In January 2024, I published the first wave of our SGE research focused on eCommerce terms looking at the impact on SEO of Google's new AI-powered Search Generative Results.
Today, I'm releasing our second wave of SGE research which is focused on the potential impact of SGE on the prominence of brand related terms.
Again, I've shared our main findings, our methodology and our data - and as always, I'm interested in your thoughts.
TL;DR:
Brands are not immune. These new types of generative results introduce more opportunities for third-party sites and even competitors to rank for your brand terms and related brand and product terms that you care about.
If Google decides to rollout the current incarnation of its Search Generative Results, then you can expect to see some erosion of current traffic levels from these type of terms.
I cannot think of an SEO that has undertaken more research and experiments into search for brands than Jason Barnard (disclosure he's a customer), so I asked him for his take on the advent of generative-AI in the SERPs.
"The only way to stay relevant in the new digital age is by investing in brand presence online. Google's Search Generative Experience (SGE), ChatGPT, and Copilot will completely take over search results because they provide a quick and easy user experience. Website traffic will fall dramatically because users get their answers in the AI results. To stay visible and front of mind for potential customers, brands must take steps, now, to appear in the generative AI results."
Jason Barnard (The Brand SERP Guy)
This time I curated a database of 251 well-known US brands across 15 industry verticals. The research data is here if you are interested.
I then expanded this list to 2,900 keywords in total by creating categories for different types of keyword across each vertical.
I manually curated the list to try and ensure that I had good coverage across all verticals and types of keyword. If there are any mistakes in the classification of the keywords, then it is entirely my fault!
Whilst, 2,900 gives me a reasonable sample size to analyse, it may not be truly representative of every single category and I didn't feel I had enough data to undertake cross analysis of Tags and Categories. So, please bear this in mind when considering the results - although from eyeballing hundreds of these SERP results, you can definitely get a feel for the patterns of SERP results which are returned for similar types of queries.
When, SGE finally rolls out to all users whether logged in or out, you can of course expect a much bigger series of studies.
If you really want to get a detailed perspective on this for your brand(s), then sign-up for the SGE Rank Tracker waitlist and talk to us about doing a private study for you, for the brand and money terms that you really care about.
Tags:
Brand (251 keywords), e.g. Apple, Ford, Chase.
Brand + Generic Term (1,002 keywords), e.g. Chase credit card, Ford dealer.
Brand + Product or Service (398 keywords), e.g. Apple MacBook Pro, HP Pavilion.
Generic Term (1,000 keywords), e.g. Electric cars, antivirus.
Product or Service (249 Keywords), e.g. 3 series, iPad Pro.
Industry Categories:
Automotive (188 Keywords)
Cyber Safety (190 Keywords)
eCommerce & Retail (234 Keywords)
Entertainment (145 Keywords)
Fashion & Apparel (179 Keywords)
Finance (218 Keywords)
Food & Beverage (211 Keywords)
Health & Wellness (266 Keywords)
Insurance (100 Keywords)
Legal (166 Keywords)
Office Supplies (119 Keywords)
Professional Services (104 Keywords)
Technology & Electronics (383 Keywords)
Telecommunications (96 Keywords)
Travel & Hospitality (301 Keywords)
I’ve shared some of the before and after SGE SERP and ranking data that we’ve produced as part of this research. This collection of spreadsheets shows you the full SERP data, rankings for Wikipedia and summary stats of the kind of domains that look like they will perform well in SGE.
SGE Brand & Product Keywords Researched - This is the file of keywords we researched. 15 verticals, 2,900 keyword terms in each.
Top Performing Generative Domains in SGE - This counts the number of ranking organic and generative URLs appearing for each domain, so you can see who's going to benefit from SGE's rollout.
Top Performing Generative Domains by Category in SGE -This shows you which domains had the greatest number of ranking pages by market niche.
Top Performing Generative Domains by Tag in SGE - This shows you which domains had the greatest number of ranking pages by type of keyword.
SGE Impact on No.1 Organic Results - This takes every top ranking organic URL across the whole dataset and looks at how it moves when a user clicks on the SGE Trigger.
SGE Impact on Featured Snippets - This shows the Featured Snippets in the SERPs before and after clicking on SGE.
Universal SERPs (Non-Paid Only) - This shows the non-paid ranking URLs on the first page of the SERPs before and after clicking on SGE.
