance policies and cover.
- Amend supplier and freelancer contracts to include compliance with AI policies.
- Provide training for suppliers and freelancers.
Bristol Creative Industries members can book a free 30 minute legal advice session with business lawyer Rebecca Steer. Slots are available on 23 May and 27 June.
The world of AI seems to have blasted into outer space on the Starship Enterprise warp 5 hypermode recently. I say this because a week in AI is infinitely faster than a week in politics – more AI tools, better AI, outwit your competitor AI with free features and yes even to communicate with Dolphins with DolphinGemma AI.
One of the biggest AI trends that is being experienced right now is the move from SEO (Search Engine Optimisation) for search engines such as Google and websites, to now using AI in the search for answers to questions.
Open AI is eyeing Google to buy. This is the biggest AI platform on earth considering the biggest search engine. A potential monopoly may be on the cards, but for now Google is now offering AI answers at the top of its search listings.
However these giants may tussle remains to be seen, but whoever owns the browser owns the gateway – and largely this will decide who rewrites the engagement for the next era of cognition.
Over time this is going to have huge implications for those businesses who have built their websites on SEO search and the Google algorithm ….. which is pretty much everyone.
ChatGPT has recently provided an image tool free of charge to those with an account. Merely speak to ChatGPT and ask what image you would like to see. It will transcribe your voice, and you can press the search button to see what image arrives.
The new image feature in ChatGPT, has led to a huge demand on the OpenAI platform sitting underneath it for power and becoming a “victim of its own success.” Leading to ChatGPT producing timeouts or error messages. However, the new feature provides a challenger to Midjourney who up to now has been the de facto image generator of choice.
So, for the absolute AI beginner, I have put an exercise on my website, for those people or families who want to see how AI works in ChatGPT with kind permission of WeareSpark.ai
In the exercise you will need to create a free account in ChatGPT or another AI platform with a verified email address. Once created, you can then visit www.thecreativesuniverse.com and go to the Resources section and open the “Research Prompt Library” word doc. This is where the exercise is located. Once you have located the word doc, then you can open your ChatGPT account and copy the first text prompt which begins “I am a brand strategist”… and finishes with…”for my brand.” Copy and paste all of text up to solid line break and put it into the ChatGPT prompt box. The AI will then go to work on providing you with answer to your exercise question.
If there are 2 or 3 of you that can work together using separate PC’s/browsers then you can see how similar or different the AI answers are given. If you all load the same question at the same time.
Once you have received your answers, then you can copy and paste the next set of text up to the line break, and again put this “prompt” into the query box. Again, the AI will go to work producing you with an answer.
Work through the whole of the document with the various prompts and once you have worked your way to the end, you will have pushed the AI to answer all sorts of questions regarding the made up fashion brand.
Doing the exercise in its entirety will give you a very good feel for what can be achieved by using AI and prompting in the right way.
The final text asks you to go into Dall-E, but with the new image feature in ChatGPT then this should produce an illustration for you to look at.
Additionally, users who wish to pay for the ChatGPT version can attach multiple documents for the AI to study. A paid for business account, also means that your precious documents remain private, instead of being shared in the “free for all” training data.
A new survey by the World Economic Forum says that “half of employers plan to re orient their business in response to AI, two-thirds plan to hire talent with specific AI skills, while 40% anticipate reducing their workforce where AI can automate tasks.” Read the full report here The Future of Jobs Report 2025 | World Economic Forum
AI dolls have been hitting the media recently around a new trend for making a doll of yourself see a recent BBC article ChatGPT AI action dolls: Concerns around the Barbie-like viral social trend – BBC News
We at LeonardoPower have provided a free Voice AI at https://aivoicepr.leonardopower.com if you sign up for free banking. You can receive an AI Voicebot for free – which is great for answering calls, meaning that you never miss one again. It’s ideal for anyone who is tied up “doing-the-do” and needs an extra pair of hands. It gives you a transcript of the call to look at when you have time.
Finally, for those that wanting to put AI to proper work such as reducing the marketing burden with AI tools to reduce the workload visit the resources section on www.thecreativesuniverse.com and find the AI Toolkit to learn more. To arrange a demo get in touch with me at [email protected]
The UK government’s new AI Opportunities Action Plan is designed to boost economic growth
But what does it mean for your business?
A new report published by the UK’s Department of Science, Innovation and Technology outlines 50 recommendations for the government to drive adoption of artificial intelligence (AI) across industries and boost economic growth. But what does the AI Opportunities Action Plan mean for marketers and the wider B2B industry?
Led by Matt Clifford CBE, Chair of the Advanced Research and Invention Agency (ARIA), the plan promotes three key goals for the government:
- Invest in the foundations of AI
- Push hard on cross-economy AI adoption
- Position the UK to be an AI maker, not an AI taker
The government’s response included promises to accelerate AI research and infrastructure development, promote AI Growth Zones to speed up planning, and public sector pilot schemes to help workers ‘spend less time doing admin and more time delivering the services working people rely on.’
And in the private sector, £14 billion and 13,250 jobs have been committed by leading tech firms following the AI Action Plan.
Finally, there are plans to develop and maintain ‘homegrown’ AI technologies, ensuring the UK economy benefits directly from the rapid adoption of these solutions.
“The UK Government’s AI Opportunities Action Plan will play an important role in helping the UK to unlock the full potential of AI and in doing so, boost productivity, enhance economic growth and improve public services. At AWS, we’ve seen first-hand the benefits that digital technologies like AI can bring.”
– Alison Kay, VP U.K. and Ireland at Amazon Web Services
According to the Department of Science, Innovation and Technology, these plans could boost productivity by as much as 1.5% per year. If fully realised, these gains could be worth up to an average £47bn to the UK each year over the course of a decade.
But what does this mean for UK businesses? And what opportunities should marketing teams look out for?
Embedding AI in your business – opportunities and risks
The AI Opportunities Action Plan effectively gives businesses the go-ahead to grab opportunities with both hands, embedding AI tools and investing in upskilling. If AI is to become the catalyst for the UK’s economic growth, there’s no better time to start adopting the latest technologies. The outlook is optimistic, but we always advise a cautious approach. It’s important to assess your readiness carefully before jumping on the bandwagon.
