Recently I was on Edward Sturn’s podcast – here is the video of it and below it is summary of some key things that I talked about.
Google’s Manual Interventions and the Impact of Public “Noise”
I also think so. I think it probably was the hype that also got them that brought them under the eyes of Google engineers and they were, yeah we need to do something because I’ve seen those cases where if anyone is mentioning anything and it is going viral then in three to four months from that point of time we see actual change in the search results.
Noise plays a part. Yes. Even last year ABC in Australia did an article about seven or eight fake stores ranking in Google search results. So there were some guys who were trying to scam people by doing caveat products, copying the brands and their images and everything else but those sites were ranking.
I think they were using black hat SEO. And as soon as the article was published by ABC next week, those sites were gone from the search results and I was surprised to see the sites were still live but nothing was showing up in the search results and even though just one week before that the sites were ranking really well. So I think it might be possible that Google manually removed the site or maybe someone from ABC reached out to Google and told them that you guys need to take down these sites or not show them in search results.
The Challenge of Creating Fair, Scalable Solutions for AI Search
And also to add on it, I think regarding these broad issues with the search results, Google will have to come up with a solution that applies to everyone. And if it’s a 50/50 case where it applies to large publishers but it don’t applies to small publishers then it’s going to be a big outrage on X.
I think the challenge for Google right now is to come up with a generalized solution for these problems because if they leave out 10% or 80% of it then it’s going to become a big negative PR scenario for them and as of now I don’t think that they have a technical setup where they can bring up a generalized solution to make sure that the AI overviews or AI mode is not game-able. It’s pretty easy to game both these systems right now and game-able might sound like a weird word but it is what it is. You can do certain things where you can make AI mode say certain things about your brand in a certain way that you prefer.
How AI Overviews Amplify Existing Flaws in Search Results
So, yeah, I think what is happening with AI mode—Harpreet you mentioned that a certain type of content where a person is creating a list and placing their own business at number one that is getting shown in AI mode a lot. But if you step back and see what is happening right now is that AI mode is essentially summarizing what is there in search results.
So it is amplifying the positives of whatever Google have done over the years in terms of building high quality search results and it is also amplifying the negatives of the search results. So the impact is getting worse. So let’s say if direct harm, let’s say if someone is searching a medical thing, if 100 people were impacted negatively by Google search results every year because they got a wrong medical information or something like that, AI mode amplifies that to thousands of people.
The Power of System-Level Prompts and Editorial Control
I feel it is going to become a challenge of editorial right now that if the AI mode system level prompt is designed by Google to look for negative stuff and it’s there, there is pretty much nothing that you can optimize for in terms of it to control that answer. I think system level prompts that Google is using for these systems is going to have more impact over time compared to E-A-T or overall marketing strategy that SEO professionals can execute.
Short-Term Tactics for Gaming AI Search (for Small Businesses)
I think the number one would be Reddit where especially if you’re looking at local search results where you want to rank for best related keywords in your city, that answer is mostly getting controlled by Reddit. So you can do some marketing on Reddit.
We discussed some tactics that you can use right now to show up but what is the long-term strategy? These are for just really small businesses who just who are at least have six months of runway and they just want quick leads for their business to survive or get through a hurdle or something like that.
A Long-Term “Web Marketing” Strategy for Large Brands
But if you’re a big brand and you have millions in monthly or yearly revenue, you should be putting a significant amount of resources in building a long-term enough strategy that even if the AI systems are changing or even if there are new AI systems coming along from different companies, you will still show up.
You have to step back and see it as you’re not doing SEO, you are doing web marketing. In terms of hedging your bet, the main thing that you should look at is a system product because with that change everything changes. You always have to keep an eye on what can be the possible system prompt of these AI mode or chat GPT or perplexity can be, what sources they are weighing and how you build a strategy according to that.
Need for Proactive Brand Monitoring
Yes. I think what’s happening right now is that overall all these systems are pulling from the web. So, the impact of anything that is being said about you on the web is just adding up to a lot.
So over time I think what marketing teams need to do is they need to have or build their own tools to do a web monitoring, and based upon that you can as soon as possible push out some content either on your website, social channels as I mentioned.
Dangers of the “AI Slop” and SEO Tooling Hype
I think what one of my friend was telling me is that he’s tracking how much money is going into these geo tools or these new AI powered tools and he’s getting to somewhere around 120 million overall has gone into this space in just the last two years. There will be more and more weird features which as Harpreet mentioned over time will add up to more and more brands getting actually hammered in search results, losing their visibility, losing traffic, revenue, people are going to lose jobs.
Future of Marketing: Agents, Automation, and New Skills
I think that the marketing will shift to more of agentic tools in their marketing and over time what will happen is some of these will actually work, will actually be able to deliver ROI. By 2030 there will be a big part of these agents running the marketing operations. You have to actually start thinking of the way that what you will be doing in a marketing team in 2030 if an agent or multiple agents are also working with you.
Right now I think it’s the right time to start thinking about computer science, not going really deep into it but at least surface level programming languages, how MCPs work as I mentioned, and all these new stuff that is coming up.
Future of Branding: Will Price Trump Brand Identity?
If in 2030 users have so many tools to do deep research analysis of the information, their data, is there any value to a brand except price? Are we heading towards a marketing scene where it’s not going to be about brand but rather it will become more about the price that you can offer your product at? So this is also something that businesses need to consider.
Final Advice: Prioritize and Protect Your Existing SEO
I feel if you’re looking to do AI SEO, whatever you are doing just make sure that you are not going to harm your SEO that you have already done. You should also be really careful around not harming your existing SEO. That will be my number one tip right now because I have heard many cases where medium to enterprise size brands, CMOs are really excited to go ahead with publishing hundreds of these AI generated pages without considering that it may harm their 10 million monthly traffic that they are already getting from Google. So this is my final advice about AI search.
Leave a Reply