Word to Your Mum: Managing the Future of SEO Without Keywords by Cindy Krum, MobileMoxie скачать в хорошем качестве

Word to Your Mum: Managing the Future of SEO Without Keywords by Cindy Krum, MobileMoxie

This presentation, originally presented at Barbados SEO, 2023, covers how people in SEO and Digital Marketing should think about the future of search. It dives deep into the impact that language models like MUM (Multi-Task Unified Model) and how they are likely to help Google limit the processing power needed to return a good result for a query, and how they are also likely helping Google do a better job at making money from targeted ads. Not all queries require the same processing power, and some can be basically the same for everyone who sees them - at least for time periods. When Google shows a Knowledge Graph, Explore cards or Featured Snippets, they know from past behaviors, that many people will find exactly what they need, but if they don't they know the next most likely queries and they present those in search as filters. MUM helps Google create and store common patterns of search related to entity topics, so that they can map out the next most likely search - they call these Journes. Journeys are represented by filters in the search, and the filters help Google understand the searcher's Journey, and predict the next step, so that they can add more filters. All these filters help Google to get the query and the journey into the correct System, which are represented by Micro-moments, and these Systems create short-cuts in the algorithmic processing necessary for Google to surface a result that the user will find valuable. The Micro-Moments are: I want to Know, I want to Go, I want to Buy and I want to Do. Know is associated with most Knowledge Graph, Explore and Featured Snippet results, Go is associated with most Local and Map search results. Buy is associated with Merchant Center, and Do appears to be associated with Google Assistant and YouTube. When Google can't classify a query into one of these systems, it will add more filters to try and figure it out - sometimes even putting the filters above the top paid ads. Journey modeling seems important for Google continuing to sell ads that convert well, and to their concepts in GA4 about cross-device and cross-platform mapping. It also seems useful with the recently announced changes that allow searchers to save Notes, Follow topics, Favorite things, and for sites that are recently visited to improve rankings for the individuals that repeatedly click on them - Which will make search results much more personalized, and will make rank tracking even harder for SEO's. Systems are likely also protecting learning models accessing bad information, and potentially using things like EEAT, not just as a quality signal but as a signal of a human origin - to prevent the Google AI from consuming AI-generated content and creating unnatural filter-bubbles om their learning models.