So what are the opportunities and risks of building AI into your strategy?
Boost operational efficiencies
There’s no doubt AI can support businesses to streamline processes and make smarter decisions. From automating repetitive tasks to optimising supply chains, AI reduces manual effort and streamlines workflows. For instance, customer service chatbots can handle thousands of queries simultaneously, while machine-learning algorithms improve inventory management by predicting demand with remarkable accuracy. These efficiencies save time and costs, while allow businesses to focus on other strategic priorities.
But implementing AI tools requires skill and understanding, and employees are often sceptical – or even fearful – so it’s important to ensure communication and training is prioritised.
Drive growth and performance
Across many industries, AI is already driving considerable growth. AI-powered analytics provide businesses with insights that were previously unattainable, helping them understand customer behaviour, market trends, and operational bottlenecks. Companies can use AI to develop innovative products and identify new revenue streams.
However, growth through AI isn’t automatic. It demands significant ongoing investment in talent and infrastructure, and a continuous improvement approach to keep up with technological advancements. This means managing expectations and setting a realistic timeline.
Beware the environmental impact of AI
AI technologies rely heavily on data processing, which demands significant computational power and energy. The environmental cost of training AI models, including its carbon footprint, electricity use and water consumption, cannot be overlooked. Training large-scale models like GPT or image recognition systems often consumes vast amounts of electricity, equivalent to running entire power plants.
Organisations must consider the impact of their AI initiatives, particularly when it comes to sustainability reporting. It’s also worth investigating tools with a lower carbon footprint and embracing ‘green AI’ solutions as they emerge.
Consider governance and ethics
As we embrace AI, we must be increasingly rigorous with our governance and ensure an ethical approach that fosters trust and reduces the risk of reputational damage. Companies should establish ethical guidelines and governance frameworks to oversee AI development and deployment. It’s crucial to ensure they’re using these technologies responsibly, and concerns around bias in algorithms, data privacy, and accountability must be addressed.
All adopters will need to battle scepticism, so building and maintaining trust with stakeholders and customers will be key. Watertight branding and communications will therefore be more important than ever.
AI-powered martech for B2B businesses
When we talk about AI solutions for marketers, we don’t just mean Generative AI models like ChatGPT and image tools. Marketing teams are building numerous AI tools into their tech stacks and new ones are popping up all the time. Here are some use cases we’re currently exploring:
AI-driven audience targeting and ABM strategy
AI tools can make audience profiling and targeting simple and straightforward. Building these tools into your account-based marketing process is a great way to gain efficiencies and cut down labour so you can spend more time crafting your messaging and optimising your content.
Market research and industry trends
AI tools are a great way to save time on market research. In the time it takes you to do a quick Google search, tools like Waldo can download reams of industry-specific intel – plus it can analyse it all for you and deliver a report straight to your inbox. It can also highlight key trends in your industry to help you narrow your focus and stay competitive.
Website personalisation and optimisation
AI-powered personalisation tools help you tailor digital content to your specific audience, as well as A/B testing to ensure your messaging, images, and UX design is optimised to convert.
Ready to streamline your marketing strategy? We can help you make informed decisions and choose the right tools to maximise ROI.
Get in touch today: [email protected]
saintnicks has won two awards at the prestigious Transform Awards Europe 2025 for their work with Ascot Racecourse.
Gold: Best Expression of a Brand on Social Media Channels
Bronze: Best Use of Copy Style or Tone of Voice
The Transform Awards celebrate excellence in brand strategy and execution across Europe. saintnicks’ work with Ascot Racecourse brought to life the brand’s creative platform, Elegance at Play – combining social-first storytelling, a distinct tone of voice, and thumb-stopping, jaw-dropping content that captured the attention of both loyal racegoers and new audiences alike.
Speaking on the win, Fraser Bradshaw, CEO at saintnicks, said:
“We set out to create a truly ownable brand voice and world-class social content that matched Ascot’s stature as an iconic British institution. To see that work recognised is a brilliant moment for the team and a testament to the power of brave, collaborative thinking.”
Looking to go further?
If you’re after a creative brand agency that will go the extra mile for your brand, drop saintnicks a line. You can find out more about their brand, campaigns, content and digital expertise here, or reach out to their Client Services Director, Francois d’Espagnac.
There’s a lot of debate right now about whether AI-powered search is replacing traditional search engines or if search engine usage is still growing faster than AI adoption. Either way, one thing is certain—search behaviour is evolving. It is increasingly important to ensure that your brand is optimised for Large Language Models, or LLMs for short. This can seem difficult if you have a brand language or a specific way of talking and this doesn’t match how the LLMs understand your content.
As businesses, marketers, and SEO professionals, this raises an important question: Should we still focus on traditional SEO, or shift our focus to optimising for AI models?
The answer is clear, traditional SEO is still critical. However, AI-driven search is changing how information is found, processed, and presented. Large Language Models (LLMs) now play a significant role in how your website is understood and ranked. LLMs, such as Google’s Gemini 2.0 Flash and OpenAI’s o3 mini, are quickly changing how consumers seek and receive information. These AI-driven systems interpret and generate human-like text, influencing decisions and shaping perceptions. Large Language Models (LLMs) now play a significant role in how your website is understood and ranked.
So, how can you ensure your brand’s content is optimised for both search engines and AI models? Here are seven key strategies to help you stay visible in search and maintain brand clarity across AI-driven platforms.
1. Focus on Entities
Entities are key concepts, such as brands, products, and services, that search engines and AI models use to understand content. For your brand to be correctly recognised and associated with the right expertise, you need to use your brand name consistently alongside relevant keywords. Instead of writing generic descriptions for example at Varn we could say “We offer great services,” it’s important to be clear and explicit. A stronger alternative would be: “At Varn, we offer innovative SEO services powered by data.”
By making these connections clear, search engines and AI-driven models can better associate your brand with specific topics and expertise. This increases the likelihood that AI-generated search responses will accurately reference your business.
2. Use clear and natural language
LLMs are designed to understand and generate human-like text, so your writing should be as clear and natural as possible. Overly complex or jargon-heavy content can be difficult for both AI and human readers to interpret.