Organic : Generative Match Results - This shows you the top organic and top generative URLs for every keyword. We use this to work out whether Google is using the top organic results to populate the generative results or not. This data is not only fascinating, it's going to be really helpful in building an optimisation framework for SGE results. Check out the section on our new Generative to Organic Alignment Score ("GOA Score") below.
If you believe having this kind of data would be useful for the keywords you care about then join our SGE-enabled rank tracker waitlist.
Sign-up to an Authoritas platform package and you’re guaranteed to be the first wave of SEOs to get a full in-depth impact analysis for your site(s).
We’re ready to automate this analysis for you and more, as soon as Google rolls SGE out to non-logged in users.
I’ve also shared an automated SGE slide deck that can be produced on demand for your SEO team for your keywords.
The presentation 'Authoritas - SGE Research - Wave 2 - Brand & Product Terms' contains all the images in the blog post (+ more bonus slides for the real chart geeks).
Hopefully, you'll find it handy if you have to do a presentation on the likely impact of SGE and want to justify action and budget sooner rather than later.
If you are wondering whether SGE results will affect your website then the answer is 'Yes'.
In our study, 91.4% of all keywords we checked had an SGE result, only 8.6% of keywords did not have an AI generated response.
(Total number of rows: 2,900).
This was slightly higher than our previous study which recorded this at 86.8%.
This chart shows high SGE penetration across all verticals, with the exception of YMYL (Your Money Your Life) categories like Insurance, Finance and Health. Although I would expect these categories to catch-up once Google has more experience and data points to judge how successful its LLM is at answering questions in these niches.
This surprised me a little. My previous research study showed 'Product and Service' terms to have a high SGE penetration percentage, but I didn't expect 'Brand' to be as high and higher than 'Generic' terms.
But the differences are not great, and my take on this is that eventually AI generated responses will be the norm not the exception across pretty much all types of keywords and categories of search intent.
The data painted a similar picture to our previous research, but shows a much higher incidence for the "generate button".
The "generate_button" type is more prevalent than the "show_more" type, appearing for 81.4% and 18.6% of all SGE responses, this compares to 65.9% and 34.1% in our previous study.
I've said previously that these different SGE types will have different click-thru rates and impact on your site traffic from organic - but I fear we'll be in the dark on this for some time to come.
In manual testing, I am also seeing a greater number of auto-generated SGE responses but I am not seeing this in automated testing. I'm speculating that this is a 'feature' of the SGE SERPs and the more a user clicks on an SGE trigger in a session, the more often auto-generated SGE results appear.
It's interesting to see that some categories, including 'Health & Wellness' and 'Food & Beverage', there's a greater propensity to show a snippet of content and a "Show More" link. Whilst, not mind-blowing - this does tell me that it will be well worth segmenting data by market niche when analysing data to build new click-thru rate algorithms for SGE.
Our previous SGE ecommerce study showed that the number 1 organic result drops down the page around 1,500 pixels, or ~1.5 times your visible viewport on desktop, when a user clicks on the SGE trigger.
It also showed that this was true whether the SGE trigger was a "Generate Button" or "Show More" link.
This painted a pretty scary picture for eCommerce brands who've spent years and millions of dollars optimising eCommerce category and product landing pages.
But we still hadn't answered whether this kind of drop will be seen across all types of keyword and industries or whether eCommerce sites are going to be the hardest hit.
The answer we have today is all industries will be impacted across all types of keyword.
The mean drop in pixels across this set of keywords was 1,243. This was less than the previous research, but certainly in-line with it. If you look at the breakdown by category below, you can see that the eCommerce & Retail vertical dropped by circa 1,500 pixels and other product/ecommerce focused verticals dropped by a similar amount or more.
Here's an example of a random sample of keywords from the research data which illustrates what's going on here.
As a reminder, when I talk about vertical drops I use either of these definitions:
AVG Vertical Drop (Pixels) - This measures how far down the page the No.1 organic result has moved due to SGE. In simple terms, it's like measuring the distance in a straight vertical line from the original position to the new position, counted in pixels.
AVG Vertical Drop (Percentage of Desktop Viewport) - This is similar to the above, but instead of measuring in pixels, it measures the drop as a percentage of your entire screen's height. Think of it as how much of your screen you'd need to scroll down to find the first organic result.
