The digital marketing industry is an industry in total upheaval. The established Titans, those platforms on which a great many of us have built our campaigns on and delivered wonderful results for many a beaming client, are being forced to transform as the AI wave breaks about them. We operate in an industry that moves quicker than any other in the world economy, offering us challenges and opportunities, thrills and scares, in more-or-less equal measure.
Digital marketing is the hare that beats the tortoise around the race track due to its nuclear-charged batteries! In our industry, the barriers to entry are relatively low: software and brand building is the biggest cost outlay, and legislation is, alas in many cases, several laps behind. And with AI improving literally month-by-month, the industry is transforming before our eyes.
The industry is undergoing a revolution. How can a local, Worcester Digital Marketing agency survive as the legal fallout begins?
Content generation has changed. Now, a single person can churn out several blogs in a day, editing them to give them a sense of individual style and to provide them, at its most basic level, with a search engine optimised, keyword rich text. They can ask the AI to include calls to action, to identify the problem of the reader and to provide them with a solution.
The same is also true of images. Now, Midjourney, Leonardo, ChatGPT itself, or actual design software such as Canva, all have the capability to give relatively realistic levels of reproduced photography, to quirky illustrations and artwork.
It will change the supply chain of the industry in ways we are only beginning to see: stock imagery libraries, often hated by those in the marketing world due to their high prices, are undoubtedly in for a continued decline, though this has been on the cards since before AI due to some claims of bad treatment over copyright by stock libraries of photographers who provided their content. In 2023, music photographer Alec Byrne took Getty Images to court over copyright infringement, saying his photographs of popular bands had been used without his permission. Perhaps ironically, the same Getty Images is now suing Stability AI, the core model behind image generation, claiming that its learning model used 12 million images to train itself without Getty’s permission. The High Court in the UK has allowed the case to proceed.
Should Getty Images win, would this smother the nascent AI industry at birth? And what would it mean for other Large Language Learning Models that depend on existing content to provide answers?
What would it mean for the millions of worldwide users who have generated their own AI content, from artistic pieces to imagery for website and blogs to text? Would all these people be found guilty of infringement overnight? Would we see a decimation of AI-generated online content, over and beyond anything that Google has achieved in its latest update this March?
My creative talent is in writing: I am a published author, and I have written innumerable blogs. In 2023 ChatGPT and Bard did certainly change the way we had to work - rather than a writer, I, like many thousands of others, became a conductor of sorts, quality checking and re-writing AI driven pieces, adding our own voices to the rather static cliches that stand out a mile on AI generated text.
From writing to imagery to video to sounds and even individual voices, AI is generating it all across the content spectrum. It is doing so at scale, and it is, in many cases, polluting the Internet.
The Digital Marketing industry’s core players react:
I use Youtube. I use it to listen to music, to watch old history and television programs, to educate myself, and to keep abreast of cutting edge news. Over the last 12 months, there has definitely been a decline in the quality of content: far too many videos feature an anonymous robot voice or videos and imagery that have obviously been generated by AI. And it’s not just Youtube either: on many local searches I have seen cloned sites, each tailored to target a geographic area and have their content keyword stuffed. The costs of duplicating a core site, and using AI to furnish each page with similar but slightly different content, is in danger of becoming a plague and devaluing the whole point of search engines to accurately sort our results.
In March, just last week from the date of me writing this world-beating content that will win me an award for ‘digital marketing agency of the year,’ (no Chat GPT involved!), Google released its March 2024 ‘Core Update!’
In its announcement on March 05th 2024, Google highlighted its aim of reducing unhelpful and unoriginal content from its search results. This makes absolute sense: Google’s entire business model is dependent on giving its users relevant and helpful information - and as an SEO practitioner, then I would advise everyone to conduct themselves according to that mantra (it will save you a lot of bother in the future).
The statement on March 05th claimed it was to keep spam out of search results, and to reduce low-quality and unoriginal sites, and to penalise those sites that used automation (AI) to carry out ‘scaled content abuse.’ Producing content at scale to boost search ranking can now get your site de-indexed entirely (and it has always been against Google policy anyhow).
I actually think this was inevitable, for if it would be permitted to continue then Google would drown in bad content and gamed search results. A reaction of this sort was inevitable.
One figure, quoted in Search Engine Land from Elizabeth Tucker, Director of Product at Google, said the update will help to reduce unhelpful content by as much as 40%.
Bing Deep Search: Using AI to understand search intent.
Microsoft has attacked the AI spam in its search results from the other side: by providing users with a Deep Search capability to answer more complex questions by understanding the search intent. In this sense, AI from Chat GPT-4 is being used by Bing to look deeper into the answers than those provided at scale by publishers using the same LLM to spam the search results.
It is interesting to see how both search engines have tried to deal with the same problem (amongst others) in different ways, and it does make me think, not for the first time, that we are in an arms race of sorts, with the same AI engine used for different purposes.
Will this cause social media sites to react? Will digital marketing agencies be caught in the crossfire?
Scaled up AI is not just a problem for search engines. It is present in social media platforms too. From creating fake profiles, deepfakes, helping disguise phishing attempts and from scaled up content generation, social media platforms are probably more at risk from AI than are the search engines. Their capacity to engage with users directly, on a targeted basis, means they pose more of a threat to the platform’s users.
Each platform will deal with it in its own way: META have set up teams to counter disinformation in the run up to the EU elections in June. Twitter, or X (!), might come out with a policy tomorrow only to change it again the following day.
But whatever happens, AI is the revolution sweeping through the digital marketing space at the moment.
So what is a small, conscientious Worcester digital marketing agency to do as the walls of the Bastille tremble?
First, we need to educate our clients as to the dangers of AI and the position it puts the core sites and platforms that digital marketeers use. This is the very core infrastructure of digital marketing: from search engines to social media platforms, to email systems and CRM tools.
Second, we need to make sure we are as up to date with the latest news on the marketing industry as we can be. We need to be in a position to feel confident in our words and to the advice we give to our clients. We are in a very fast-moving situation as the digital marketing industry threatens to consume itself, and we need to be aware of what steps the major parties will take, and how that might threaten the work we’ve done to date to elevate our clients presence.
Third, we need to spread the risk. We need to encourage our clients to try small discovery budgets on relevant platforms and channels. Standing still in digital marketing is akin to standing on a rail track: the train is coming, but it’s hard to see how far away it is! And spread the risk not just across platform, but by media type. You like writing? Great - but let’s now get some video content up there!
Fourth, and perhaps key: Make good, human relevant content that adds value to the user. If you continue to do that, then you should be relatively immune from updates designed to combat the AI content. (I like to believe that virtue goes rewarded, but this sadly isn’t always so: there are stories of good sites being penalised in Google’s latest update from SEO experts who claim they haven’t used AI at all: though it is not often they tell us which sites these are!).
Fifth: Build your social proof. Try to become known in your circles. Again, this should help insulate you from any blow from the giants! What I mean by this is getting your brand and name on third party sites, from social media to old fashioned news.
Do let me know if you have any other ideas about how you would prepare your agency, clients, or business now the floodgates have opened!
Sources:
Stability AI rejects photo copyright claims from Getty Images in High Court: (Paywall)
Google’s comment on its Core Update: https://blog.google/products/search/google-search-update-march-2024/
Bing Deep Search: https://blogs.bing.com/search-quality-insights/december-2023/Introducing-Deep-Search