Is SEO for AI all about vectors?
Instead of relying on exact keyword matches, search engines now use vector embeddings – a technique that maps words, phrases, and even images into multi-dimensional space based on their meaning and relationships.
3 strategies mentioned:
- Add semantic topic modeling to your keyword research mix
- Focus on where topic meets intent over keywords
- Make good content (which means make it for humans, not bots)
Of course, the best way to find out what AI might think about your content is to use AI to tell you.
YouTube is making changes to avoid the most annoying ad placement occurrence on the platform, a midroll ad cutting off a sentence.
Starting May 12, 2025, We’re improving the quality of mid-roll ads on YouTube. That means we’ll show more mid-roll ads at natural break points, like pauses and transitions, and fewer ads where they may feel interruptive or cause viewers to abandon the video, like in the middle of a sentence or action sequence.
Another example of AI (I’m assuming) turbocharging contextual features.
As a follow up to my post about Alexa+, this is the inflection moment for voice assistants.
Adding “true” AI capabilities will either expand their use cases and, therefore, adoption. Or it will prove that they’ll merely be an extension of other tech and not a new platform of sorts.
GenAI as customer lock in
Alexa Plus is $19.99 per month on its own or free for Amazon Prime members — a better deal, considering Prime costs just $14.99 per month or $139 per year.
Prime is the digital version of Costco or Sam’s Club. People pay a membership fee with the expectation of receiving more value through savings. Then a bunch of secondary benefits sweeten the deal.
By paying a fee, members are more likely to spend money to justify the initial fee to themselves.
Amazon’s primary retail customer is Prime. Everything else funnels there.
via The Verge