I first covered the concepts of Early Stage User Queries and Query Stacking in my 2020 post, User Experience Forecasting. In this post, and given Google has now unveiled it will be taking into account these concepts (albeit under different guises), I’m going to expand on the concepts further as I have with clients over the past 3/4 years.
2023 is the dawn of the AI era in Search (and by Search, I mean both organic and paid).
In the Google Marketing Live Keynote, streamed by Google Ads on YouTube on May 23rd, they revealed a number of new product integrations around Paid Search, but also how Google is going to visually show Ads in search results – but also how Google will display search results (Paid and Organic) based on the context of the entire “search journey” and not just queries in a silo.
Early Stage User Queries and Query Stacking aren’t new concepts, but now they are becoming more prominent it’s important we understand how to translate these concepts into actionable recommendations to benefit clients.
They reflect the need for content to be accessible and understandable to users at all stages of their information-seeking journey, and for search engines to be able to interpret and respond effectively to a wide range of queries, including those that are less precisely phrased or that evolve over time.
Vidhya Srinivasan took to the stage and showed us three queries in succession highlighting that Paid text ads would appear in the Generative AI result, and how Paid ads will appear within Generative AI product suggestions… But then the third query is the most interesting from an SEO perspective.
The three queries Srinivasan used in the example were:
- [outdoor activities to do in maui]
- [hiking backpacks for kids]
- [is it easy for them to learn surfing]
And from this Search Event, Google understands “them” is referring to the “kids”, and surfing possibly relates to Maui.
But this goes further, as to how Google processes our search queries.
Well, we know from Paul Haahr’s 2016 SMX West presentation (link) that Google has been performing a deeper level of query understanding for a minimum of 7 years.
We know that Google:
- Does the query name any known entities?
- Are there useful synonyms?
And the final point that Haahr made on this slide is – context matters.
Context Google is now user from previous search queries in the search event to determine hidden means and provide additional value in the subsequent queries through both organic and paid results.
Google is potentially using the Knowledge Graph, and its existing knowledge of how entities, concepts, and “things” are related, and the degrees of relatedness between them, to further inform the new SERPs and provide added value results that go beyond satisfying multiple common interpretations of a query, but to a consensus based and almost crowd sourced way of catering for multitudinous common interpretations of the query.
Google is also using information from our own websites to inform other Google products, an example being on the Map Pack and Google highlighting that specific products are “mentioned on the website”.
The reason I say “crowd sourced”, I believe for some SGE responses there will be an element of caching, especially for high MSV or breakout search queries.
How Can We Learn From This?
The key takeaways from this align with other Google announcements, and not wanting to sound like a broken record, but we need to:
- Add unique perspectives to our content.
- Ensure our content has a beneficial purpose and a diversifying value proposition.
- Build brand and gravitas (EEAT, links, make people care if you weren’t in the results).
In addition, it will not take long for third-party tools to catch up – as well as bending existing Google Search Console data – to identify URLs appearing for queries we wouldn’t have directly associated them with, and reverse engineering the complex user journey as to how your URL and your value proposition being put forward by the content (main or supporting) on that URL can be associated or tied to secondary and tertiary user needs.
From here, we can improve the user experience of the content, include additional cues and elements, to then further improve the user’s ability to forecast their experience with your product or service (the user experience forecast). You can hear me talk about user experience forecasting on the Voices of Search podcast.
Early Stage User Queries
Early stage user queries, as described in the text, refer to search queries that are initiated by users who are trying to find information or achieve a specific objective, but may lack the precise vocabulary or understanding of the subject to form an effective search query. This lack of familiarity or expertise often leads to vague, broad, or otherwise ineffectively phrased queries that may not immediately yield the desired results.
These users are at the beginning, or “early stage,” of their information-seeking journey, and are typically still in the process of refining their understanding of the subject at hand as well as their specific information needs. Because they may not yet know the most relevant keywords or phrases to use, they may struggle to find the information they’re looking for.
Query Stacking
The concept of “query stacking” is closely related to early stage user queries. It refers to the practice of conducting multiple, successive searches, each one designed to bring the user closer to their ultimate objective. These follow-up searches may refine, expand upon, or otherwise modify the original query, based on the results and insights obtained from previous searches.
For example, a user might initially search for “how to bake a cake,” then refine their search to “chocolate cake recipe,” and finally refine it further to “chocolate cake recipe without eggs” based on the results of the previous searches and their evolving understanding of their own needs and preferences. In this process, their end goal of finding a suitable chocolate cake recipe remains consistent, but the way they phrase their query evolves over time.