When creating and writing content, imagine you are explaining your services to someone with no prior knowledge of your industry. Keep your language simple, direct, and conversational. If your subject matter is technical, take the time to explain key terms in a way that is accessible to a general audience.
By making your content easier to understand, you improve its accessibility for users while also increasing the likelihood that AI models will accurately interpret and feature your content.
3. Structure your content for AI and search
Content that is well-organised and clearly structured is easier for both search engines and AI models to process. This means using descriptive headings, subheadings, and logical formatting to guide readers and search algorithms through your page.
For example, if an AI bot encounters a section titled “Benefits of Optimising Your Brand Language for LLMs” followed by a well-structured list, it can quickly determine that the following points describe the advantages of LLM optimisation. This helps AI models extract and summarise relevant information more accurately.
Breaking up content with bullet points, numbered lists, and short paragraphs also improves readability. Both human users and AI bots can more efficiently scan and process your information, leading to better search rankings and improved user engagement.
4. Link your content logically
Think of your website as a well-organised library where every piece of content has its proper place. If your pages are connected in a logical and intuitive way, AI models and search engines will have an easier time understanding how different sections of your website relate to one another.
If your homepage links to main sections (like “Products” or “Services”) and those lead to specific sub-pages, a search engine or AI can follow that path to understand how your content is grouped. This again provides even more context to the information you are publishing, improving AIs understanding of your brand, or entity. A clear and connected website architecture not only enhances user experience but also signals to search engines that your content is well-structured and authoritative.
5. Build authority through digital PR
Authority and credibility are just as important for AI models as they are for traditional search engines. If trusted sources reference your brand or website, AI models are more likely to feature your content in their responses.
To build authority, focus on securing high-quality backlinks from reputable industry websites. Publishing guest articles, participating in expert panels, and being featured in respected publications all help establish your brand as a reliable source of information. Digital PR efforts not only improve traditional SEO rankings but also enhance your brand’s visibility in AI-generated search results.
6. Answer questions directly
AI-driven search is heavily focused on answering user queries. To improve your chances of appearing in AI-generated responses, structure your content to provide clear and direct answers to commonly asked questions.
Consider incorporating an FAQ section into your website or structuring blog posts around key industry questions. When answering these questions, be concise and informative. Well-structured, easy-to-digest responses are more likely to be surfaced by AI models when generating answers for users.
7. Create AI brand language guidelines
Just as brands create tone of voice guidelines for marketing and social media, it is now essential to establish guidelines for AI-generated content. AI models pull from existing online content to generate responses, so ensuring consistency in your brand’s language across digital platforms is key.
Define the messaging and terminology that best represents your brand, and ensure that AI-friendly content aligns with these guidelines. Regularly review AI-generated responses related to your business to identify any inconsistencies. By being intentional about your brand language in AI-driven search, you can maintain control over how your business is perceived and ensure that AI-generated content reflects your true brand identity.
Final thoughts on optimising your website copy for search engines and LLMs
The way we search for information is changing rapidly. The rise of AI-driven search means that brands need to optimise their content for both traditional search engines and LLMs. However, this doesn’t mean abandoning traditional SEO; it means evolving your strategy to align with how AI models interpret and present content.
By focusing on clear, structured content, entity-based optimisation, and AI-friendly brand language, you can improve your visibility across both traditional search results and AI-powered search platforms. As search continues to evolve, staying ahead of these trends will be critical for maintaining brand presence and ensuring your content reaches the right audience.
If you want to learn more about optimising your website for AI search, contact our team at Varn for expert guidance.
21.03.25Article by: Tom, CEO
Google has launched a new experimental AI search tool, AI Mode, in a bid to compete with the likes of ChatGPT and Perplexity AI. Blending powerful generative AI with their traditional search interface, Google’s new chatbot goes beyond the familiar ten blue links, delivering detailed answers with advanced reasoning and real-time information. In this article, we’ll explore what Google’s AI Mode is and how it differs from other AI-driven search tools. We’ll break down its key features and functionality, highlight strengths and weaknesses compared to existing tools, and discuss the potential impact on user search behavior.
What is Google’s AI Mode?
Google’s AI Mode is a new search experience (currently only available to Google One AI Premium members in the US via Search Labs) that uses Google’s latest AI model (a custom version of Gemini 2.0) to generate rich, conversational answers directly in Google search results. Instead of just showing a list of website links, AI Mode gives an AI-generated overview in response to your query, complete with relevant information gathered from multiple sources and accompanied by citations/links for reference. It is particularly designed for complex or multi-part questions that typically would require multiple searches – for example, comparing detailed options or exploring a new concept step-by-step. As with other AI powered search tools, users can ask follow-up questions in a conversational manner, allowing them to dive deeper into a particular topic within the same search session. This effectively turns search into an interactive dialogue, powered by Google’s AI and backed by Google’s vast index of information.
Google’s AI Mode uniquely combines generative AI with Google’s established information systems. It can tap into the Knowledge Graph, real-time data about current events, and even shopping data for product information. Whilst the current version of Google AI Mode available via Search Labs hasn’t shown product listings as part of any of our test searches, this is still in experimentation mode and so we will likely see many new developments over the coming weeks and months.
Key features of Google’s AI Mode
Google’s AI Mode introduces several notable features and enhancements over a standard search experience:
- Advanced reasoning for complex queries: AI Mode uses a custom version of Gemini 2.0 that excels at reasoning through complicated, multi-part questions. You can ask nuanced questions that might have previously required piecing together answers from multiple searches. For example, you could ask a detailed planning question or a comparison between technical options, and the AI will break down the problem and address each part in a structured answer.
- Conversational search with follow-ups: AI Mode supports follow-up questions and context carryover, turning search into a conversation. After getting an initial answer, you can ask a clarifying question or request more detail, and the AI will remember the context. This multi-turn conversation ability creates a more natural, interactive search experience, allowing deeper exploration of a topic.
- Integrated web links and citations: Google’s AI Mode provides source links so you can verify information or read more about the topic you are searching for. The AI-generated answers are presented in flowing text but include inline citations or a list of sources. The information is backed by verifiable content, and Google has emphasised factual accuracy – if the system isn’t confident in an answer, it will default to showing regular search results instead. This focus on factual reliability helps address concerns about AI “hallucinations” by prioritising trusted sources and showing users where the information is coming from.