Remember, we're using a desktop browser (1920 x 1080) so the visible viewport is about 940 px high after considering the vertical space taken by your browser's header and footer.
As soon as the SGE button or link is clicked to expand, then almost every No.1 organic position result falls below-the-fold of the page.
This is true across all vertical markets - although you can see from the chart further below that there are significant differences. For example, the impact will be more severely felt in Office Supplies, Technology & Electronics and Fashion & Apparel than Legal and Insurance.
When you look at the typical design of the AI element of the SERPs you can see why.
Here's a SERP for 'smart tvs' as you can see there's textual AI content giving the user advice on what factors to consider when purchasing a smart TV, as well as links to articles, review sites and buying guides. This is followed by a long list of products, which when clicked cause a pop-up on the right-hand side of the page containing Google merchant centre stores.
The number 1 organic ranking has dropped from 631 pixels down the page to 3,537 pixels. A whopping great drop of 2,906 pixels or 269% of the desktop viewport - over 2.5 scrolls down the page!
You do not need a degree in mathematics to work out that this kind of SERP result is going to have a signficant adverse effect on the visibility of and visits to BestBuy's number 1 organic ranking.
But, for other terms the impact will be more muted. Take this result for 'kids stay free'.
The number 1 organic listing here for Usnews.com has 'only' dropped by 781 pixels!
If you generate a lot of your site traffic today through product queries to your product listing pages (PLPs) or product detail pages (PDPs) then you are going to get a rude awakening soon.
The picture doesn't change as much when you look at the 5 different types of keywords. This tells me you need to be as vigilant monitoring the SERP for your brand related terms, as you are for your generic money terms.
SGE content on Google's SERPs features a mix of links, with an average of 10 links per SGE instance, coming from around 4 different websites.
This has not changed much since our last research.
• Average Total Links: 10.8 (0.6 higher than our eCommerce research)
o On average, each expanded SGE element will show a minimum of 9 links and this holds true across all categories. So your old number 1 rank is the new number 10 :(
• Average Unique Links: 5.0
o Out of these links, on average, there are about 5 unique links per instance. This means there is some repetition of the same links across the SGE content.
• Average Number of Unique Domains: 4.3
o On average, these links come from 4 different websites (domains).
• Average Number of Links per Domain: 2.8
o This suggests that, on average, each domain has about 2 to 3 links pointing to it within the same SGE content.
What does this mean? Well, the good news is that if you can rank in the new SGE results, then you have a chance to capitalise on your relevance and try and secure multiple ranking positions which could mean you could win a decent slice of traffic in these instances.
Last time around, this part of the research got a lot of attention and at least one SEO mentioned that they were seeing a different picture to us. So, I've enlisted the help of Harsha, our resident data scientist to help me look at this in a number of ways.
As a reminder, our last study showed that 93.8% of the generative URLs did not match any URL in the top page 1 organic search results, 4.5% were Exact Matches and 1.6% were Domain Matches.
I made the point at the time, that I felt this number was high because our keyword set was focused on eCommerce products and Google ranks PLPs and PDPs well today in organic, but has now started showing articles from 3rd party websites in SGE - completely different content formats, from completely different sites.
So, before we get into the detail, I have shared the data (Organic : Generative Match Results) and you can always do your own calculations if you like. Please note that it is highly likely that many of the SGE results will have changed since this data was collected and we have had a rather major core update since then too!
But, like the last study, this study also shows that Google is selecting pages from a much more diverse set of sources than appear in the organic results.
Exact Matches - 20.1% of the time
Domain Matches -17.8% of the time
No Matches - 62.1% of the time.
Please bear in mind, even if you have an exact match, the data we've already looked at shows that there's usually 10 links and if you match exactly or at domain level you can expect around 3 links at most, which means that there's at least 7 new links competing with you for the searcher's attention and clicks.
The generative pages that do not match a top 10 ranking organic result are fewer than last time around but it does vary by industry vertical and to a certain extent by keyword type.
You can see that the overall match rate by category varies with much lower scores in product-oriented eCommerce verticals than service-oriented categories like Legal and Professional Services.
When you look at the 5 different types of queries we segmented our research into you see a similar picture. Google is showing a diverse set of domains across the board, even for brand terms.
