About last march, the data on several of my websites suddenly became a little weird. It's not the clear slopes of the flow, it's the more subtle changes — the new content, the slowness of the recording, which is usually seen in the search for results in two or three days, which is now a week or even longer. The most disturbing thing is that even if these new pages are recorded, the total ranking is on 20 or 30 pages。
My first reaction was that the algorithm was renewed. It's been a long time, and it's sensitive to google's maybe 100-degree whistling. It's true that in those circles, the search engine is adjusting the identification and weight of the ai-generated content. But i didn't take it too seriously, maybe, a bit lucky. I used it as a tool of consumption, just like many others. Efficiency is real, and a subject is thrown out, and a very well-structured and well-spoken article can be published with a complex manual adjustment of the keyword structure and paragraphs. The website is updated on a daily basis, from two to three times a week. Looking at the number of content added backstage, the end of the story was solid, and the site was “growing up”。
The consultation was that there was a lot of content and that the website was not “stronger”. The job, i fell into a very mechanical kind of work looking for hot-spot keywords, generated in ai, slightly modified, published. And then you stare at the entries and rankings, and there's not much water in the big sectors. I had thought that the question of advice could be found in a “manual correction” that was not sufficiently thorough. So i spent more work on adjusting and even trying to create the same subject with a different ai model, and then put together and embellished. But the effects are still the same. The site is much more stuffed, like a man who eats but is indigestion, but the search engine doesn't seem to like "absorption."。
The switch point was once a chat with a partner who did the technology stream seo. Instead of talking about algorithms, he consulted me on a question: "do you think that what you're producing today, and what's going on six months ago, is actually going on?" i thought that the zone could not simply be divided by the keywords, but the inner core, the logic of the language, the way information is organized and even the pattern of examples, does carry a very strong “mechanical smell”. The search engine reptile captures mass information daily, and it is estimated to be much more sensitive to this “favorite” than we are human beings. It can't be straight enough to conclude that it is a slag content, but it can be hesitant and wonder what kind of user it is best placed in front of. The consequence is, indeed, slow entry and low ranking。
It was recognized that the consultation could be done without “quantity” and “outside optimization” and that the whole approach to doing business had to change in terms of “quality” and “differentiation”. The “generation-division” line must have run out of credit. I'm going to have to make it look like a human being before talking about optimization. I ran into ai's smart seo assistant just to figure out how to do that efficiently. It attracts me to the point that it's not another content generator, it's more like an analyst on both sides of the seo and content。
I started trying to use it differently. Like, i don't want to write a good article. I'll start by sorting out a central idea and some of the cases that i want to express, even a draft that's a little bit more rhetorical, and then give it to ai's smart seo assistant along with a few key words. I changed the focus of its directive, not to write an article on xx, but to "see how that approach combines with these keywords, and what information users really want to see when searching." in its response, it will provide an analysis of the intent of the subsequent search, indicating at what point in time a user in front of one of my keywords could be aware of, seeking a conceptual interpretation and still dealing with the programme. This gives me a lot of clarity before writing, perhaps the final words of the organization。
Practice has found that the biggest difference is that the consumption process of content has been reversed. The usual content of the keyword resolution is now the content of the common resolution “what i want to express” and “what users want”, and the keyword is the bridge to natural integration. In that process, ai's smart seo assistant played a role in helping me to fast-track and calibrate the mark so i didn't get to high or run away. It's like writing a tool at a time, and i'd like to sort it out in nature, but the data profile shows that more searches are focused on a detailed error code. I re-focused on who would suffer in detail, then developed other techniques. The speed at which the article is recorded and the rate of subsequent hits is much better than what i've written before。
So back to the question of who's going to ask, the content of the ai generation can't make it as fast as i think i can now that the word “generation” alone will not be far. The ultimate purpose of the search engine is to connect questions and answers, and it prefers information-based content that can really handle questions. Ai is a weak deputy, but it cannot be the end, much less the beginning. As a consultant with an understanding of search intentions, as an aider for efficiency, rather than as a writer for alternative considerations, the road may be too busy. In the future, algorithms will certainly become more intelligent, and the value of content will become more and more similar to that of “human”. By then, we could no longer be fighting for ai, but who saved more “human” insights and temperature from the root of ai. It's actually a lot more difficult to think about, but it's more interesting。
