My team and I have been helping brands reach their SEO traffic and conversion goals through content and links for over 10 years. Recognized by industry leaders and household brands as an authority in both organic content and digital PR.
Content Caffeine #19: Perplexity BS, ditch vanity metrics, zero-click update, AI spam news, and more
Is it an 'answer machine,' an 'information aggregator,' or a search engine? According to WIRED, Perplexity is a bullshit machine. The accusation was made following an analysis of Perplexity by WIRED and a developer, Robb Knight. Their investigations showed that Perplexity appears to be accessing and scraping websites from which coders have attempted to block its crawler (Perplexity Bot). Even worse, Perplexity is repackaging and publishing original reporting—sometimes inaccurately and with minimal attribution. Forbes is the latest media company to accuse Perplexity of plagiarizing its content. And, while digging into how their own work was being used, Forbes also found stolen scoops from Bloomberg and CNBC. But, the ethics of the above are a discussion for another time. The issues above sum up all of AI—LLMs and more. As WIRED points out: When a user queries Perplexity, the chatbot isn’t just composing answers by consulting its own database, but also leveraging “real-time access to the web”...to gather information, then feeding it to the AI model a user has selected to generate a reply. It's easy to forget these facets of the technology or to misunderstand it. LLMs are tools that predict the next most probable word, and then the next, and the next. Sometimes, depending on the prompt, the results are wholesale plagiarism. If your use case is a search/answer engine, then, more often than not, the output tends towards plagiarism. What's my opinion? The tech will never be better than it is now. It will not become more reliable, it will continue to hallucinate (er, lie), and it will continue to plagiarize. Eventually, we will stop trying to use it for 99% of the current use cases. Because, there is no escape from an AI-driven misinformation loop. Words on a pageMark Williams-Cook, Founder of AlsoAsked, recently dropped this on LinkedIn: Longer content does not necessarily rank better….The web is not short of quantity of content - it's short of quality. His post provoked several interesting comments. I agree with Mark. Vanity metrics, like length or word count, are outdated measurements of success. But, these metrics continue to dominate because they are easier to scale than providing an information gain or fully serving search intent. Just hit (or exceed) the word count the top posts hit and you're good, right? Not so fast. The problem is that SEO is competitive. You're not trying to match what's there. You're trying to make it meaningfully better. You must figure out why someone has asked a particular question and work smarter to provide them with the answers they need. That has nothing to do with the number of words on a page. Zero-click senseMarketers and SEOs are making a big deal out of the new zero-click search study from SparkToro and Datos. The study shows that the majority of Google searches—58.5% in the U.S. and 59.7% in the EU— result in zero clicks. It's created a bunch of discussion online—for many of the wrong reasons. Danny Goodwin recently made several valid points about the report. In summary, he says:
Danny also reacted to the data showing that 30% of all clicks in the U.S. go to Google-owned properties (e.g., YouTube, Google Images, Google Maps). Let's put this into perspective. Saying that fewer clicks is really bad news colors our perception. Gradual drops have occurred for a long time, and are probably one of Google's tactics to own more of our attention. Yes, this may affect sites where Google offers an option, e.g., travel, but sites like Expedia and other aggregators have contended with this for years. A decrease in the percentage of traffic going to the open web vs Google-owned property does not mean searchers aren't getting what they want. For example, if you end up on YouTube, you're viewing content created by a brand, business or creator who chooses to use that channel. The content may be on a Google property, but it 'belongs' to the creator (who benefits from this). As Danny says: Yes, Google is keeping people within its ecosystem, but that shouldn’t necessarily be “concerning” if you’re optimizing for more than just Classic Search. This graphic distorts reality. Here's why. The sharp drop in August/September is jarring to the eye, and we know this was a period of huge volatility. Take a step back and look at the US numbers from the beginning of the graph—September 2022— to the end of May 2024. Yes, there's definitely a drop. But that represents a 2% aggregate drop of clicks directly from Google search to the open web. In the last 2 years! Hardly an apocalypse. And within that 2 percent drop, we don't know how much of the traffic was funneled through other Google properties to the open web—think YouTube or Images. We've been waiting for AI to dramatically change the landscape of search, but that hasn't happened yet. And, change happens slowly—which is a positive aspect. I'm not being naive. No drop is great news. We can either wring our hands over this situation (which is not too different from what's been happening over time), or we can adapt our SEO and grow our channels as usual. Look, if there are serious drops, or things start accelerating because of technological change, I'll be the first to sound the alarm. Staying informed is crucial for our SEO strategy. Likewise, if and when I see any compelling data, I'll take action. But the data from this study does not impress me. I hope these insights help you. For more tips, follow me here. Additional Resources1. A report from the Reuters Institute for the Study of Journalism has found that a majority of news consumers are suspicious of AI being used to produce the news. "People broadly had fears about what might happen to content reliability and trust." —Nic Newman, senior research associate at the Reuters Institute and lead author of the report. 2. Cloudflare offers to block crawlers scraping information for AI bots.
|
|
Content Caffeine
My team and I have been helping brands reach their SEO traffic and conversion goals through content and links for over 10 years. Recognized by industry leaders and household brands as an authority in both organic content and digital PR.
Content Caffeine For content-obsessed marketers and SEOs Hi there, Welcome back. I truly appreciate your replies to my last newsletter. You asked for it, so I've dug into what Google disclosed about queries and ranking in their trial documents. There was a boatload of info in the 93 documents I read! Hopefully, the key points below (together with my comments) help you plan your SEO and content strategies for the coming months. See you next time,Nicole P.S. I read and enjoy all your emails....
Content Caffeine For content-obsessed marketers and SEOs Hi there, Welcome back. Would you like an inside look at how Google handles Search? I found dozens of details 'hiding' in the Google antitrust trial documents, so I'm sharing a couple of my insights below. Reply with "Yes" and you'll get more in the next newsletter. That letter will arrive after Labor Day, so I wish you a great long weekend with your family and friends. See you next time,Nicole P.S. Did a friend or colleague forward...
Content Caffeine For content-obsessed marketers and SEOs Hi there, Welcome back. Today, I'm sharing our latest poll results (below). But first… Holy moly! Google lost the antitrust trial! (Check the comments on my post). Google will appeal, but I'll give you my opinion on the situation soon. See you next time,Nicole P.S. Drop a question re the Google lawsuit on this post and I'll give you my thoughts in the next newsletter. Thank you! Poll Results Thanks for your feedback. I'll keep fine...