This does mean you need to develop an approach to monitor all types of keywords and ensure you upgrade your competitive analysis dashboards to track your performance in organic and generative results.
It also means that in the future SEOs will need to pay as much attention to the similarity or how well aligned the organic and generative results are, as this will directly play into your content strategy in each area.
We've been looking at this in detail as we start to think through the implications of a global SGE rollout on SEO strategy.
I believe brands will soon need to measure their performance in organic and generative results separately, and in combination together, to get a full picture of competitive performance (not to mention the other SERP feature types like Video or Featured Snippets if they survive).
It's likely that for certain types of keywords and in certain segments of your market you will have new types of SERPs and new types of competitors. They will be stealing your current visibility and traffic for many high performing organic keywords and you will not be able to do much about it, until you understand why they have been selected to rank in the generative results.
The first thing you need to establish is how well aligned the top generative ranking pages are with the organic results (or vice-versa if you prefer). This is what the GOA and OGA Scores aim to tell you.
Where there is strong alignment:
But where there is weak alignment then you will need to focus on:
The following SERP for 'executive club' has a GOA score of 1.0 - meaning we could find every single generative URL within the top 10 organic URLs.
Here's the extracted URLs in order of position.
https://www.britishairways.com/en-us/executive-club
https://www.executivesclub.org
https://www.executiveclubapt.com
https://www.forbes.com/advisor/credit-cards/travel-rewards/british-airways-executive-club
https://executiveclubofsi.com
https://blackandbluesteakandcrab.com/executive-club
https://www.executivesclub.org/about-us
https://play.google.com/store/apps/details
https://apps.apple.com/us/app/british-airways-executive-club/id1388235740
https://www.nerdwallet.com/article/travel/british-airways-executive-club-guide
https://www.executivesclub.org/about-us
https://www.nerdwallet.com/article/travel/british-airways-executive-club-guide
https://www.executiveclubapt.com
https://www.forbes.com/advisor/credit-cards/travel-rewards/british-airways-executive-club
https://www.executivesclub.org/about-us
https://www.britishairways.com/en-us/executive-club
https://www.executiveclubapt.com
https://www.executivesclub.org/about-us
https://www.executivesclub.org
https://www.nerdwallet.com/article/travel/british-airways-executive-club-guide
https://www.forbes.com/advisor/credit-cards/travel-rewards/british-airways-executive-club
https://executiveclubofsi.com
https://executiveclubofsi.com
https://www.britishairways.com/en-us/executive-club
https://www.executiveclubapt.com
https://www.executivesclub.org
https://www.executivesclub.org/about-us
https://www.forbes.com/advisor/credit-cards/travel-rewards/british-airways-executive-club
https://www.nerdwallet.com/article/travel/british-airways-executive-club-guide
In my experience results like this are fairly rare. Results with a much lower aligment between the organic and generative URLs are much more common.
Take this example for 'ai computing' - it has a GOA score of 0.11.
Here's the extracted URLs in order of position. Only the ranking page on www.techtarget.com is worthy enough to grace both lists.
