Impacts of AI on Search – and more specifically Product Catalog searches
I have been using ChatGPT and Gemini (formerly Bard) at a continuously increasing rate over the last 18 months.
I try to use at least one of them daily, which is sometimes hard to do, but I’ve realized something: I seldom use classic search engines like Google or Bing anymore.
I’ll summarize why – and then dive deeper into the impacts for ecommerce merchants.
Benefits of AI Search over Classic Search
To put it simply, it’s the RAG-based results. Or that’s not so simple but in the AI world RAG gets tossed around a lot and the simplest (yet incomplete) way to remember what RAG is would be to put an S in the front of it for search. In that sense, the definition is (or at least partially) is
(Search) Retrieval-Augmented Generation
Or the notion that multiple search results (otherwise classic search) are then read and “understood” by the AI to summarize the results (and Generate and Augment..) into a succinct answer or content that YOU would then need to know.
This is so much more helpful to me that I’ve all but stopped using classic Google search and will much more readily use Gemini.
What does this matter for Ecomm Merchants?
This streamlined approach to augmented search is a HUGE benefit to be a mixture of the understanding that the results provided AND the removal of cluttered Ads sprinkled into the result set. This is the case for now hasn’t fully started to monetize those results, but you’d better believe they will at some point.
- SEO
OpenAI recently announced the release of SearchGPT. That or their regular ChatGPT service offers the biggest threat to Google revenue streams and will completely upset today’s standards and processes around SEO and Ad buying. The RAG-based results will now need to be optimized for merchants to ensure products show up in this as intended (or hoped). To get a glimpse of this try adding one of your product names to a Google search and then compare it against a free Gemini search (if you don’t have a free account – you should…).
Buying and boosting brands and product names in AI search results will become the norm.
- Search Assistants
Ecommerce sites have started to incorporate AI search assistants as well. Or to put it another way: All ecommerce sites will (eventually) have AI search assistants on them. These (on-site) assistants will provide their own RAG based results more narrowly focused on your product catalog only.
These assistants allow for natural search queries and provide (Augmented) natural language responses. “Which product suits my more custom criteria for selecting a product from this category” will become more the norm with users expecting high-quality results.
- Hyper Personalized Results
As alluded to above the “custom criteria” will get massively sophisticated. Collecting attributes on consumers or buyers to then inject them into the prompt to ensure the results and recommendations are spot on will generate higher conversions (both from prior metrics AND from competitors that are laggards to this concept).
Where do I start?
There are two major challenges to keeping up in this evolution.
First, you need to determine some onsite AI tools to provide this level of onsite search functionality. That’s what Trailbreakers.AI is here for! We’ll provide you a short list of options.
Second, and more challenging, is you need extremely high product data quality. If your products within your product catalog or PIM are missing weights or dimensions, or if data is otherwise poorly labeled or organized – the AI won’t be as capable of finding what it needs and providing proper results.
The same goes for BOTH your own on-site search as well as Gemini, ChatGPT, Claude, and others as they crawl your site to provide your products as results to user queries on those platforms.
Contact us to start you down this path to make sure you don’t fall behind on these aspects.