- Deep integration with Google’s data ecosystem: A key advantage of AI Mode is how it leverages Google’s enormous data and knowledge base. It doesn’t rely solely on a pre-trained model’s memory; it actively pulls in fresh information from Google’s index, Knowledge Graph (for facts about entities), and even up-to-the-minute news or product info. This means answers can include very current information (something a static model might miss) and factual data like dates, figures, or product details drawn from structured Google data. By contrast, standalone AI chatbots without this integration might give outdated answers if their training data is old.
- Parallel search processing (“Query Fan-Out”): When you submit a question in AI Mode, Google’s system will often break it into sub-queries and search for each in parallel. For example, a question comparing two products might spawn separate searches about each product’s specs, user reviews, pricing, etc. The AI then combines all of those results into one answer. This parallel processing allows more breadth and depth in the response than a single traditional search could provide.
With these capabilities, Google’s AI Mode is poised to change how users interact with search, especially for in-depth inquiries. Next, let’s compare how this new mode stacks up against other AI-powered search tools available today.
Google AI Mode vs. other AI-powered search tools
Google is not the only player integrating AI into search. Competing offerings like Perplexity AI and OpenAI’s ChatGPT (among others) also provide AI-driven search or Q&A experiences. However, each takes a different approach.
Google AI Mode vs. Perplexity AI
Perplexity AI is a newer AI-powered search engine that, like Google’s AI Mode, answers questions by fetching information from the web and then summarising it with an AI model. Perplexity has gained a niche following for its clean interface and strong focus on citations. How does it differ from Google’s AI Mode?
- Independence and integration: Perplexity is an independent platform, not a general-purpose search engine with its own vast index like Google. It relies on querying the web and then uses an AI (such as GPT-3.5 or GPT-4) to formulate an answer. The key difference is integration with data systems: Google’s AI Mode benefits from Google’s internal data (knowledge graph, etc.) and infrastructure, potentially giving it a broader and deeper pool of information to draw from. Perplexity, being separate, doesn’t have a proprietary index on the scale of Google’s, so it’s limited to what it can fetch via search and any indexed sources it has.
- Real-time information: Perplexity does fetch information in real time (that’s one of its selling points – it’s not limited by a training cutoff). In practice, Google AI Mode and Perplexity both can provide up-to-date info, but Google’s integration means it can also pull from live updates (news, etc.) seamlessly. Perplexity will show you what sources it found and often includes the time or date of those sources. Google will similarly include fresh sources and even say when it’s using real-time info. Both are strong in freshness, but Google might have an edge for truly live data (e.g. Google can directly incorporate something from minutes ago if it’s indexed or in its news feed).
- User base and access: Perplexity is available to anyone for free (with some limits) and has a premium version for more advanced GPT-4 answers. Google’s AI Mode, at least in early 2025, is restricted to invited users or Google One subscribers with AI features. Over time, Google will likely roll it out more broadly.
Strengths & weaknesses: Google AI Mode’s strength against Perplexity is the combination of breadth and depth – it can answer more complex questions by drawing on more sources and using better reasoning, all integrated in one place. Perplexity’s strength is being lean and focused: it often gives very concise answers with minimal fluff and clearly shows sources, which some users (especially researchers) appreciate. However, users have to go to a separate site or app to use Perplexity, whereas Google’s AI Mode is in a place where billions of searches are already happening. Overall, Perplexity pioneered the kind of experience that Google is now building natively, but Google’s version could eclipse it by virtue of superior data integration and user convenience.
Google AI Mode vs. ChatGPT
ChatGPT, developed by OpenAI, isn’t a search engine, but it is a prominent AI tool often compared in this space because it answers questions in a conversational way. It’s important to clarify the context: ChatGPT (the default free version) does not have direct access to live web information. Still, many people use ChatGPT as an information tool, so how does Google’s AI Mode differ?
- Data sources: Google AI Mode pulls from the live web and Google’s index every time you ask a question. ChatGPT’s default knowledge, on the other hand, comes from its training data (which, as of March 2025, includes data up to around October 2023, with limited knowledge of more recent events unless using an update or browsing). This means out-of-the-box ChatGPT can’t reliably handle queries about very recent events or dynamic information (unless you’re using the paid version of course).
- Purpose and usage: ChatGPT is a general AI assistant – you can ask it to write code, draft emails, brainstorm ideas, educate you on a subject, etc., all in a conversational flow. Google’s AI Mode is narrower in purpose: it’s meant to enhance search. So while it can also handle coding questions or explanations, it doesn’t for example directly write a long essay unless that’s part of answering your query. ChatGPT often excels at creative tasks or open-ended discussions that go beyond factual Q&A. If you asked ChatGPT to write a short story or solve a puzzle, it would do so from its trained knowledge. Google’s AI Mode might not even engage with a prompt that isn’t essentially a search query. Thus, ChatGPT’s strength is its versatility and depth in pure conversation (with no requirement of citing sources), whereas Google’s AI Mode focuses on being an accurate research tool embedded in search results.
- Citation and trustworthiness: By design, ChatGPT does not provide citations for its answers, and it can sometimes “hallucinate” facts or sources, which is problematic if you need to verify information. Google’s AI Mode always ties back to sources and will avoid answering if it can’t ensure accuracy. For someone looking for an answer they can trust or use in research, AI Mode’s approach is more transparent. ChatGPT is great for quick explanations or drafting, but if a user needs to double-check facts, they have to manually ask for sources or use the browsing tool. In contrast, Google AI Mode includes the links up front, making it easier to trust (or at least verify) the response.
- Model capabilities: ChatGPT (especially GPT-4 version) is extremely powerful in reasoning and language, and in some contexts it might produce a more detailed or eloquent answer than Google’s AI Mode. However, ChatGPT’s weakness is it might not know the latest specifics or data points post its training cutoff. Google’s model in AI Mode is also highly capable and is specifically tuned for providing “high-quality responses” in search.
