AI Overview for Software House Industry: Key Success Factors at Google – July 2024 Study

AI Overview is an innovative solution that enhanced traditional search results with AI-generated content. It appears above the standard organic and below paid results, offering users more complex and personalized information.

Based on our Agency's research conducted in the Google US in July 2024, we can share the latest data on the software house industry, i.e. companies that develop software, including websites and applications. For this industry, presence in AI Overview on the American search market is key to effectively acquiring sales leads.

Executive summary

1. AI Overview is increasingly rarely displayed in search results. Compared to the May 2024 study, sales phrases have dropped from 28% to 2%, while blog phrases have dropped from 45% to 11%

2. The quality of the link profile is less important for appearing in AI Overview as an answer source – Trust Flow for sources in AIO is lower than for pages in the TOP10 search results.

3. Tech industry services should invest in indirect channels to acquire traffic from Google. LinkedIn and Reddit are services that appeared in the TOP10 search results even for sales phrases reflecting the user's purchasing intent.

4. Ads appeared in 91% of sales-related phrases when the AI ​​Overview module is generated. The median for the number of sponsored links is 5.

5. The content generated by AI Overview contains 2 to 10 links to sources for blog queries and 3 to 9 for sales queries.

6. AI Overview did not replace the featured snippet from search results. The People Also Ask section appeared with every query for blog queries – it is a permanent element of SERPs and therefore can be treated as a guideline for content creation.

7. Google is systematically introducing changes and conducting tests of both AI Overview and traditional search results. For example, the option to save AIO results has appeared, the ability to generate AIO has disappeared, and the size of the AIO module itself has decreased relative to SGE by about 30% (measured in pixels).

Why did we do the study?

The question in the article's topic is one of the biggest challenges in the era of growing popularity of AI. The current change may be an opportunity for companies basing their business on the search engine, but also a huge threat, including losing their dominant position, organic traffic, and consequently revenue. This applies to both websites positioning themselves on the US market, i.e. the technology industry, but also all English-language websites that want to prepare for the future implementation of AI Overview in Europe.

The growing market share of AI-first search engines, numerous estimates regarding the outflow of users from traditional Google, and systematic changes in AIO have prompted us to conduct the following analysis. We want to introduce B2B service owners and marketers to the current state of the technology industry in AI Overview, but also arm them with knowledge that will help them better understand the changes taking place in AI-first search engines and find their chance to increase visibility and revenue.

Details of the study

We had conducted a study of search results and AI Overview in Google US in the B2B technology industry (software house). For the analysis we used:

• search results retrieved July 18, 2024
• search results for almost 2,000 phrases: 991 for sales phrases, i.e. phrases appearing on service or industry pages, and 991 for blog phrases (assigned to articles and blog sections).

The study was conducted using White Dog's proprietary tool based on reverse engineering and machine learning. This approach allows for the analysis of potential success factors in a specific niche in an objective manner on large data sets. 

AI Overview Research Results for the Software House Industry

1. What domains are the most common source of information in AI Overview? Charts for purchase and blog queries.

   
   
   
   
   
   
   In both types of queries, the LinkedIn platform appeared very often in the TOP10 search results. This should be particularly emphasized in the context of sales phrases (services), because LinkedIn does not offer services, and yet it is considered by Google to be a valuable domain worth showing in the TOP10 ranking. It is also worth noting the high share of the Reddit platform for blog queries on the first page of search results. Competition, in the sense of the search engine, is becoming increasingly greater.
   
   This emphasizes the importance of synergistic SEO, Social Media and Content Marketing activities.

2. AI Overview and changes to traditional Search Results

   We also examined the prsence of featured snippets known from traditional search results, in the context of AI Overview. There are industry hypotheses that snippets were a preview of AI Overview – they work similarly to the AIO module, with the difference that the source of the direct answer for the user is only one page.

   Here are some stats related to featured snippets and AI Overview.

   • People also ask, a module in Google containing 4 questions and answers, always appears for blog-like pages regardless of the generation of AI Overview. In the case of phrases expressing purchase intent, PAA appeared for 86% of phrases when AIO was generated and for 77.4% when AIO was not generated. This can be a guideline for the method of creating content - it is important to answer user questions or long-tail phrases.
     
     

   • The Things to know module for sales phrases appears with a frequency of 0.45% and 11.4% (with AIO and without AIO, respectively) and 6.42% and 14.1% for blog phrases.
     
     
     

   • In the case of blog phrases, in 22% and 12.6% (with and without AIO) of them the Discussions and forums module also appears, which emphasizes the importance of user-generated content and the need to comprehensively cover the user's path not only with content in your domain, but also on portals such as Reddit or Quora. For sales phrases, the values ​​are marginal - 0.45% and 3.9%.
     
     

    

3. How does content length affect search  rankings and AI Overview?

   
   
   
   The sources in AI Overview are about 400 words longer than the URLs in the top 3 search results. The graph shows the median. The x-axis shows the positions in the search results from 1 to 30 (grouped by 3 for better readability), while the y-axis shows the number of words in the content of the pages examined.
   
   
   
   

   In the case of blog pages, the median for AI Overview source pages is 2,300 words, which is less than the TOP3 search results. There is also a correlation between content length (y-axis) and position in search results (x-axis) – the longer the text, the greater the chances of reaching TOP10.

4. Link Profile and its impact on AI Overview

   
   

   
   The quality of link profiles (expressed by the Trust Flow parameter from the external Majestic tool) of domains found in AI Overview sources for sales phrases is lower than in the case of pages from the TOP3 search results

   
   
   
   Similarly, in the case of pages appearing in the TOP10 for blog queries. It can therefore be assumed that external linking is not a key factor taken into account when displaying sources in AI Overview.
   

5. Large Language Models prefer structured content, i.e. the impact of paragraphs and their length on AI Overview

   
   For both blog queries and sales queries, the number of paragraphs in the text and the number of words per paragraph for domains appearing in AI Overview is higher than for TOP3 pages.
   
   
   
   
   

AI Overview vs Search Generative Experience

We also conducted a study of the technology industry in early May 2024, just before the Google I/O conference, during which the development of Google search in the context of gen-AI (including a reduction in the level of hallucinations) and the transition from SGE to AI Overview were announced. Thanks to this, we can compare SGE and AI Overview.

SGE to AIO: stats

1. Less AI Overview in search results.
   
   • For sales-related phrases, the AIO share dropped from 28% to 2%.
   • For blog phrases, the AIO share dropped from 45% to 11%.
2. No possibility to generate AI Overview. In SGE, users could still use the option of extended response to the request. After 14.05, i.e. in AI Overview, this is not possible, but Google provides the option to save the result.
   
   
   
3. Fewer results from People Also Ask. This snippet appeared for all queries regardless of intent or AIO (previously SGE) generation. On blog pages, coverage is 100%, while for sales queries it is 86%.
4. More user-generated content in search results. For blog-related phrases, this module appeared in 15% of cases, while currently it is 22%.

See also:

• Generative Engine Optimization (GEO), jako optymalizacja pod AI Overview w wyszukiwarce Google?
  
• Czym różni się GEO od SEO?
• Jak wdrożyć strategię GEO?