https://blogs.nvidia.com/blog/what-is-ai-computing
https://blog.seas.upenn.edu/new-chip-opens-door-to-ai-computing-at-light-speed
https://ainowinstitute.org/publication/policy/compute-and-ai
https://www.ibm.com/topics/artificial-intelligence
https://www.forbes.com/sites/forbestechcouncil/2023/11/28/preparing-for-the-future-of-ai-computing
https://www.nvidia.com/en-in/about-nvidia/ai-computing
https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence
https://www.komprise.com/glossary_terms/ai-compute
https://www.sciencedirect.com/book/9780323953993/ai-computing-systems
https://www.rigb.org/explore-science/explore/blog/whats-difference-between-ai-and-regular-computing
https://www.investopedia.com/terms/a/artificial-intelligence-ai.asp
https://www.complexica.com/narrow-ai-glossary/computational-intelligence
https://csweb.rice.edu/academics/graduate-programs/online-mcs/blog/computer-science-vs-artificial-intelligence-and-machine-learning
https://www.investopedia.com/terms/a/artificial-intelligence-ai.asp
https://csweb.rice.edu/academics/graduate-programs/online-mcs/blog/computer-science-vs-artificial-intelligence-and-machine-learning
https://www.simplilearn.com/tutorials/artificial-intelligence-tutorial/what-is-artificial-intelligence
https://www.investopedia.com/terms/a/artificial-intelligence-ai.asp
https://csweb.rice.edu/academics/graduate-programs/online-mcs/blog/computer-science-vs-artificial-intelligence-and-machine-learning
https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence
https://www.rigb.org/explore-science/explore/blog/whats-difference-between-ai-and-regular-computing
https://www.linkedin.com/pulse/ai-versus-traditional-programming-whats-difference-ron-new
https://builtin.com/artificial-intelligence/types-of-artificial-intelligence
https://www.complexica.com/narrow-ai-glossary/computational-intelligence
https://www.simplilearn.com/future-of-artificial-intelligence-article
https://builtin.com/artificial-intelligence/types-of-artificial-intelligence
https://csweb.rice.edu/academics/graduate-programs/online-mcs/blog/computer-science-vs-artificial-intelligence-and-machine-learning
https://www.complexica.com/narrow-ai-glossary/computational-intelligence
https://www.investopedia.com/terms/a/artificial-intelligence-ai.asp
https://www.linkedin.com/pulse/ai-versus-traditional-programming-whats-difference-ron-new
https://www.rigb.org/explore-science/explore/blog/whats-difference-between-ai-and-regular-computing
https://www.simplilearn.com/future-of-artificial-intelligence-article
https://www.simplilearn.com/tutorials/artificial-intelligence-tutorial/what-is-artificial-intelligence
https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence
The GOA Score is designed to help you make a data-driven and consistent decision about how well aligned the SERP is between SGE and organic and therefore what action to take.
For each keyword, the GOA score looks at how many generative URLs out of the whole set of generative URLs appear in the set of organic URLs for the keyword.
If all of the generative results were sourced from this organic set then the score would be 1. This is the maximum score possible.
The OGA Score is the reverse of this, as it simply looks at how many organic URLs appear in the generative result set.
No. of unique generative URLs with a match in the organic set / No. of unique generative URLs = GOA Score
No. of unique generative URLs with a match in the organic set / No. of organic URLs = OGA Score
In the example above for 'executive club', you can see that the number of unique generative URLs is 7 and the number of organic URLs is 10.
All 7 generative URLs have matches in this set of organic URLs . Therefore the GOA score is 7 / 7 = 1.0 signifying a perfect alignment and the OGA score is 7/10 = 0.7.
Having a top performing organic page seems to be essential if you want to rank in SGE for this term.
We can use conditional probability statements to determine whether we should use the GOA or the OGA Score, e.g. How likely is that event B would happen, given that event A has already happened?
If we want to understand, "Given a web page ranks in generative, how likely is it that it is also in the organic results?"; we should be looking at the GOA (Generative to Organic Alignment) score. Here it is 1 (or 100% if you prefer).
If we want to understand, "Given a web page ranks in organic, how likely is to also rank in generative?"; we should be looking at the OGA (Organic to Generative Alignment). In this example there is a 70% likelihood.
In the other example above for 'AI computing', you can see that the number of unique generative URLs is 9 and the number of organic URLs is 9.
Only 1 generative URLs matches to this set of organic URLs . Therefore the GOA score is 1 / 9 = 0.11 signifying a very weak alignment and the OGA Score is also 1/9 = 0.11.
It's unlikely therefore that your organic ranking is the key factor in determining whether Google selects your page to appear in SGE.
Using conditional probabilities again:
"Given a web page ranks in generative, how likely is it that it is also in the organic results?"; we should be looking at the GOA (Generative to Organic Alignment) score. Here it is 0.11 (or 11% if you prefer).
"Given a web page ranks in organic, how likely is to also rank in generative?"; we should be looking at the OGA (Organic to Generative Alignment). In this example there is only an 11% likelihood.
When you look at the distribution of GOA (or OGA) Score by vertical, you can see that there are certain eCommerce-oriented categories where the alignment is weak, e.g. Fashion & Apparel, Office Supplies, Technology & Electronics.
These are key niches where Google really is changing the types of content it shows prominently to users. These are the kind of niches where a change in SEO strategy is imperative.
Only for a few niches like Insurance, FInance and Cyber Safety is the alignment score strengthening beyond the midway point.
