First Published: 11 September, 2026
In Part 1, Marketing in the AI Era, Sophie Neate explored how AI is reshaping marketing. This follow-up article focuses on what marketers can do about it; with three practical GEO actions you can start today.
We recently applied these techniques to two established ecommerce categories. This involved restructuring content around questions customers were asking our Sales team and improving discoverability. Both Categories subsequently experienced strong growth in both organic visibility and traffic, alongside improved visibility in AI-powered search. While many factors influence performance, the results reinforced the value of focusing on customer intent and answer-led content.
1. Understand customer intent with query fan-out
AI search platforms may turn one question into multiple related queries, a process called query fan-out. Ahrefs highlights that these queries include re-phrasings of the same question, related topics, comparisons, and personalisation to the user’s behaviour or location. The result is a more comprehensive and relevant answer with follow up questions for the user. For Marketers, this means visibility is becoming increasingly about being relevant across a wider topic, not just ranking for individual keywords
A great place to start is actual questions that your customers ask via the sales team, your contact centre, and your website. This is real customer intent and will give you a solid foundation. For example, a customer question about “What’s the best mountain bike under $2000?” might trigger related queries about your height, frequency of use, the type of terrain you want to ride, and comparisons with different bikes.
The next step is identifying the fan-out queries against these questions. There are a lot of tools that can help identifying these and with understanding your AI visibility, which is best depends on your exact needs. A few that I have tried, include Otterly (better for monitoring), SemOne and Microsoft Clarity (although MS Clarity is behaviour analytics tool it offers some good AI insights – and it’s free). Ahref’s Brand Radar is also well worth a look.
A common mistake is creating content for every fan-out query. Instead, look for repeated themes and use these to improve existing content or build topic clusters around real customer needs. When creating this answer led content, don’t limit yourself to text only. You-tube video Q&A’s can be very powerful, but make sure you include a transcript, set up chapters (where relevant) and ensure the video title matches the question answered.
Creating the right content is only part of the job. Search platforms and LLMs also need to discover, crawl and understand the content.
2. Improve content discoverability
Crawlability and indexing are still important foundations. If your most important content isn’t indexed, then AI search platforms might overlook your content, irrespective of how good it is.
Search engines don’t crawl every URL on your site. They basically allocate a crawl budget per site. Crawl budget is more of a concern for larger sites or ones that frequently change. The amount that is crawled is determined by how easy the site is to crawl, the popularity and frequency a URL is updated.
To maximise the crawl efficiency of your site and help make your topic cluster be discoverable, make sure you pay attention to the following:
Not all LLMs access information from the same sources. While Google remains important, Bing is also relevant. If your target audience use Microsoft products, Bing becomes more important. This is because Microsoft Copilot uses Bing’s results. Does Bing influence ChatGPT? There is a lot of conflicting information on this one. Microsoft is an OpenAI partner, in some scenarios there is evidence that Bing influences ChatGPT, particularly for local. However, the extent of Bing’s influence is still unclear. At this stage, I believe it’s worth paying attention to your Bing ranking.
3. Help search engines understand your content
Schema is structured data that is added to the site’s HTML to support search engines understanding your content. There are a range of different schema. The value of each schema will depend upon your business, for example:
These are just three techniques, implemented together, they can help improve your brand’s chances of being discoverable and recommendable by AI search platforms. Before you implement anything though, I recommend testing to see how you perform across different platforms against your most important customer questions and tracking your performance before and after implementing these actions. To help form an accurate picture, make sure that you are in incognito mode, run multiple repetitions (same question/prompt 3 times). It’s also important to run the same exercise logged into the AI platform and logged out.
There are of course other techniques to consider, such as building consistent brand signals across third-party sites like directories and encouraging positive customer reviews. A simple way to get started is by optimising your Google Business Profile.
Each platform and customer market are different, and LLMs continue to evolve rapidly. But the fundamentals of marketing remain the same: understand your audience, create content that meets their needs, and continuously test, learn and adapt.
About the author :
Gareth Lathey is a Digital Marketing Manager at OfficeMax New Zealand, with more than 15 years' experience across marketing, digital marketing, ecommerce, search and customer experience. He is passionate about translating organisational objectives into great customer experiences and commercial results and has a particular interest in how AI is changing the way customers discover, evaluate and engage with brands online.
Category
.svg "MA-Logo (1)")
