If you don’t enjoy writing, ChatGPT and other generative AI may have seemed like your saviour in 2023. But, says Sheila M Averbuch, as the poor content hits the fan, companies need to go back to basics with real writing by humans. So, what types of content should you be commissioning for which marketing needs?
For people who struggle with writing, generative AI was a curse disguised as a blessing in 2023, offering what seemed like a fast, easy way to create content without paying a writer. A series of catastrophes then revealed that this AI solution doesn’t just have pitfalls, it is a pit – one your brand may not escape without injury.
Content strategy has been a particular fail: deciding which type of content should do what, and where. Witness the letter from CEO of Guardian Media Group to Microsoft, revealed by CNN: it accused the tech giant of damaging The Guardian’s reputation.
“What’s most intriguing is the rush to trust these tools and the eagerness to replace human thought with AI”
Microsoft’s home page had run an apparently AI-generated poll alongside a licensed Guardian article about the death of Australian woman Lilie James: the poll asked readers how they thought James died – was it accident, murder, suicide? The Guardian blamed AI and said the incident showed exactly why it didn’t want these tools anywhere near its news.
Microsoft promised to disable polls and investigate, but this wasn’t the only gaffe on the site, which is the default start page of the Microsoft Edge browser. Its news section has featured a rash of bogus stories drawn from what proved to be unreliable web sources, a trend that insiders say is due to reliance on AI over human editors, whole teams of whom Microsoft laid off in 2020.
Haste makes reputational waste
What’s intriguing is the rush to trust these tools and the eagerness to replace human thought with AI – and not just for editorial decisions. At Samsung Electronics, one team member uploaded a recording of a company meeting and asked ChatGPT to write meeting notes. A colleague uploaded confidential semiconductor source code and asked ChatGPT to optimise it.
The incidents led to fire-fighting by the electronics giant and retroactive efforts to create and enforce company policy around generative AI.
The obvious needed to be stated, and Samsung did: uploading anything into these tools sends it outside the company and makes it impossible to retrieve. Share your secrets with the machine, and they’re secret no more.
When it comes to marketing communications, those fails are racking up, too. I know a business owner who got ChatGPT to write a blog post for him in the style of Shakespeare. A strategist friend told me about the marketing plan that her uninformed colleague had generated using ChatGPT, which had zero originality or insight. She counsels businesses to avoid generative AI if originality is important to them.
“It takes a human,” she told me, “to understand context, tone, relevancy. And generative AI can’t do original content – is that what you want for your business? Especially if you’re in an industry with a focus on quality, you’d better not use AI for that.”
Choose the right content to hit your target
So you’ve decided that you’ll put the right foot forward in the coming year and trust humans rather than the AI to write content about your business. What material should you be spending your budget on, and why? Here’s what I advise my clients to do:
Context: Low awareness in the market of your offering or the problem you solve
- Short form – post LinkedIn updates curating excellent insights from thought leaders in your industry (don’t forget to tag them – see more in my article about social media engagement).
- Long form – commission an experienced business journalist to help you write thought leadership articles, with insights drawn from your first-hand experience in the field. Good starting points: capture your insights on trends, pinpoint a common problem you’re seeing among customers, or clear up misunderstanding in the market around an issue you’re expert in. Pitch the finished piece to media outlets, or publish in full on LinkedIn or Medium.
Context: Providing proof points to potential customers
- Short form – reshare content on LinkedIn published by your customers and tag them
- Long form – create customer success stories (also known as case studies), where an impartial journalist interviews you and your client to write about the problem you help them solve
Context: Need to meet targets for qualified sales leads
- Short form – post consistently on LinkedIn and monitor interaction with your content. Reach out individually to potential targets who’ve engaged with your material. Comment intelligently on content shared on LinkedIn by targets.
- Long form – commission an e-book to reflect your insights and market research on the current business problems that prompt customers to reach out to you. Hosted on a dedicated landing page and require name and email address in order to download
Context: Need to establish your expert reputation in a competitive field
- Short form – scan the morning headlines for breaking news touching on your industry or the problem you solve and share this with a comment on LinkedIn. Tag the journalist who wrote the story.
- Long form – commission a thought leadership article that sets out your industry predictions and ‘trends to watch’ in the coming year.
Be an original
Whatever the business context, there is a content type – short or long form – that can meet your objectives.
Remember that it’s not just what content you publish that matters; the ‘how’ is also relevant. Instant content created by a generative AI says a lot about a company’s leadership, its commitment to corporate social responsibility and risk.
There’s no way to know whether the content in the generated article was fully plagiarised from elsewhere, but what is certain is that generative AI’s have largely been trained to write using copyrighted material stolen from experienced writers.
As you look to the coming year, start as you mean to go on – as an organisation committed to quality and sustainability, who cares as much about its methods as its results.