We believe capturing the full list of visible (and hidden) generative pages and comparing them with at least the top 10 organic ranking pages is essential to help you formulate the correct optimisation approach.
For example, we could try and look at this from both sides at once. The ‘intersection’ divided by the ‘union’ is generally called the Jaccard similarity coefficient (AKA Jaccard Index) in Set theory.
The formula is:
(A ∩ B) / Size (A ∪ B)
I think that’s more than enough formulae for one day. Suffice to say, how well aligned the generative and organic URLs are for your keywords is very likely to have a bearing on the methods you employ to optimise.
And as you can see from the chart below, this is likely to vary by type of keyword or category.
Hint: Our SGE Rank Tracker is ready and waiting for Google to roll this out worldwide and we are planning to give you these kind of critical insights.
Well, if you've earned a Featured Snippet then it is still appearing - but I still feel its days are numbered!
Here's a search results page for 'digital printing'. You can see that the text of the AI generated answer and the Featured Snippet (when you eventually get to it) is pretty similar.
A user has plenty more distractions before getting to Xerox's Featured Snippet. This will clearly have an impact on the contribution value of the current featured snippet to Xerox's organic traffic and sales.
The bad news is that Featured Snippets are pushed a long way down the page by the expanded SGE content and there's every chance a user might have solved their task by the time they reach the Featured Snippet. The average vertical drop of Featured Snippets with generative content was 1,014 pixels. This was 1,487 pixels in the last study; the difference is down to different type of keywords and verticals showing different length AI SERP results. You can see that when we breakdown the data by vertical category.
Either way, your cherished Featured Snippet is not going to be delivering as much traffic to your site in the future.
The silver lining, if there is one, is that if you have worked out how to win lots of Featured Snippets then it won't be a giant leap to win lots of generative listings.
_for_featured_snippets_by_category.webp){kind=tracking}
I could, in theory, analyse every domain in the result set to assess the impact SGE will have on its SEO Visibility.
I did this previously for Wikipedia as an example. But given the number of brand terms in this research study I'm not sure this analysis will enlighten you much more than you already know.
SGE is going to push all your current rankings for all the important non-paid SERP features you care about (organic, featured snippets, people also ask, videos, etc) down the page by a full page scroll at least.
Where your domain does not rank in generative results, then you are going to lose traffic and probably sales.
You really should think about getting a bespoke SGE report about the risk to your site. We'd love to help :).
We analysed every domain that appeared in the SERPs for each keyword to understand which brands were the most prominent in organic and whether this prominence was mirrored in the new generative results.
The top performing domains across all keywords are hardly surprising, with the exception of www.caranddriver.com which is due to the fact that automotive brands were slightly over-represented in our random selection of 251 brands. For this reason, it's much more useful to see the top performing domains in generative results for each market niche rather than across the whole corpus of data.
The standout for me is Quora.com which finished in the top 20 performing generative domains in 11 of 15 categories. The only categories where it failed to make the top 20 were Automotive and YMYL categories such as Legal, Insurance and Finance.
Wikipedia was in the top 20 generative domains in every category. It was the first or second best performing domain in 11 of these categories.
Every SEO's favourite expert site Forbes.com (not shown in the table above- but I have shared the top performing generative domains by category data) ranked in the top 20 in Cyber Safety, Travel, Finance and Telecommunications. Notably, it was the second best performing domain in the Finance category which certainly raised an eyebrow. It was also the 3rd best performing domain for brand keyword terms, when I looked at the top performing generative domains by tag data.
We'll keep an eye on these rankings in the future as my gut feel says a Google 'correction' may be in order.
I'm not going to insert a chart for all 15 verticals - if you would like to see them then you can find them all in the shared SGE PowerPoint presentation or in the spreadsheets mentioned above. For now, I am going to select three of the most interesting categories.
You can see that the major car publishing websites are performing the best. These are the websites that are providing independent consumer advice on the different aspects of buying a car. You'll see that Google is favouring these types of sites across categories. I would not read too much into the relative position of the different auto manufacturers as this is largely a symptom of the keywords selected.
The reason blog.google.com is appearing top is that it appears for so many product terms. It is actually this page (https://blog.google/products/shopping/shopping-graph-explained/) which ranks - the page explains how Google sources the products it is displaying in the SGE lists that appear in the SERP. Behind this, you have the ever-present Wikipedia - but note the prominence of Quora ahead of all of the usual major retailers.
