The Death of Recommendations?
Most people aren't following the forthcoming SCOTUS case Gonzalez v. Google. Arguments will be heard in February. To recap the facts: US citizen Nohemi Gonzalez was killed in a 2015 ISIS terrorist attack in Paris. The family sued Google, Twitter and Facebook arguing they had abetted the attack by hosting ISIS content and "recommending it" to users. Lower courts held the claims were barred by Section 230 of the Communications Decency Act. At issue is the "promotion" of the terrorist content by algorithms, which goes beyond simply hosting the content. Should an "algorithmic recommendation" be outside the scope of Section 230 liability protections? Google argues, "A decision undermining Section 230 would make websites either remove potentially controversial material or shut their eyes to objectionable content to avoid knowledge of it. You would be left with a forced choice between overly curated mainstream sites or fringe sites flooded with objectionable content."
- Algorithmic recommendations are a core social media engagement tool. Critics argue they're manipulative, impact mental health and have wrecked havoc on US and global democracy (see, e.g., January 6).
- There's bi-partisan support for Section 230 "reform." But how to do that without ushering in Google's "parade of horribles" is tricky.
- SCOTUS wouldn't have taken the case if they weren't going to change the law in some form. They're probably going to qualify 230 immunity.
Brand Value and AI Content
Consumer technology site CNET said it was experimenting with AI content creation, and had published 75 articles so far. Bankrate is too and apparently ranking well with its AI content. CNET takes pains to suggest humans are still directly involved, which seems to be less the case for Bankrate. While CNET and Bankrate have recognized brands – we can debate their relative strength – both rely heavily on SEO for ad revenue, lead-gen commissions and affiliate revenue. It seems clear publishers will soon use AI tools to reduce costs ("do more with less") and generate more content for SEO, which holds the promise of more revenue. Auto-generated content is against Google guidelines, though there's ambiguity if humans rewrite or edit it. Regardless, Google will likely struggle to address the tsunami of AI content that's coming. At a time when many users perceive Google's results to be declining in value, an AI-content explosion could become a kind of existential threat – different from that of AI-assisted search.
- Brand credibility will likely become even more critical, as trust in generic online content continues to erode. Yet brand building is complex.
- We shouldn't assume AI-content "will suck" but much will be mediocre. This may further diminish confidence in random search results.
- With its HCU, spam and product review updates and E-E-A-T guidelines, Google is aggressively trying to boost search quality.
Most SMBs: 'No MSPs'
The conventional wisdom in the SMB-SaaS world is that small businesses need lots of technology support and prefer outside experts to help them with digital tools. However a new survey (n=500) finds that most SMBs prefer instead to buy software directly from vendors rather than "from an IT reseller, consultant or managed service provider (MSP)." Only 14% of SMBs with fewer than 20 employees would seek out an IT reseller to buy software. As headcount grows so does the the inclination to use third parties. This is somewhat counter-intuitive because larger firms tend to be more sophisticated and can afford dedicated IT staff. Indeed, only 22% - 24% of very small businesses (<20) have an IT person or outside consultant according to the survey. The reasons given for buying through a third party, MSP or consultant vary by company size but include: training, education, decision support, maintenance and, for the largest firms, getting the best price.
- The "why" is missing here. It may be that the smallest SMBs buy directly out of necessity because they can't afford to work with consultants.
- But the issue of trust (for small SMBs) is likely a factor. They may distrust third parties who have perceived conflicts of interest (e.g., commissions).
- Digital agencies weren't included. But once a trusted SMB-agency relationship exists, customers often want broader help and advice.
- Near Memo episode 96: Apple Business Connect goes live ... Is stack ranking of staff effective? ... Productization of AI.
- Five times when ChatGPT got local SEO wrong.
- Strengths, limits of using ChatGPT to write product descriptions.
- How to improve your local business reputation (on Google).
- Google "stack ranks" employees as a likely prelude to layoffs.
- Opinion: AI bots will have bigger impact than search or smartphones.
- GPT Radar is an AI-content detector.
- Removing indented search results has positive impact on CTR.
- Traditional print gains share as marketing budgets scrambled.
- Gen Z: 73% believe social media ads sometimes/always violate privacy.
- TikTok fined 5M euros in France for making it hard to decline cookies.
- US Chamber says it will sue FTC to prevent death of non-competes.
- Bossware: Woman ordered to pay employer for "time slacking."
Listen to our latest podcast.
How can we make this better? Email us with suggestions and recommendations.