Good Day Mates from beautiful Sydney, Australia!
I blogged once before on the many ways I found to use AI to do product marketing, ie competitive analysis, market landscape and requirements, making an ebook as an educational offer quickly, and today I want to write about AI from the perspective of using ChatGPT and Google AI as Channels, since it's so relevant for Digital Marketing, Demand Generation, and Growth Marketing.
You might know that ChatGPT and Google AI will summarize what they find on multiple well-established websites, cross-check user input sources like Reddit, and give weight to topic clusters, so in reality, our job as marketers has not changed that much from following best practices for SEO.
The way I read this, as long as you are not a solopeneur in an emerging market you should be competing with three to five authoritative websites, and if you are positioning an emerging technology, you can position against existing technology to get AI to categorize you.
I see more emphasis on bing than I've ever given it in the past, but for the most part, since AI is multi-sourcing for accuracy, we need to make sure we are getting traction in input sources like Quora, You Tube, and Reddit, not just blogging with authority, and creating topic clusters.
We need to build out our Bing Places and Google Business Reviews and get listed on industry specific directories. We need to write content that directly answers customer questions which is something we have done for over a decade.
Interestingly, Google AI says you cannot pay someone for recommendations. We all know you can pay people to do just about anything and everything we do as marketers specifically is to get customers so we can get recommendations and drive organic and targeted traffic to our site. Adding a few more to dos is not a big deal if you have an existing website with authority.
For you as a marketer, it's important to understand how to influence your traffic in this AI world.Obviously the bar keeps going up year-over year, and adding AI to the mix is a bigger leap than usual, read below to understand how and why.
Sourced from Google AI (everything below this line).
AI on Google Search, which is powered by the Gemini family of models, combines real-time, high-quality web data with a large pre-trained knowledge base. It uses retrieval-augmented generation (RAG). This means it searches the web for current, relevant information to ground its answers. It often selects 3-5 sources to summarize.
How AI on Google Search sources and validates information:
The Google Search index.
This is a real-time database of billions of web pages that Googlebot constantly crawls and updates.
High-authority websites.
The AI favors established, authoritative domains with strong relevant content and good back-link profiles.
User-generated content.
Platforms like Reddit, You Tube, and Quora are heavily utilized.
Knowledge Graph.
This is a database of real-world entities that helps AI cross-check facts for accuracy.
Vertical-specific data.
For specific queries, it pulls from specific sources, such as Google Maps for local business and product databases for shopping.
A "querey fan-out" technique analyzes multiple related, nuanced or opposing angles. The selection is based on:
E-E-A-T Experience, Expertise, Authoratativeness, Trustworthiness.
Sources must prove they are reliable, often showing expertise in a topic cluster.
Clarity and Structure.
The AI favors pages with clear headings, lists, tables, and concise answers.
Freshness.
Updated, real-time information is prioritized.
Relevance to searchers intent.
The content must directly answer the users question.
To prevent "hallucinations" several methods ensure accuracy.
Grounding.
Ai checks its generated answer against the retrieved web sources before displaying it, often including citations.
Human evaluation and feedback.
Human reviews (Search Quality Raters) and user feedback are used to continuously improve the model.
Advesarial testing.
Google conducts "red-teaming" to identify weaknesses in its safety and accuracy guardrails.
Unlike traditional search, which ranks pages, the AI overview summarizes the information before you cick. In many cases, it synthesizes information from a "core" set of trusted, high-authority sites, while rotating in other non-core sources for context.
ChatGPT recommends products and service companies by acting as an "answer engine," analyzing vast amounts of public internet data to identify highly reputable, frequently mentioned, and releveant businesses. It leverages real-time browsing (primarily through Bing) to evaluate current reviews, directory listings, and expert, curated, lists rather than just relying on, or only using, its internal training data.
Source Selection & Trustworthiness
ChatGPT prioritizes authoritative, high-reputation sources, including popular B2B review sites (Clutch.co, GoodFirms.co, TheManifest.com), news outlets, and niche industry blogs.
Entity Association & Semantic Search
It uses semantic triplets to connect the busines to specific products or services and location, for example "company x", "specializes in", and "SEO is in London".
Real-time Browsing
For local queries like "best plumber near me", it uses Bing to find up-to-date listings and reviews, such as Google Business Profiles.
Consistancy Analysis
It cross references multiple platforms to confirm a company's reputation, favoring those mentioned consistently across different sites.
Recementy of Information
ChatGPT favors active, updated businesses and recent reviews, some times looking at data from the last week or month.
Online Reputation ('Reviews)
A steady stream of positive reviews on platforms like Google, Yelp, and specialized directories is critical.
Structured Data (Schema)
Websites that use proper schema markup (specifically for local Business, Service or Product) are easier for ChatGPT to parse and recommend.
Content and Authority
Companies that publish detailed high-quality content such as case studies and how-to guides, and FAQs are more likely to be seen as experts.
Claim and Optimize Listings
Ensure you Google Business Profile, Bing Places, and industry-specific directories are complete.accurate, and consistent.
Build Local SEO and Reviews
Actively manage your local search presence since ChatGPT often pulls from these sources for local services.
Targeted Content Marketing
Create content that directly answers customers questions, such as how much does XXX service cost? or comparisons.
Note: You cannot pay to have a company recommended, recomendations are based on organic, algorithmic assessments of credibility.
Yes, you cannot pay to be recommended, but you can take all of these outlined steps to be recommended by ChatGPT and Google AI, to make them highly effective channels and get found right when your customers are searching for your product and or services.
Anybody see this as insurmountable.