What sticks out like a sore thumb to me is Forbes.com. Forbes is clearly a publishing phenom and it is well represented in generative and organic results across most verticals. I'm just not sure this is the first website I would be looking for when needing financial advice.
These results were collected before the recent March Core Algorithm and AI-Spam updates - so it will be interesting to see whether this position changes considerably when we re-run the data again.
I also decided to look at which pages performed the best in the new SGE results.
As I mentioned previously, Google links to its Shopping Graph help page in many product-oriented SERPs, so it's a clear winner.
It's interesting to note that Alphabet owns 6 of the 7 top performing pages in our result set at least.
You really need to run this analysis for your own set of keywords to assess whether any of your pages, or worse your competitors' pages, are outliers and knocking it out of the park.
As yet, I have not seen any non-Google pages with significant SGE gains, other than Yelp's search page which is appearing for many local searches.
I'm so used to skipping over paid ads when using Google myself, that I also skipped over them completely in my last study and I didn't provide any insights into ads or shopping at all last time around.
So, for all you paid search lovers out there, here goes. ;)
In our study, just over 50% of keyword had paid PPC ads.
_by_tag.webp)
As you would expect, the prevalence of paid results was much greater in ecommerce oriented commodity categories such as office supplies.
_by_category.webp)
As you can see this does vary by category, but it's broadly 50:50. Just over 51% of the time the PPC ads will show above the generative URLs.
I don't want to sound too cynical, but there's only one direction this is going to go in the long-term.
64% of the time Shopping Ads appear below the generative results. I do feel this is likely to change, as Google has plenty of opportunities to squeeze in new ad units within the SGE results, especially where it is currently showing non-paid listings from its Merchant centre.
Data was collected in late February 2023 over a couple of days. We used a Chrome browser from a logged in Google account holder in the United States using a PC running Windows 11.
The browser’s viewport was set to 1920 x 1080 pixels. We calculate the vertical drop in pixels as a percentage of the browser’s viewport to give you an indication of the magnitude of a drop. e.g. If an element drops 1.5 times the viewport, then it’s tantamount to a page and half scroll down the page. This describes the process we followed to capture the data.
This is what we call the SERP post-click, and you will see this represented as 'Universal' or 'U' in the study. So, a ‘U rank’, would be the rank of this URL after a user clicks on the SGE Trigger to generate expanded generative results.
Understanding rankings - We calculate the x, y position of every object on the page pre-click and post-click. We know humans tend to read the SERP results in an F-pattern, so the top ranking is the one with an x, y position closest to the top left of the page (for left-to-right reading languages).
We can count rankings by SERP Feature Type, so you can see if you rank for the first item in organic, people also ask, generative results and so on.
We also count the URLs' Universal Rank on the page, so this is a URL's rank including all SERP types. Generally, unless specifically indicated otherwise, we exclude any paid adverts, paid shopping or irrelevant Google SERP features where you cannot rank a URL from your site (e.g. Travel Finder). We also only include 'Visible' results, since there are sometimes hundreds of hidden links in generative results and carousels.
N/B: When comparing how many places a URL has dropped after SGE clicking, we join the pre-click and post-click SERPs together and create a new 'Index', this index shows the position of the URL on the page amongst all non-paid items. You can use just this column to see how many places a result has dropped.
We appreciate the fact that the visible viewport is less than 1,080 pixels due to the browser header and footer, and consequently could have made these vertical drop calculations even more accurate. However, during our study we noticed that Google continues to experiment with design and markup changes and the current calculation is good enough for the task at hand today. If we changed this to 940 pixels it would only make the data look even scarier!
There was a large number of brand terms in this dataset and this could have skewed the results.
There was also a larger number of automotive brands than brands from other sectors - this could give a misleading view that car sites like caranddriver.com or kbb.com were performing better than other domains across all keywords. For this reason, we've analysed top performing domains by vertical not across all keywords, so domains are compared to their peers.
Please bear in mind, the sample sizes get quite small when you breakdown the analysis by keyword type or vertical. It's usually enough to see a pattern in the type of SERP results that are returned, but it's not guaranteed that it will be like this across all keywords in a segment.
