Will AI replace SEO professionals?
Generative AI tools seem smart, but lack the wisdom and experience needed for effective SEO. Here's why people still must be at the helm.
It’s a topic concerning many in the SEO industry at the moment. And with increasing discussions and scaremongering in the media about how AI will lead to job losses, it’s no wonder.
“Will AI replace SEO professionals?”
No, I don’t believe it will. But there is palpable fear in the industry, and understandably so.
There are people nervously considering side-moves into more stable professions. But what even is a stable profession when AI seems to encroach on them?
Why is there such unease in the SEO community at the moment? We’ve faced the “SEO is dead” trumpets before and lived to tell the tale. So why does this time feel different?
Why generative AI feels threatening
AI isn’t just threatening from the “what if it goes rogue and takes over the world” perspective. Although I’m sure there are people losing sleep over that.
Generative AI is being touted as a time-saving, workforce-reducing miracle tool. It can come up with strategies and automate processes. Plus, it doesn’t have employment rights.
The risk of AI eroding our careers isn’t something completely hypothetical.
Everyone will have seen ChatGPT being used cautiously (if we’re lucky) in our workplaces to streamline once admin-heavy tasks. Engineers are writing code with it. Data scientists are finding trends.
The concern is that it’s only a small step between streamlining processes and removing human intervention entirely.
Using AI in SEO
I do not doubt that you have been trying to use AI within your workflows. It will have helped you write meta titles or enabled you to complete a complex Python script.
AI has an important place in SEO.
We should look at ways to speed up and automate our work where possible. This frees up the time we need to find more strategic growth opportunities.
Limitations of AI in the workplace
At first glance, it seems like AI could solve many commercial problems. It can reduce employees’ workload significantly and, in turn, reduce the associated costs.
However, severe limitations still make it a tool rather than a replacement in search marketing.
Existing information only
One of the biggest limitations of generative AI in digital marketing is that it can only rely on information it already has access to. It can create but only from existing ideas.
Marketing efforts have to keep up with changes in human behavior and appetite. Marketers are much more likely to understand the impact of trends on their marketing efforts than AI would.
Lack of experience
AI has phenomenal theoretical knowledge. It has instant access to documentation and can identify and use relevant information. It is smart.
However, the argument could be made that it is not wise.
There is a difference between the utility of a graduate student who has just spent the last three years learning about their specialization and a seasoned professional who has been dealing with the reality of a subject for a decade.
Remember when you first started as an SEO?
You likely read the beginner’s guides, attended the conferences, and practiced the interview pitches.
When it came to it, though, your first attempt at auditing a website probably yielded a lot of superfluous information.
Yes, you may have been technically right in all that you highlighted, but without the context of knowing how stakeholders think or how workloads get prioritized, it probably didn’t move the needle much.
As a beginner, you may not have realized what actions would have the most impact.
Based on direct inputs only
AI is only as good as the prompts provided to it. There still needs to be some human input for the AI to generate something worthwhile.
As frustrating as many feel meetings are, that time discussing ideas and plans with colleagues can be an invaluable source of direction.
For example, a sneer, rolled-eyes or enthusiastic nodding can convey buy-in or disagreement with a strategy. A human might pick this up and adjust their communication style or plan altogether.
Similarly, economic or legislative news might cause an SEO to reconsider their roadmap. Unless the AI had been instructed to change the plan, it would not know.
Another aspect that limits AI is that it is not privy to the wider context that its output feeds into. Although AI could be programmed to take data from multiple sources, it may never be enough to give it the full context of a situation that a human would have.
For example, AI would need to be given access to current news stories, economic conditions, and brand sentiment to understand whether a PR story will land well and at the right time.
Dig deeper: How to integrate generative AI in your SEO
Get the daily newsletter search marketers rely on.
Why SEO needs humans
Regardless of how much AI can be used as part of an SEO’s workflow, it cannot replace a person entirely.
So many aspects of our roles require compromising, prioritizing and pivoting in response to sometimes subtle inputs.
SEO strategy requires buy-in
For most SEOs, the most time-consuming part of our roles is not the planning or implementing SEO initiatives but getting the buy-in.
An AI system may be able to conceive a solution to a technical SEO problem. Still, it will not be able to drop into the right conversations with engineers to make sure it is carried out properly. It will be impossible to discuss with the team lead why the ticket needs to be prioritized above others in the queue.
SEO needs to factor in other priorities
SEOs have to learn to compromise to get their solutions implemented.
There may be other demands on resources within a company, meaning the technically correct SEO solution isn’t the most pragmatic.
Expecting AI to understand a company’s internal politics well enough to choose the battles to fight isn’t realistic yet.
SEO solutions are unique to the website
SEO strategy is often created in an environment of unique constraints. Each website and each vertical it operates in will be different.
AI will need to fully understand these factors to develop an effective strategy.
Each company will have unique internal challenges, like a lack of a CMS or subject matter expertise. It will be hard for AI to understand how that might impact the viability of its recommendations.
Crafting prompts requires an SEO understanding
For AI to be able to generate useful recommendations, it has to be prompted by someone who understands SEO.
For example, I recently had a discussion with an engineer trying to solve an indexing issue. The engineer had turned to ChatGPT for a solution before coming to the SEO team.
Unfortunately, no matter what they told ChatGPT about the issue they were trying to solve, ChatGPT continued to give the same advice – a solution I had to advise against.
Reading through ChatGPT’s response, I could see why it made those suggestions, but it didn’t know the nuances of the website we were working on, which rendered that solution untenable.
The engineer also didn’t understand enough detail and could not prompt the AI to find a different solution. If the engineer had implemented the solution as the AI had provided, it would not have solved the problem.
The value of SEOs in an AI world
There are risks that not everyone will share the view that AI cannot entirely replace a seasoned SEO. Replacing human workers with AI is likely an attractive prospect to employers, with cost-saving being the main reason.
However, this is a false economy. As we have seen above, there will still be a need for someone within the company who knows enough about how search engines work and how people search to be able to craft the prompts.
We must continue to discuss AI and its benefits to our SEO efforts but also be mindful of its limitations.
These are important discussions to have with budget-holders. Demonstrate how to effectively use AI for SEO to save time and money.
Use AI to free up time to focus on seeking growth opportunities and share that with your stakeholders.
Dig deeper: How AI will affect the future of search
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.
New on Search Engine Land