Show HN: Aventos – An experiment in cheap AI SEO (opens in new tab)

Thanks for the feedback. To answer your question is: 1) yes, you still need good content as if you're ranking for SEO. However trad seo(i'm shorthanding bcuz i'm lazy) requires you to really to be in the top 5 results. Although this has beeen made worse by google AIO, 0 click answers, and a bunch of other things. For AI SEO you need to be in the top 20-30. Easier-yes but fundamentals like writing good conent is still needed 2) The between trad SEO and AI SEO is that the content you have to create to rank well has changed-often needing multiple pieces to rank well. For instance the prompt "Which AI company has the safest AI models?" has google results show a lot of listing various companies by ranking them safe and least safe-almost promotional content. But if you look at what chatgpt is searching and getting, it's a lot more safety index, reports, and technical information.

There's SEO experts that believe that trad seo and ai seo is the same-that's fair. We're of the group that believe it's pretty much the same, although we're starting to come at an inflection point and it'll change in the future.

To answer your second question: we don't have an icon to give recommended actions. I think that's a mistake we made during planning, where we saw "ai recommended actions" and questioned the validity of them. For instance, it doesn't make sense to edit and change wikipedia becuase that's incredibly hard although a few platforms do recommend that. The idea was that users would improve their keyword research and blog content strategy and we just assist in figuring out what to write for, which we still think is the best approach- however it seems that wasn't very clear. It's different from improving page rank because the strategy has been shifted. Instead of opimizing for the 1st and 2nd result, you're optimizing for topics as a whole and can even appear on the second page of google. If you're saying improving your pagerank via making good content that's unique and high quality-then I fully agree that you need to improve your pagerank

So many LLM submissions here are self promotion spam from accounts with no other activity, which further makes them seem low effort and decreases the average amount of trust they've earned or deserve.

That’s an awesome point. I think it's an aversion on my end to charge a higher price. Partly because it's still a pretty green field so there’s a lot that hasn’t been built yet- features that companies would like to use and we would charge for.

Will we charge a premium in the future? Yeah, but only where we’re confident there’s real ROI and have a great product to back that. For instance, AI attribution in this space is incredibly difficult right now. And in my opinion, without hard numbers (revenue, sign-ups, conversions, etc.), I can’t justify asking for premium pricing.

Maybe I'm thinking about it wrong tho and making excuses for myself, I think you have a good point

How third parties query your product: For ChatGPT specifically, they open a headless browser, ask a question, and capture the results like the response and any citations. From there, they extract entities from the response. During onboarding I’m asked who my competitors are and the response is going to be recongized via the entities there. For example, if the query is “what are the best running shoes” and the response is something like “Nike is good, Adidas is okay, and On is expensive,” and my company is On, using my list of compeitotrs entity recognition is used to see which ones appear in the response in which order.

If this weren’t automated, the process would look like this: someone manually reviews each response, pulls out the companies mentioned and their order, and then presents that information.

2) Distribution of queries This is a bit of a dirty secret in the industry (intentional or not): usually what happens is you want to take snapshots and measure them overtime to get distribution. However a lot of tools will run a query once across different AI systems, take the results, and call it done.

Obviously, that isn’t very representative. If you search “best running shoes,” there are many possible answers, and different companies behave differently. What better tools do like Profound is run the same prompt multiple times. From my estimates, Profound runs up to 8 times. This gives a broader snapshot of what tends to show up everyday. You then aggregate those snapshots over time to approximate a distribution.

As a side note: you might argue that running a prompt 8 times isn’t statistically significant, and that’s partially true. However, LLMs tend to regress toward the mean and surface common answers over repeated runs and we found 8 times to be a good indicator- the level of completeness depends on the prompt(i.e. "what should i have for dinner" vs "what are good accounting software for startups", i can touch on that more if you want

oh wow, didn't even noticed that. That's a good catch. sry for having it happening to you. Will be removing the tours

you got it perfectly. Taking scraped results and presenting the data to customers. Based on these results, SEO/AEO experts then make suggestions like "write about X" to help you "improve" in the results. Is it part dark magic part truth? probably, but that's the SEO industry

I think the biggest issue in the industry is there's no way to defintiely tell how useful are these converted users or whether performance is drastically better by showing up in chatgpt. Although attribution itself is a difficult issue in the online marketing industry