- Accessibility: ChatGPT is accessed via OpenAI’s website (or API) and requires an account sign-up, with the GPT-4 version paywalled under ChatGPT Plus. Google’s AI Mode, once fully launched, will be accessible to anyone on Google Search for free. That is a huge difference in potential reach. ChatGPT’s interface (the free version) is purely a chat with no extra web content, while Google’s AI Mode lives alongside the web content it’s drawing from.
Strengths & weaknesses: Google’s AI Mode is strongest where ChatGPT is weak: real-time factual queries with need for source attribution. It provides an answer you can cite or trust to be up to date. ChatGPT’s strength is in open-domain creativity and instructive dialogue – it’s often more flexible in what you can ask. For an SEO expert or researcher, Google AI Mode might be the preferred tool for gathering information with confidence in the source; ChatGPT might be what you use to brainstorm how to use that information or to generate content from it. One could imagine using both: e.g., ask Google AI Mode for the latest stats or details on a topic (with sources), then use ChatGPT to help craft a report or article around that info. Another point: ChatGPT, being model-based, sometimes injects more of a conversational filler and can occasionally deviate. Google AI Mode, guided by actual search results, is more likely to stick to the point. In summary, ChatGPT is a broad AI assistant with knowledge (albeit time-limited), whereas Google’s AI Mode is an AI-enhanced search specialist grounded in live data. Each has their place, but for the specific job of answering search queries with current info, AI Mode is built to excel.
Impact of AI Mode on search behavior
The introduction of AI Mode in Google Search has significant implications for user behaviour and how people interact with search engines:
- Fewer clicks, more instant answers: One immediate effect is a potential reduction in clicks to external websites. When the AI Mode provides a comprehensive answer on the search results page, users may feel less need to click through multiple links. For example, if someone asks a detailed question and the AI summary fully answers it, that user might never visit the sites that provided the information. This trend began with featured snippets, but AI Mode takes it to a new level by answering much more complex queries directly. For users, this can be a time-saver – they get what they need faster. For businesses however, this could lead to a drop in website traffic and fewer on-site conversions.
- Longer, more conversational queries: Users may start phrasing their searches in a more natural language and detailed way. Instead of typing a few keywords, users might pose a full question or even multiple questions at once, knowing that the AI will parse and answer them in one go. Over time, people could grow more comfortable “talking” to search like they would to a human expert. This will naturally lead to an increased number of long-tail searches, something we’re seeing throughout AI search and which should be incorporated into your SEO strategy.
- Continued need for traditional search: It’s worth noting that not every search will use AI Mode. Simpler or navigational queries (like “Facebook login” or “weather tomorrow”) might still be served best by a quick snippet or a link. Google has signalled that if the AI isn’t confident, it will fall back to regular results. Users will likely learn when AI Mode is most helpful (e.g., when answering “big” questions) versus when it’s not necessary. Also, some users might not trust the AI answer fully and will click sources to verify or see more. So while behaviour is shifting, it’s not a complete replacement of all search habits – rather, it adds a new mode for certain kinds of informational needs.
Mobile and voice implications: As search becomes more conversational, voice search is likely to become much more popular. AI Mode’s development might bleed into how Google Assistant or mobile voice queries are answered (more conversationally, with summarised info). If AI Mode makes it easier to get a direct answer, people might be more inclined to ask their phones a question out loud and trust the spoken response.
Google AI Mode: summary
In summary, AI Mode is changing search behavior by making search more of a dialogue and less of a directory. It is important that we place additional focus on conversational search within SEO, and that we optimise content for voice search, long-tail keywords, and individual entities – but we need to do so whilst making sure we don’t ignore traditional search. Google may have seen a large drop in their market share in recent months thanks to the introduction of other AI powered search tools, but they may just start pulling that traffic back thanks to the launch of AI Mode. We’ll keep an eye on these developments, and will let you know when Google AI Mode is ready for the general public.
In the meantime, if you’re concerned about search performance in this new era of AI and would like to make sure your website is optimised for AI search, give us a call – we would love to hear from you.
21.03.25Article by: Aimee, Head of Data & Innovation
Voice search has actually been around since 2008 when Google first introduced voice search on its mobile app for iphones, and has since continued to grow in use. Voice searches can be made from a range of devices including virtual assistants (e.g., Siri & Cortana), smart speakers (e.g., Google Nest & Amazon Alexa), and smartphones. Factoring in voice search optimisation into your SEO strategy has especially come to light recently with developments in AI search and featured snippets, as these typically appear for voice searches. Our blog will give you a rundown of how voice search works, how and why to optimise for it, so you can adapt your SEO strategy for a range of search mediums,, and how to track performance.
How does voice search differ from traditional search?
Although traditional searches have become longer and use natural language, voice searches tend to be more conversational, as if users were having a real-life discussion. For example, a user may type “Top 5 holiday destinations 2025” but verbally ask “What are the top 5 holiday destinations in 2025?”. Voice search queries tend to be informational, especially if a user makes a request via a smart speaker, or commercial if using another voice-assisted device such as a mobile phone. In comparison to traditional searches, voice searches produce more clear and concise results, often with AI Overviews and featured snippets. It’s also worth noting that a large portion of voice searches come from users who are on the move, particularly when using a mobile device, to get quick answers.
Voice search works through an automatic speech recognition (ASR) system, which translates speech to text as follows:
- An individual uses their voice to make a search such as “What is the best recipe for Mother’s Day?”
- AI will translate this speech into text, which search engines are able to process
- AI identifies a user’s search intent using natural language processing (NLP)
- Then, depending on the device a user is searching from, the answer is either verbally given back to the user, or provided in the form of typical search results including an AI Overview and SERP features.
What are the benefits of optimising my website for voice search?
One difficulty with voice search is that some people using smart speakers for voice search will be unlikely to investigate what source was used. For example, a user asking an Alexa “What is the best material to make a knitted jumper?” will not be able to click through to a website selling wool and be converted into a potential customer. So you may be asking the question, if appearing in voice search results may not lead to increased traffic on my website, why do I need to optimise for it?