I don't generally comment on algorithm updates, but the recent updates have been huge. Searchengineland is a pretty good place to start if you want the lowdown.
There's every chance that these updates will have impacted the generative results too.
I have manually re-checked quite a few examples and have seen plenty of examples where quite spammy looking 'made-for-search-engine' pages are no longer ranking in SGE results. This could be unrelated or a coincidental, but from my perspective the quality of SGE results does seem to have improved.
I will re-run these terms again in due course and update this post if there are any meaningful differences.
Please bear in mind, that if you cross-check manually any of the keyword ranking or SERPs data that I have shared it's very likely to have changed.
My previous SGE research for ecommerce terms demonstrated that SGE will push the #1 organic result down the page by ~1,500 pixels once an SGE element is expanded. This research supports this view, albeit the mean drop was lower at just over ~1,200 pixels, this did vary considerably by vertical market with several ecommerce-oriented verticals dropping by 1,500 pixels or more.
It's not clear yet how big an organic traffic drop brands will see from SGE.
One thing is clear. It will have the maximum impact if you stick your head in the sand and do nothing to prepare.
The first step starts with analysing your market and we can of course help you with this.
Sign-up to the SGE Rank Tracking waitlist.
The extent of the damage to your online business that SGE will cause will vary considerably. It's evident from our data that certain brands are poised to do reasonably well, whilst others may well have their organic traffic decimated by SGE, leaving them with little option in the short-term other than to invest more budget in paid search and shopping ads and optimising their merchant feeds.
There's no doubt in my mind, that the rollout of SGE will create unprecedented risks to your hard-earned organic traffic. But, there will of course be new opportunities to succeed. You need to be ready now. The most competitive brands in the most competitive niches are already thinking about optimising for SGE, if not doing so already.
Can you really afford to wait and see?
I'll leave you with this:
Google recently announced a reshuffle of its Search Team. A couple of items caught my eye in Elizabeth Reid's, Google's new Search Lead, LinkedIn post announcing her new role.
Google has already served billions of SGE results.
"More is coming soon!"
In the words of everyone's favourite brand SEO expert, Jason Barnard:
“Google's SGE, ChatGPT, and Bing Copilot have changed search forever; Brand is now the primary driver of results. To stay visible and viable in the new world of search, brands must act today and start building a consistent, comprehensive, and cohesive digital marketing strategy, driven by brand and marketing and packaged for SEO.”
We're ready for SGE.
Are you?
The rollout of SGE will create unprecedented risks to your hard earned organic traffic, as well as new opportunities to succeed.
You need to be ready. The only question is, whether you want to be ready now or later?
We conducted a study of 1,000 popular commercial search terms to try and develop a picture of the probable impact of Google's new SGE experience on organic rankings and performance. The research depicts a scary picture of potentially dramatic falls in organic visibility for many sites.
Review our latest assessment of how Google's SGE feature is going to change the SERP as you know it. This has implications for rank tracking, content strategy, page optimisation, content quality and more.
If a web page listing is elevated into the featured snippet position, we no longer repeat the listing in the search results. This declutters the results & helps users locate relevant information more easily. Featured snippets count as one of the ten web page listings we show.
Up until last year, Wikipedia truly dominated as the cited source in knowledge panels for brands. The truth is much more complex, of course – Google gets its information from multiple sources and gets corroboration / cross checks that information across multiple other sources before including a brand in the Knowledge Graph.
You can extract local business listings from Google in two ways:
If you want a quick and easy way to find your own IP address on Google Search, then simply go to your local version of Google and ask the question, “What is my IP address”. Google gives you the answer right at the top of the search results.
The quality of the job ad is probably the most significant ranking factor in Google for Jobs. On top of displaying key elements such as a company info, role description, skills and responsibilities, I warmly recommend to add extra layer of information for Google to digest. For example, working hours, salary, benefits, and a more in-depth company information could make the difference between a good and an excellent job ad copy.
MANY of us have seen SEO click through rate (CTR) studies, performed on large data sets, but what can we learn from these, and, more to the point, are they truly representative? Given the ever changing nature of the SERPs – are click-through rate (CTR) studies too crude and limited in their scope to cater for the multi-faceted nature of a typical SERP? And in fact is there even such a thing as a typical SERP anymore?
