Commerce content has been something of a poster child for performance marketing in 2023. The practice involves producing content, such as product reviews, roundups and listicles, in an editorial voice that upholds a publisher’s values and standards. It is proving highly effective in driving performance via affiliate links.
Commerce content has been such a hit because, in the best traditions of the partnership economy, it allows publishers to speak with authenticity to consumers while driving revenue growth via a brand partnership.
Its power and potential has made commerce content an increasingly critical element of publishing’s business model, with 81% of premium publishers saying that it forms an important part of their media mix, and 40% expecting to dedicate more than 20% of their budget to commerce content by the end of 2023 - up from 26% in 2022.
Trouble in (publishing) paradise
So far, so good. But, there’s trouble on the horizon, too. Producing a significant quantity of commerce content to meet the demand publishers are experiencing is time- and resource-intensive, and research suggests a majority of publishers still rely on in-house teams to do much of the creating. Scaling this content is not easy, and the problem is exacerbated by the lack of commerce content writers publishers can call upon from within their ranks.
These two issues mean that publishers are missing out on significant amounts of potential revenue. The industry has been looking hard for a solution to the conundrum, and the consensus is that AI might be able to offer some help.
AI to the rescue?
Artificial intelligence offers content creation two things: speed and scale. Deploying AI to produce quantities of commerce content that will help consumers make purchasing decisions and drive revenue for publishers and brands is an exciting idea.
This year has seen numerous digital publishers embark on their own AI explorations. Buzzfeed has launched a steady stream of AI content, from personalised “Infinity Quizzes”, which used OpenAI’s API to write romcoms and generate breakup texts, to an AI-powered “chefbot” called Botatouille, through its food content brand Tasty.
Spiritual wellness publisher Ingenio powers a dream interpretations site with AI, as well as a clairvoyant chatbot named Harmony, who offers spiritual guidance. G/O Media, which owns outlets including The AV Club, Gizmodo and The Onion, has experimented with AI to generate list articles, while The New York Times, The Guardian, Bloomberg, Reuters and others have all published stories written by generative AI.
While not all reviews have been favourable, the creative opportunities are tantalising, especially around personalisation. It is encouraging that the most popular forms of commerce content are new product launches, buyer’s guides, reviews and top-three listicles, because they lend themselves to the kind of content that OpenAI’s ChatGPT and others can produce at scale and at speed.
But why stop at a buyer’s guide - not least when AI could produce personalised product descriptions, recommendations, social posts and even advanced chatbots that help consumers in the lower reaches of the marketing funnel move towards a conversion. The sky really is the limit. Or more accurately, the sky might be the limit one day. Which leads us neatly onto…
AI: issues and limitations
Trust, authenticity and accuracy form the bedrock on which commerce content builds its success - just as it does for the entire partnership economy. The problem with AI is it has well-documented issues in all those areas.
In March 2023, an open letter, signed by industry heavyweights including Elon Musk and Apple co-founder Steve Wozniak, called for “all AI labs to immediately pause… the training of AI systems more powerful than GPT-4”, over fears AI would soon be capable of seizing “control of our civilisation”.
In November 2023 - within days of the UK hosting the AI Safety Summit 2023 - the board of OpenAI (the developer of ChatGPT) stunned the industry by firing Sam Altman, its CEO. Turmoil followed in which the viability of the industry’s most prominent player was questioned, before Mr Altman was reinstated four days later.
So, there are fundamental issues regarding public trust, and even ChatGPT - when asked - highlights its inability “to evaluate the accuracy or quality of content”. The platform goes as far as to say that it is “essential to have human oversight to ensure that the affiliate content produced by AI is factually correct and meets the required standards.”
Food for thought for the future
None of this diminishes AI’s potential in helping to unlock commerce revenue in the future, but we’re not quite there yet. We’ll need to see whether publishers have the ability to develop the kind of AI tools that will help scale and speed up their commerce content offering.
Publishers will also need a light touch in deploying this technology, and balancing the appetite for increased revenues while maintaining the trust of their audience. I look forward to seeing how it all plays out.