Firstly, voice search covers a wide range of devices, not just Alexas or Echos, and secondly, the steps you would take to optimise your website for voice search are also helping to optimise your website for traditional searches. So improving your website’s SEO for voice search will likely come with many benefits such as appearing in SERP features and AI Overviews, and ranking well organically. Voice search is about capturing these users and drawing them into your website, as they may become customers, so it’s great to include this in your overall SEO strategy.
How to optimise your website for voice search.
Because questions used in typically written searches are becoming longer and more conversational, similar to voice searches, optimising the below elements for voice search essentially allows you to hit two birds with one stone. Matching search intent, including relevant keywords, adding schema where applicable, and optimising for local search and mobile will have positive effects on your rankings. Here are a few ways you can optimise your website’s SEO for voice search, whilst helping to enhance its general SEO performance.
Keywords and search intent
Keyword research enables you to understand your target audience and tailor your website’s content. You can find keyword opportunities through resources such as Google Search Console, Google Ads, or Google Analytics, as well as third-party tools such as SEMRush, Ahrefs, and Moz. It’s important to include long-tail keywords, conversational phrases, and semantic keywords to cover a range of relevant queries your audience uses to find your website’s products and services.
To strengthen your content strategy, match search intent to your content. If a user is searching for “the best Greek salad recipe”, they will likely be looking for recipe SERP features with the most appealing name, enticing description, and attractive image. The above query is an informational one, so if your business sourced fresh fruit and vegetables, you could add a recipe under blogs or a designated recipes folder. Along with this content, you should also include recipe schema, adding relevant data such as a title, description, optimised image, and reviews, to increase the likelihood of appearing in recipe SERP features.
FAQs and Schema
As SERP features and AI Overviews are now in search results, the typical organic blue links that make up the top 10 results get pushed down the page. But as the screen on a mobile device is much smaller, users are likely to focus more on these top features in voice search results as opposed to website URLs. Therefore, it’s important to optimise your content, following a question-answer style while adding FAQs to pages where relevant. Alongside keyword research, you can see which questions users are searching by looking at the ‘People Also Ask’ SERP feature and looking at commonly asked questions in customer feedback.
Marking up your website’s content with a range of schema types will increase the likelihood of appearing in SERP features. Here, user-generated content (UGC) such as reviews and testimonials are great additions to your pages and schema as these make your site more credible, a factor that influences web rankings.
Optimise for local search
As highlighted earlier in our blog, a large portion of voice searches made from a mobile device are by users who are on the move. These users are often making local searches, such as “What’s the nearest supermarket near me?” or “Where can I get my phone fixed?”. Therefore, to optimise your website for local SEO, make sure that your Google My Business profile is up to date, particularly if you have a physical store. Including location-specific landing pages on your website is also beneficial as these can rank for location-specific keywords in SERPs.
Mobile optimisation
Due to Google’s mobile-first indexing approach, you need to make sure your website is up to scratch on mobile. Not only does a fast mobile site speed matter in Google’s eyes, but people using voice search, particularly when on the move, are looking for quick answers and quick buys. Users want to be able to find the information they are looking for easily on your site. Having a smooth, functional, and logical navigation on your mobile site plays a part in conversions. If a site is too slow, elements are unresponsive, or a user cannot find what they are looking for, they may simply leave the site.
How to track your website’s performance in voice search.
Although you aren’t able to directly see your website’s appearance in voice search results, there are various elements you can track to get an understanding:
- Featured snippets: as these features appear in voice searches, this can be a key indicator of your website’s visibility for voice search.
- Advanced Web Rankings and other search tracking tools: these tools allow you to see what keywords generate featured snippets, thus providing key insights into how you can optimise your content for voice search.
- Link Google Search Console to Google Analytics (GA4): as longtail keywords are used in voice search on mobile devices, this can give a good indication as to the searches being used and what pages of your site voice searchers are landing on.
Key takeaways
Appearing in voice search results isn’t an easy one-route path; Following a range of SEO tactics including question-answer style content, long-tail keywords and phrases, and schema markup as well as optimising your website for mobile and local search will improve your website’s performance in voice search results.
Want to learn more about how AI is changing search? Check out our AI Search and Innovation blogs, or contact one of our SEO experts today.
17.03.25Article by: Georgina, Future Talent Graduate
By consistently providing valuable content, companies can build trust and credibility with their audience. This trust not only helps in retaining existing customers but also attracts new ones.
High-quality content can significantly boost your SERP visibility, making it easier for potential customers to find you. Also, by giving your audience valuable content you increase your usefulness to them
If you’re stuck on the notion of content marketing, this is how it’s done.
1. Align cross-channel messaging
Consistency is key to building a strong brand identity. Ensure that your PR efforts and social media content marketing initiatives are singing from the same hymn sheet.
Start by developing a comprehensive brand messaging guide that outlines key messages, tone of voice, and brand values.
Then hold cross-team messaging sessions and utilise social listening tools to ascertain the kinds of messages you need to develop & ensure these messages are aligned across all your outputs. Consistent messaging reinforces your brand identity and helps avoid confusion among your audience.
2. Develop thought leadership content
Position your executives and subject matter experts as industry leaders through a well-defined content strategy and strategic content creation and placement.
Identify key topics and trends in your industry where your organization can provide unique insights. Then create a content calendar that includes opportunities for thought leadership pieces, such as guest articles, speaking engagements, and webinars.
You could have your CEO write a series of LinkedIn articles on industry trends, which can then be pitched to relevant publications as op-eds. Thought leadership content like this enhances credibility, builds trust, and can lead to valuable media opportunities.
3. Leverage earned media coverage in marketing
Don’t let positive press mentions gather dust – incorporate them into your content marketing efforts to maximise their impact.
Create a system for tracking and cataloguing media mentions and awards. Develop a series of content pieces that highlight recent press coverage, such as “In the News” blog posts or social media highlights.
You could create an “As Featured In” section on your website homepage, showcasing logos of publications where your company has been mentioned.
Third-party validation, from respected media outlets, can significantly boost your credibility and persuasive power.
You could also seek to leverage user-generated content, like customer-created media, can complement media coverage by enhancing engagement and expanding your brand’s reach through authentic contributions.
4. Coordinate strategies
Social media marketing is a critical component of coordinating social media strategies for both PR and content marketing. Ensure your efforts are coordinated for maximum impact.
