AI, Search, SEO and Analytics
It has been a growing problem for websites in recent years that Google Search (the dominant search engine world-wide) has provided more and more "zero click" responses. For example with Featured Snippets, it shows a whole paragraph extract from an article that may contain enough information for the user - so they do not have to click through to the article itself.
The worry for websites with quality content and high search traffic is that AI will use that quality content for its responses while reducing search traffic to the website. Another complication may be the use of AI to generate website content that will compete with yours - this is already happening in areas that can be monetised.
I'm aware that this article could well be out of date within a month given how fast the tech world is moving. Right now we're in the sweet spot of "AI with everything" where every platform feels like it has to unveil its own AI functionality. Do all these platforms need AI? Of course not and it is likely that many enhancements will be quietly forgotten soon. Do WhatsApp users really need Meta AI at their fingertips? Of course not.
Google Search and AI
Right now in New Zealand, Google has scaled back its incorporation of AI into Google Search after its experience of launching it into the United States. It was initially promoting posts from Reddit and Quora as answers to your questions but quickly revised that after well publicised results (the glue on pizza story) and citations from those platforms fell dramatically in June (Search Engine Land article). It is fairly certain that AI will rise again for Google Search and be implemented worldwide - but will likely be limited to certain types of queries.
"how old is michael caine" = no AI
"what is the best digital camera for beginners" = AI response
How will ranking your content-rich website work in the age of AI? It is fairly early days for search engines incorporating AI so any recent changes in your rankings are more likely due to adjustments to the Google algorithm than due to the impact of AI.
Microsoft’s Bing Search already has its AI tool Copilot embedded - so let’s compare a current NZ search to see the NZ responses for both Google Search and Bing.
When to Plant Kūmara in New Zealand
Kūmara is a sweet potato that is very popular in NZ and I do want to grow kūmara this spring and haven't tried it before.
The images below show the same desktop search for Google Search and Bing.
Bing is already pushing a larger depth of information than Google, using the width of the screen on desktop to do it. Google has abandoned the desktop width ever since it pushed its Ads to the top of the screen (my article from 2016). Bing's mobile view contains similar information to the desktop views and uses a swipe to bring in the multiple sources.
On desktop, Bing offers its AI offering Copilot on the second tab. Clicking on it will input your search and respond. It is drawing the content from more than one source and labelling which source it is using - that is good for the user in that they can easily click through to the full article using the numbered links - and the sources are also listed by domain below the response.
So how do content rich websites compete with AI Search?
It is likely that users will not click through to your content even though the content is forming part of an AI response. It will become increasingly important to keep users on your website and one method is to offer a better search experience within the website. Whereas traditional website search just generates a list of pages with the most relevant at the top of the list, using Large Language Models (LLM) your website can answer questions and surface all of the relevant content that already exists on your website. Because only your own content is used, you can be confident in the information that will be displayed.
Use GA4 to report on the AI traffic you receive
Finally, how do you measure if you are getting traffic from these AI platforms (not just AI in Search but also ChatGPT, Perplexity etc). One method is to set up a new Channel group in Google Analytics 4 and add an "AI platform" channel. The definition will need to be constantly refined as the market develops but the image below is a good starting point as of July 2024.
Running a report of the Landing Pages for this channel will let you know which content is being offered by AI and causing a user to click through.
