AI and Content Creation
Content creation is a staple of all B2B and SaaS company’s marketing strategy. The reasons are pretty evident. Google dominates our worlds and thus we’ve been trained to use it to search for answers. In terms of applications, it has been pretty effective, to say the least. How many searches does the average knowledge worker make a day? According to Google, the average person searches Google 3-4 times daily. ChatGPT agreed, indicating that 77% of users conduct 3 or more searches daily.
I’d posit most of us with desk-based jobs use it a lot more.
While the emergence of AI competitors (from Claude to ChatGPT) may depress search volumes, I’d argue that is unlikely for some time. After all, the nature of searches varies greatly between platforms, and thus in many instances, they are not direct substitutes for the search.
Anyway, the net effect is that a common aim of most companies is to be easily found on Google.
What this translates to is ‘page 1’ of Google (although the concept of pages is less clear since infinite scrolling took off).
And a content strategy plays an integral role in helping to deliver on this (although the likelihood of this being achievable is (a) highly context-specific and (b) increasingly harder each year.
The Content Playbook
The content playbook is well established by now. Create compelling long-form content that addresses some pains that your Ideal Customer Profile (ICP) has and position your company as a possible solution. What this content topic actually is will depend on the strategy. Some content may be thought leadership where there is less attention paid to Search Engine Optimisation (SEO). In contrast, other content is very much keyword-focused targeting words associated with commercial intent.
The trends in recent years have not been great for content creators.
1- Google has continued to reduce the footprint for organic content (naturally favouring paid-for placements)
2- Competition has intensified with many companies using the same playbooks
3- Key fundamentals have been declining - visitor numbers, time on page (dwell time), etc
The net effect is that it is getting much harder to stand out and rely on a content strategy to deliver significant traffic.
In some ways, it is a classic case of the Tragedy of the Commons as popularised in the book by Garret Hardin: "Freedom in a commons brings ruin to all." In essence, what Hardin argues is that when individuals act independently and in their own self-interest, in a shared resource system, collective ruin follows. In the book, the example is rooted in the example of over-grazing a commons. Acting individually it is rational to have your cow graze the commons, but if everyone acts in their self interest, too many cows destroy the commons for all. And that is precisely where we find ourselves with content creation.
Of course, on the flip side, others have doubled down on content and you have heavily VC-backed companies like Intercom and Stripe, to name but two acting as almost publishing houses.
What to do?
The Emergence of AI
In recent years we’ve witnessed the emergence of AI, with lots of AI companies looking to target solutions with scant regard to the implications. Content creation is one. In some ways that is the last thing we need. More content. From AI.
Not only does it increase the noise-to-signal ratio it muddies the water for all.
In short, AI content sucks.
The tone is off
It is subject to hallucination (or passing off facts when they are not)
It is essentially regurgitated content with no attribution of the sources
It lacks humanity (real examples or ‘experience’ or ‘nuance’)
Add these together and it is increasingly easy to see that it is just not good content.
So for those looking to use AI for content creation, I’d suggest you consider an alternative route. I’d also argue there is a strong chance Google will eventually penalise AI-generated content - not due to any strong moral reasons, but simply because it is bad for business. I sure I’m not alone in hitting the back button as soon as I realise I am reading-AI generated content.
That all said. I am all for using AI as a complement (as distinct from a substitute). I’d argue the following are all good uses of AI when it comes to content creation.
Topic ideation
Content structure
Summarising text
Spelling (or grammar) checks
Fact-checking (although Google is likely to be more reliable)
Sourcing a quote (once attributed).
Helping optimise content for SEO
We all face a choice.
I’m for calling BS on AI-created content and calling it out. After all, peeing in the pond is never good, and that is what this is.
So as managers please ensure that those working in marketing functions understand that public facing AI-generated content is not acceptable. (I’ve steered clear of legalities here but there are growing examples of companies trying to blame AI for factual inaccuracies in published content. I’m not sure that argument will hold much sway).
And the next time Linkedin encourages you to ‘Start a post, try writing with AI’ simply ignore it. I’m even tempted to see if I can create an #AI-BS hashtag and to get it trending.
Summary
While this might seem as a rant against AI in a world drowning in new AI applications it is not intended to be. I’m drawing the line with AI generated content. I’m sure I am not alone.
Do you really want to be publishing low value content, that may have technical inaccuracies, that reflects badly on your brand and may ultimately be penalized by Google in a future algorithm update?