Develop a unified social media calendar that incorporates both PR and content marketing initiatives.
Start by using social listening tools to identify trending topics and conversations where your brand can contribute meaningfully. Create an alignment between the proactive and reactive to ensure you’re always part of the conversation.
When developing a PR campaign, plot out your key campaign moments and creative cross channel activation plans combining content and media. But don’t just leave it to those pre-planned moments, plan for contstant engagement.
A coordinated media and social media approach ensures consistent messaging, regularly engagement which helps amplify your reach across different audience segments.
5. Create data-driven content
Original research and data can fuel both PR pitches and compelling content pieces. Consumers prefer learning about products through articles, highlighting the effectiveness of data-driven content.
Identify gaps in industry knowledge that your organization is uniquely positioned to fill. Start by conducting regular surveys or data analysis projects that can generate newsworthy insights. Then leverage these insights in the form of news and content, solely built around your own proprietary data. Don’t stop at written content, think of new and engaging content formats to spin out your findings.
Original data sets you apart as a thought leader and provides valuable, exclusive content for media outlets.
6. Repurpose content across platforms
Make your content work harder by adapting it for different channels and formats.
For each piece of content, create a plan for how it can be repurposed across multiple platforms, including video content. Tactics here vary based on the desired requirements, but where possible look to breathe new life into content in formats beyond that of their original form.
You can turn blog posts into social media snippets or longer-form videos and find ways to break up longer videos into smaller chunks that can be activated as previews or teasers. Repurposing content ensures consistency in messaging while maximizing the return on your content investment.
7. Build a unified content calendar
Planning PR activities and content marketing initiatives together ensures alignment and maximizes impact.
Create a master calendar that includes all PR events, content releases, and marketing campaigns. Optimizing content for search engines like Google and Bing should be a key part of this planning to increase web traffic and achieve content marketing goals.
Look at aligning all your critical moments and messages into one unified activity calendar. Timing media releases and news announcements with social content for maximum effect.
When planning a product launch, coordinate press releases, blog posts, social media campaigns, and email marketing to create a cohesive narrative. A unified calendar prevents conflicts, identifies synergies, and ensures a steady stream of coordinated content.
8. Collaborate on goal-setting
Establish shared objectives between PR and content teams to work towards common targets, considering the unique characteristics of each social media platform. Hold joint planning sessions to identify overarching business goals and how each team can contribute.
Look to develop shared KPIs that reflect both PR and content marketing objectives. Set a joint goal of increasing website traffic from earned media mentions by 20% over the next quarter. Becasue shared goals foster collaboration and ensure that all efforts are aligned with broader business objectives.
Integrate metrics and measurement
Implement a comprehensive framework to track the impact of both PR and content efforts.
Develop a dashboard that incorporates key metrics from both PR and content marketing activities and use tools that can track the customer journey, from initial PR touch points through to content engagement and conversion.
You could track how a press release drives traffic to a landing page, and how that traffic then engages with your content and converts. For example; content marketing examples such as blog posts, case studies and whitepapers can be used to illustrate how metrics like engagement rates, lead generation, and conversion rates highlight the success of your content marketing efforts.
Integrated measurement provides a more complete picture of your marketing efforts’ impact and helps identify areas for improvement.
Foster cross-departmental collaboration
Break down silos between PR, marketing, and digital teams to create a truly integrated approach by developing a unified content strategy.
Implement regular cross-team meetings and collaborative projects and use collaboration tools that allow for easy sharing of ideas, content, and feedback across departments.
Create mixed-team task forces for campaigns, ensuring representation from all aspects of your business. Becasue cross-departmental collaboration leads to more innovative ideas, better resource allocation, and a more cohesive brand presence.
Conclusion
Integrating PR and content marketing strategies is no longer just a nice-to-have – it’s essential for creating a powerful and cohesive brand presence.
By aligning messaging, leveraging each other’s strengths, and fostering collaboration, PR and content teams can create a synergistic approach that amplifies their impact and drives better business results.
As the lines between these disciplines continue to blur, organisations that master this integration will be well-positioned to build stronger relationships with their audiences and achieve their communication goals more effectively.
By following these steps, businesses can create a solid content marketing strategy that drives profitable customer action and helps them achieve their marketing goals.
A successful content marketing strategy not only enhances brand visibility but also fosters deeper connections with your audience, ultimately leading to sustained business growth.
SEO is an evolving discipline, and as we enter what very much feels like the next stage of search (and decide on what new acronym to use to summarise it!) It’s important to reflect on what is working well but also what may be needed to futureproof your strategy. That’s important across various disciplines, but especially in SEO.
Here we take a look at backlinks, brand mentions, and off-page SEO within the context of AEO optimisation for platforms like SearchGPT. Expect to learn how the citations used by SearchGPT differs from Google, why you need to optimise for your brand name in addition to your content and pages, and how to start getting coverage in the right areas for your business.
What backlinks are being used by SearchGPT?

When looking at the kind of backlinks it is important to think about the context. Looking at an example for Varn, we conducted a search on the ‘best off-page SEO agencies’ and it returned the below result. Content pulls through from Varn’s Off-page SEO page but the sources section at the bottom is what we are really interested in.
Examining the content we can see that whilst some commercial pages are shown including Varn’s there is a high number of long-form content pages listing out SEO agencies. These kinds of lists are becoming more common across the web and appear to be what SearchGPT is prioritising when it comes to citations. If your website is mentioned in this kind of content, you are more likely to appear in SearchGPT for relevant results. Varn for example are featured in the Social Shepherd article but not in similar referral style pages from Omnius or Designrush.
These represent golden opportunities to reach out and ask for a placement. In addition to this we are also seeing more reddit results be leveraged within the content, whether this will stay or not is difficult to determine, but for this thread around the best SEO agencies in the UK, getting Varn featured is also a bonus. The question is, will not having a backlink in this kind of content have an impact?
Are brand mentions becoming more important?
Brand mentions have been a part of SEO for over a decade but their potential importance has often been debated. In the early days pre-2012, their importance was minimal, with a pure focus on backlinks. Following the Penguin and Hummingbird updates, the algorithm developed in terms of context and entity recognition, and this coincided with the Google patent on ‘implied links’ , a hint that unlinked brand mentions could be a ranking signal.
