[ad_1]
Most content today is arbitrage, simply moving information from one place to another.
Very few blog posts are created new information. Most are for remixing, curating and copying existing content, transferring the same core information from one website to another.
If your content is just shaking up common knowledge, then I have bad news: the robots will eat your lunch. Generative AI is the ultimate arbitrage machine capable of doing this producing thousands of copycat articles faster than you ever could.
To stand out in a sea of commodity content, you need to go beyond routine copy-pasting of information and find other ways to add value.
Fortunately, there are three ways you and your gentle human brain are uniquely qualified to add value beyond AI: experiment, experienceAnd attempt.
The best way to add value beyond AI content is through experimentation: going out into the world, testing ideas, and gathering new information that has never existed before.
LLMs are trained on a staggeringly large data set and continue to consume new information daily. But they are not omniscient. They have gaps in their knowledge: information for which they have not been trained, or more importantly: information that does not yet exist.
When you experiment, you create something new and unique, unique to you and never seen before. If someone wants the information you’re offering, there’s only one place they can get it. It doesn’t exist in the data available for LLMs (at least, not yet). This is something you, and only you, can do.
How to do that
This may sound intimidating, but experiments can be large or small, substantial projects in themselves or quick value-adds to otherwise mundane topics.
You can conduct extensive sector researchlike Aira’s report on the state of link building:
Analyze data generated by your company and its productslike the benchmark report I co-authored and used 150 million page views of Google Analytics data:
Run tests to understand how things workas Patrick Stox did to investigate the impact of block high ranking pages with robots.txt:
Collect data to prove (or disprove) known ideas.like Rand at SparkToro who brings receipts for the idea that Email is the most reliable marketing channel:
This has always been a great marketing strategy (and a great link-building tactic: everyone wants to link to original data, as evidenced by the backlink data for Aira’s report).
But it becomes more effective in an age of near-perfect information, when the marginal cost of creating content is virtually zero and the answer to any common problem can be summoned at a moment’s notice.
Sharing basic information no longer has lasting value: the days of achieving outrageous results as the first brand to share a simple ‘how to’ or tutorial are numbered. Nowadays you have to to create information, but also sharing.
Everything created solely by generative AI is trapped in the realm of theory. It will always be less valuable than the same advice from an authoritative source, someone with clear and relevant experience.
In a world where it is easy to get answers to questions, readers will be more concerned about the source of the answer. You can stand out from faceless AI content by proving to the reader that you’ve gotten your hands dirty, and actually done so experienced what you write about.
If there are fifty websites – or five hundred– by providing an answer to their question, readers can afford to be critical of the source they choose. If they want to learn more about budgeting, they will likely choose the experienced financial advisor over the anonymous CRM solution and a blog post written by “Content Team”.
When they want to buy a new camera, they prefer the reviewer who has bought, used and compared real cameras:
About every brand that pulled product descriptions from popular e-commerce stores or wrote them in theoretical statements:
The more crowded a topic becomes, the more important first-hand experience becomes as a method of differentiation. Your job is to prove the provenance of your advice.
How to do that
This is something we try to do regularly on the Ahrefs blog.
You can write about topics you have firsthand experience withlike Chris, an experienced SEO agency, who has our Beginner’s Guide to SEO Reporting:
Interview people about topics you don’t doas Mateusz questions real marketers about their favorite benchmarks:
Please provide concrete evidence of your experience, whether that means taking a screenshot of the tool, filming the interview or sharing a photo of the book you are referring to:
Share anecdotes and stories that provide context to the information, such as SQ reflecting on his experience writing more than 100 articles:
Get skin in the game, like my attempt to read and review them all SEO newsletter available:
The reverse is also true: you should avoid writing about topics you have no experience with and cannot justify having that experience in.
Most companies I see scaling AI content are cost-motivated. They’re not using generative AI to create new, innovative experiences: they’re trying to save money and are willing to sacrifice quality for publishing speed and headcount reduction.
This provides a clear path to differentiation: make better things, spend more energy and create content that is more than just words on a page.
How to do that
Many of the brands I follow (and the products I pay for) have earned my attention through large, strenuous content campaigns.
There are web comicssuch as that of Poststam Email deliverability guide (with Dunning the super owl):
Video serieslike Paddle’s Netflix-style documentary series about the takeover of a company:
Bookssuch as that of Ahrefs beautifully illustrated children’s book:
Free toolslike Veed’s TikTok downloader:
Unique experiences on the pagesuch as those from Typeform The Star Wars Guide to Net Promoter Scorecomplete with hand-drawn AT-ATs:
This kind of content is rare. It is expensive and difficult to create, requiring specialized skills and cross-departmental collaboration. But difficulty is a problem: if it’s hard to create, it can’t be immediately pumped out by an old company with an old AI tool.
While it is often difficult to justify the effort and expense of these projects, it becomes easier every day. Thanks to generative AI, publishing functional, “vanilla” content (words on a page with a few stock photos) is simply not a differentiator.
The more effort you put into building tools, publishing books, or creating unique experiences, the more likely it is that real people will remember your brand, care about your business, and ultimately buy from you.
Final thoughts
Generative AI makes it very easy to share reasonably well-written, reasonably accurate information on a dizzying array of topics. Humans will never beat AI at this game, and honestly, we shouldn’t try to.
We must accept the growing division of content. Let AI handle the lower end of the content (basic information, definitions, summaries and synopses, lists) and focus the expert human energy on the higher end.
In the age of generative AI, there is no advantage to be gained by simply shifting common knowledge from one place to another. We must find new dimensions of differentiation and leverage our unique strengths: creating new information through experimentation, getting our hands dirty and sharing first-person experiences, and making an effort to create what others won’t.
[ad_2]
