Anticipating SEO Trends: My Watchlist for 2024

There are already many blog posts with SEO tips for 2024; while this may look similar, I want to share my own insights based on last year's developments.

Indeed, this topic has been widely mentioned in recent SEO strategies discussions, both positively and negatively. It's clear why it became a hot topic last year and will continue to be in 2024.

The Good - Generative AI is Helpful

We all know generative AI can help with a ton of stuff but let us get the obvious out of the way: content.

If used properly, gen-AI can help you supercharge your content production process; emphasis on properly. You can get it to help you write an outline, proofread, copyedit, and even get additional ideas.

If used smartly, it can really help cut down time spent on some of the tedious tasks. This can even include researching as there are already some AI tools online that offer the functionality (e.g. Perplexity).

Even outside of content, gen-AI can help with some of the technical stuff for SEO. For instance, there are already some ChatGPT plugins that provide the structured data needed in JSON format for your webpage. For freelancers, this could be a timesaver. If you really want to take advantage of this, you really should build your own webapp or even just python scripts to generate this in bulk.

The Bad - Generative AI is Not Perfect

When I was first starting to mess with generative AI early in 2023, the very first lesson that I learned was "garbage in? garbage out". At its very core, these gen-AI models (whether large language models (LLM) or the image models) rely on data to be good. After that, it would then rely on any data that users would enter in their prompt.

This led to the 'prompt engineering' trend. It does hold some value as talking to LLMs today is similar to talking to an intern. You really have to be as specific and descriptive as possible if you want to get good results.

With that said, playing around with LLMs can take time to sort of "master". You need to know how to formulate your prompts well and how to adjust if ever you are not getting the results that you want.

There is also the issue of hallucinations. This is the technical term for LLMs making stuff up if they really don't know the answer to your question or problem. It can be easily remedied or prevented by system prompts and retrieval-augmented generation (RAG) but you'd really have to be a nerd to make the most out of this stuff.

For some traditionalists and/or purists, these reasons may be enough for the distrust in generative AI. These are valid reasons and using LLMs may not really be for everyone. It has its own use cases.

The Ugly - Generative AI Can Be Misused

Before gen-AI has gotten better at emulating writers, websites used to have content mills. These were freelance writers and agencies willing to offer bulk content writing deliverables at bulk.

Heck, I started my SEO journey by being a ghost writer. I didn't even know what a ghost writer was back in the day. (Hint: it's not about writing spooky ghost stories)

A webmaster or a website owner could get a full article for about $5 before. That was a really cheap price back then.

Now, content generation is cheaper for everyone. You could easily get articles from ChatGPT for free. Even with the paid route using API calls, you could get an article for less than half a dollar. That's insane.

With that said, there have been a lot of bad actors taking advantage of LLMs to implement blackhat SEO strategies.

Remember this guy?

This is the site's traffic now. Google slapped a manual penalty on the site.

And this brings me to my next item that I will look out for in SEO 2024.

There really is no escaping Google if you are in SEO, unless you are dealing with a hyperspecific niche in a superlocal target market that uses Baidu, Bing, or Yandex as their search engines.

Google Algorithm Updates

Google really pummeled a lot of websites during the later part of 2023. There were 6 major Google algo updates that started in August.

• August 2023 core update
• September 2023 helpful content update
• October 2023 spam update
• October 2023 core update
• November 2023 core update
• November 2023 reviews update

You would think that the updates aim to correct and penalize websites that have been using generative AI to mass-produce content but it doesn't seem to be that way.

Going back to Jake's case study, the traffic dropped not because of an algo update by Google. They were slapped with a manual penalty.

Why is the difference (algo update penalty vs manual penalty) important here? There is a high probability that Jake's site was slapped with the manual penalty only because he bragged about it on Twitter.

This means that Google algo updates so far does nothing to prevent other webmasters from doing these kind of SEO strategies. I think we know why.

Google's Hand in Generative AI

Even before Bard and SGE's beta was rolled out, Google has sneakily changed their guidance on using generative AI for websites.

Before, AI content has been generally regarded as spam. Now, their tone has changed.

We all know why. Google has released Bard, its answer to ChatGPT. In addition, they have their own Vertex AI offering on Google Cloud which provides developers access to their own LLMs and other gen-AI models.

It makes sense for Google not to aggressively penalize generative AI content since it would disincentivize potential customers. Of course, they want more people to use their products.

LLMs have improved so much in the past few months that any content generated can become indistinguishable from something that a human wrote from scratch. Even with just a small touch of human editing, it can be usable and readable.

As of this writing, there is currently no tool that can detect if a piece of content has been generated by AI. Even OpenAI, the creator of ChatGPT themselves, have scrapped their own "ai writing detector" because of its low accuracy.

There is really no incentive for Google to crack down on websites who heavily use generative AI for producing content. These said sites could even be using Google's own LLMs for that matter.

Search Generative Experience

SGE has been a widely-discussed topic among SEO experts ever since it was revealed last year and for good reason. It can completely derail a lot of SEO strategies simply by existing.

If you have been doing SEO for a while now, you might have heard of zero-click searches. Majority of these zero-click searches currently come from featured snippets.

This is a sample of a featured snippet.

Now here it is on SGE.

This even looks like a short article by itself. There is even updated information which is highlighted.

Would a user still click on the articles on the right?

There is still no global rollout date yet, but you can just imagine how frustrating this might be for publishers and websites alike.

This is why I would most likely focus on building the next item to try and mitigate the negative impacts of SGE next year.

If you have a website, chances are that you made one to get customers or an audience. Ecommerce websites use SEO to get more sales and convert more users. Publishers make websites to get more eyeballs and pageviews for programmatic revenue.

Whatever your purpose is, you will have to build and maintain your brand. Why is this important for 2024?

Should Google and Bing serve SERPs with lots of gen-AI snippets, the difference for getting clicks could be brand awareness.

For example: if a user is currently doing research for the best smartphone for a particular price range, they would most likely click or read from one of their trusted sources; regardless if they are mentioned in the snippet or not.

Loyal users would even help your rankings in the long run. Remember the slides that came from Google's antitrust lawsuit proceedings? User interaction could be ranking signal.

My theory is this: the more users flock to your site from SERPs regardless of your ranking, the better your rankings will be in the long run. This could also be one of the reasons why bigger brands rank better overall (well besides their enormous link profiles but you get my point)

With user interaction possibly being a major difference in 2024 for SEO and AI, it makes sense to look out and focus on the next item.

I would argue that the SEO's primary job is to optimize the site and bring users to the door (clicks). Anything after that is only secondary to an SEO team's responsibilities.

With that said, having a well-built site can bring more organic traffic overall. Even if your site ranks well but results in a lot of bounces, you'll eventually lose your rankings or get minimal traffic as a best case scenario.

Recirculation is a term that I learned just last year. It pertains to how a user or a reader circulates around your website. The more they circulate, the more pageviews you get from them.

Search Experience Optimization (SXO)

SXO is also a term that I have just encountered last year. It basically means merging user experience and search engine optimization.

If you think about it, it makes sense that the two go hand-in-hand. Accelerated Mobile Pages (AMP) was started because Google wanted to train webmasters to improve their website's UX.

As I've said earlier, user interaction could be a ranking signal. Even if it wasn't, it still pays to have a website with fast-loading pages and seamless navigation. You basically want the user to stay a while and read more, which could net you more pageviews or conversions.

Personally, this is something that I find challenging since this is another dependency on tech and UX teams. This calls for a lot of discussion and the need for buy-ins for enhancements on the websites that I handle.

For SEO experts and practitioners everywhere, it seems like our work is cut out for us this year.