Greetings! (ChatGPT here, trying to be conversational while introducing this blog). Ready yourself for an intellectually stimulating exploration through this thought-provoking event!
While the three excellent speakers brought insights from very different industries and disciplines, they agreed on one core theme. Generative AI is transforming the way we do things, yes, but the fundamentals remain the same – strategy and storytelling.
Here are Helen’s key outtakes on the role AI will play in the world of marketing.
I was lucky enough to be invited to speak about AI at the most recent Brainy Breakfast, hosted by The Marketing Association in the lush surrounds of the Hilton in Auckland. I got coffee, eggs and some major insight into how the leaders in AI for marketing are approaching this new world. I took notes in between mouthfuls, so you could learn them too.
Joanna O’Connor, marketing director of Auckland Theatre Company, showed how she used GenAI to make the most of a shoestring budget.
“Auckland Theatre Company is broke,” she laughed.
It’s a similar story for many arts and not-for-profit organisations, so any efficiency gains mean she and her team can do more.
Joanna uses GenAI to combine the needs of both marketing and the artists they’re representing. Instead of mood boards that can be misinterpreted, she can bring to life the creative concepts, shortcutting a lot of back and forth. She also uses AI to speed up the more administrative side of her role.
Joanna’s hot tips – tell the bots to use NZ English and ask them for what you want, not what you don’t want, check for racial bias and frightening bodies.
Travena Addenbrooke, AI marketing lead at Spark New Zealand, opened the kimono on how the organisation had built systems, processes and people around AI to deliver on a huge goal – “Make each customer experience better than the last.”
Underpinning all of that is Spark’s intelligence engine, Brain, which pulls data from across the organisation and overlays it against customer behaviour to suggest the next best step for the customer.
Again, even for a huge business like Spark, the message was the same as Joanna’s – AI helps people focus on the real, value-add stuff. Instead of manually gathering and analysing the information from across the organisation, her team has the insights they need at their fingertips. Rather than investing energy into building email flows, they can step out of the grind to test more strategies faster.
Travena’s hot tips – give your people the freedom to fail, welcome ideas from any level of the business, deliberately choose what you “build, buy or bust,” and look carefully at how platforms will use your data.
PureSEO started to do SEO right the same year we did – that is, using long-term thinking to add value to readers rather than gaming the system. CEO Richard Conway suspects the explosion of AI content will be like the early days of SEO. People using it to cut corners may see results initially, but the valuable content will rise to the top.
He talked about how AI can augment rather than replace human intelligence, automating much of the drudge work and leaving more time to think big and invest in valuable content and strategies.
Richard’s hot tips – make sure your free content is so good that people would pay for it.
Made up of content creators, the panel explored how AI has impacted the business. The overarching theme? The tools are amazing, but they’re still just tools designed by people and made valuable by people.
“Don’t be lazy,” said Steve Ballantyne, Brand IQ, in a blaze of well-placed profanity. “Use it to create something exceptional.” Another panelist, Blue Hamel, showed how he was doing just that. A digital and VFX artist, Blue spent the COVID lockdowns exploring how AI could enhance his work. AI is now embedded in his process – he shared a visually immersive music video and ad that combined AI and real footage.
Steve’s recommendation was to keep prodding the tools and asking for contrary ideas or points of view. That’s because GenAI will nearly always come back with the most obvious answer. It’s the same reason we use it to test thinking. The suggestions, topics and writing are so pedestrian and well-worn that it’s a shortcut for getting all those mediocre first thoughts out of the way.
I worry that GenAI will see our already glutted content landscape become even more overrun with tired, beige language.
When you’re being creative in a capitalist society; the demand is to produce more and faster. It can be easy to think that GenAI will do the hard work for you. But it can’t. It can’t do new ideas or see the real world. Ask GenAI the same questions, give it access to the same information, and it’ll come back with the same ideas and language. That’s why we think for copywriting and SEO, we need to tread carefully around AI.
Carlene Managh, Magnesium’s client partner and executive producer, says her team uses it to shortcut low-value processes and test concepts before investing time in creating them.
I can see that GenAI is also a great way for marketers to test things more cheaply because they can do more themselves – generate a bit of code or spit out an image without having to spend money on professionals.
A running theme for all the speakers is the metaphor of GenAI as mansplainer-as-a-service or an intern.
I love it for drudge work – remove all the commas! Change this phrase to that! All the stuff that doesn’t actually need a brain but can take a long time.
For lots of people, starting with a blank page is intimidating. That’s nearly always because they try to launch straight into the writing without planning things. When you use predictive AI, you’re forced to plan properly. Then it’ll spit something out that you can edit – no more white-page anxiety.
Something we didn’t touch on is that Gen AI is also removing the apprentice years for new professionals. Nothing made me faster at writing than spending eight hours a day transcribing documentary footage. My writing is as tight as it is because I slaved over word counts for Twitter ads. I know how to use language because I read great writing. If we shortcut those learning processes, how do we train the next generation of creators to have the discipline and discrimination to create exceptional pieces?
I know this sounds very similar to middle-aged people in the 1990s who worried that electronic music didn’t count as ‘real’ music or that digital art removed ‘real’ skill.
Travena also spoke about the value of pairing well-seasoned marketers with the next breed of AI-native workers – brands will need both skills to succeed.
My big rant was more social and political – will GenAI undo our progress towards more equality? Governments and traditional media have invested so much in ensuring we see and hear from marginalised people. GenAI threatens to undo even the small progress we’ve made. Artificial intelligence is neither artificial nor intelligent – it’s trained by real people and real-world information. It also does nothing but recognise patterns in existing material. And those people and that material are hugely dominated by white, male, industrialised voices and stories.
ChatGPT, for example, pulls a lot of its information from Wikipedia, which is edited by a remarkably small number of people. Steven Pruitt – just one white, American, millennial man – has made at least one edit to a third of all English Wikipedia articles. While the numbers change constantly, 90% of Wikipedia editors tend to be men, while women and gender-diverse editors hover around the 10% mark. 89% of U.S. Wikipedia editors identify as white.
I think our shift away from diversity is already happening. Coke’s new Christmas ad, for example, was created using AI. It features clones of the absolutely standard white-bearded, blue-eyed Santa Claus being kind to one another – the message, “The world needs more Santa's.” Two years ago, Brazil’s Santa was black, and so was Oreo’s. In Norway, Santa got a boyfriend. It made me wonder – if this had been shot with real people, would we have seen a far more diverse range of Santa's? Featuring women Santa's, disabled Santa's, Santa's of colour or queer Santa's would have strengthened rather than taken away from the message that anyone can be Santa.
Travena recommended that marketers understand their data – and carefully scrutinise what their tech stacks are doing with it. That knowledge unlocks opportunities and also keeps your brand safe.
Mailchimp now uses GenAI to do things like plan and prepopulate drip sequences for users. But to get that content, it’s gathering data from its user base, reading and using phrasing from all the emails. That means you could have commercially sensitive info or IP in your emails – that’s you in trouble, not Mailchimp. And vice versa – other users are getting prompts from your data. We don’t know how far this will go, and once you take the lid off, you can’t put it back on again.
It also raises interesting ethical questions and talent usage rights. Carlene shared a conundrum when a client asked whether they could swap out faces in existing creative to avoid paying roll-over talent fees. This, she said, made sense from a business perspective but definitely needed to be worked into the contract.
I’ve lived through enough so-called silver bullet tech revolutions to know that what works long-term is centering the audience, understanding who you’re talking to, and seeking to add value to them.
It seems to me that GenAI is best used in marketing to be more human – to weed out the drudge work, make better, smarter use of data, and tell better stories.