From 2015-1018, brand mentions became more important and started to play a significant role in signalling authority if they were placed on authoritative sites. This trend has continued up to 2024, with the added dimension of LLMs now clearly leveraging brand mentions in the answers they surface to users. The areas of the mentions may also play a part, as Neil Patel discusses:
“The more mentions, the more likely AI models are to consider your content high quality.
That could be things like user-generated content reviews, comments, and forum posts. These are the types of information large language models are likely to use, so use social listening tools to discover these comments.”
What does SearchGPT say?
Our question: Do you use brand mentions on forums, social posts and wider non-traditional SEO sources in your algorithm when serving answers on brands to users?
SearchGPT: I don’t have my own algorithm for ranking brands, but I can analyze and synthesize information based on a mix of sources, including brand mentions on forums, social media, news articles, and other online discussions. However, my responses prioritize authoritative sources like official websites, structured SEO-optimized content, and well-cited information.
Are unlinked mentions still a tactic worth pursuing? Google’s grip on search has not vanished just yet, but with their dip below 90% of the market share for the first time since 2015, ensuring your time is spent on the most valuable tactics will be crucial.
How to start earning citations to support AEO
So we’ve discussed the kind of citations that SearchGPT uses for content around ‘best X service providers’ and we have explored the added importance of brand mentions on these types of pages and wider sources on the web. How do you actually go about driving these for your business?
- Contribute to industry blogs: Write insightful articles for reputable industry blogs and publications. This not only showcases your expertise but also increases the chances of your brand being mentioned and linked. This was great for links, and it’s still great for building brands.
- Media outreach: Build relationships with journalists and influencers to feature your brand in articles, interviews, or reviews.
- Create shareable content: Develop engaging content tailored for platforms like Facebook, Twitter, LinkedIn, and Instagram. Encourage sharing by crafting content that resonates with your audience. This will help drive visibility and also increase the chances of content getting picked up on blogs and news sites.
- Earn citation placements: Reach out to the websites that reference your competitors and appear as citations within SearchGPT and other LLMs, this is the content that these tools are already using when serving content, so getting placements will be highly valuable.
Key takeaways
Brand mentions and off-page SEO is here to stay with the new AEO paradigm, if you want to continue to earn visibility remember to:
- Monitor and research the kind of citations LLMs deem valued in your industry
- Be conscious of brand mentions on the web and develop a strategy to drive more
- Work with an agency that prioritises future-proof outreach to drive visibility
If you would like to discuss how LLMs serve content to users and how you can get featured in commercially valuable searches no matter the platform, get in touch with Varn today.
Historically, Schema Markup has been used to target rich results in SERPs. However, we have recently seen a shift in the SERP landscape, with AI overviews now taking up valuable real estate at the top of SERPs and lowering the priority of rich results. The question now is – how can we optimise content to be best understood by AI, so this can be featured not only in the AI overview space but also in AI search engines?
What is schema markup?
Schema is a machine-readable markup language that can be added to a page’s HTML and is used to define what is on a page for better comprehension by search engines [5]. There are various types of Schema (e.g. Product, Organization, ImageObject) – with a full list available on schema.org/ – but it is important to note only some have associated rich results, and that these are not guaranteed even with schema markup added to your page. However, don’t let this put you off from adding other types of schema to your site – the more information you can markup about your content the better!
How to add schema markup to your site
Schema can be added directly to page HTML, or often there are options to add this directly through your CMS. Plugins such as Yoast can also be used to generate Schema – this code can often be taken as a starting point to create the more interconnected graphs we will touch on later in this article.
When writing Schema, we recommend making good use of tools like Google’s Structured Data Markup Helper or the JSON-LD Playground to validate code as you write it. Google’s Rich Result Validator will show if your Schema markup (either through a page URL or code snippet) is correct and eligible for rich results, whilst Schema.org’s Validator will show if a page’s Schema implementation or code snippet is correct and error-free.
Why is schema markup important for AI search?
Schema in the context of Semantic (Entity) SEO
To understand how to optimise for AI search, we must first understand the shift from lexical to semantic search algorithms. Lexical algorithms rely solely on keyword matching, whilst semantic algorithms focus on comprehending natural language, alongside the meaning and intent behind a query, to give answers that go beyond surface-level associations [2]. Semantic search aims to deliver more relevant, helpful, and tailored results for users, resulting in a better user experience and more intuitive search behaviours.
Semantic search structures and understands content through modelling this into entities, their properties, and the relationships between them [3]. Here, an entity refers to a specific concept in the real world – a person, place, organisation, idea, etc. Google combines all these components together into a knowledge graph – a graphical representation of entities, their attributes and the links between them.
AI platforms such as ChatGPT and Gemini rely on semantic search and schema markup to interpret and process information, as this allows them to extract information and comprehend content much faster and with fewer computational resources [6]. Therefore, optimising your Schema markup for semantic search will also help optimise for AI.
So how can I best use schema for semantic / AI search?
Essentially, accurate and thorough Schema markup presents semantic value to search engines – it adds greater contextual meaning and defines relationships between entities on a site [1]. Adding Schema markup provides this information to search engines in a quick, clear and easily understood way, meaning search engines do not have to process and infer this information themselves.
Historically, Schema has not been used as extensively as it could to define the relationships between entities, as adding independent Schema types was enough to go after rich results. However, when looking to optimise for semantic search, the focus has shifted less from defining entities themselves to defining the relationships between them, focusing on building up Google’s knowledge graph for your content to ensure this is properly and thoroughly understood.
In real terms – we want our Schema Markup to be one connected graph, rather than a series of separate blocks of code so search engines (and AI search!) can best understand these relationships [4].
For example, you may have:
- NewsArticle
- WebPage
- WebSite
- Organization
As 4 separate blocks of code on your site, this may be sufficient to try and gain a rich result, but this does not define the relationships between any of these entities.
What we would instead want is:
We (and search engines) can now clearly see how these entities are linked in one graph, helping convey greater meaning and optimising for semantic search.
Schema as an SEO strategy