Experimenting with AI – Lessons, Advice and Cautions from Fellow Marketers
First Published: 25
We asked Marketing Association members from across New Zealand how AI has made a difference for them over the last 10 months, and what advice they’d give other marketers based on their own experimentation with different AI tools and technology.
AI was everywhere in 2023. New AI tools seemed to be released every day and existing ones kept improving. No wonder it was one of the biggest topics of conversation among marketers and that everyone has been feeling the pressure to stay up to date with the latest AI technology. As Scott Galloway puts it “AI won’t take your job, people who understand AI will.”
Has AI been game-changing for marketers or not?
The one thing all our contributors agree on is that AI has helped them speed up mundane and repetitive tasks and given them more time to concentrate on strategic initiatives or relationship building. As they say, remember where you add value and use AI as a way of making more time available to do that.
While some describe AI as ‘game-changing’ and ‘transformational’, others feel that despite the hype AI hasn’t changed their roles massively. Moumita and Adnan point out that software like Grammarly and Otter as well as martech platforms and CDPs already used AI and machine learning. What we have seen in 2023 is the acceleration of the tech with the advent of generative AI and LLMs. It’s also brought AI to the forefront of client conversations.
Dale is now regularly fielding questions like “and what is your approach to AI?” or “and how are you going to use AI in this [insert brief here]?”, with both Gavin and Adnan agreeing that AI has accelerated the path to value and reduced the lead time required when producing outputs for clients.
Learning by doing when it comes to AI
When it comes to AI, the only way to learn is to just get stuck in. That’s the key advice from everyone we spoke to. Test, experiment, explore. Sobia suggests starting with small experiments, learn from them, and then gradually integrate AI into your workflow and processes. Adnan agrees, saying that while courses and training seminars are a useful starting point, the real value and tailored use-cases emerge when you integrate these tools into your daily activities.
Moumita encourages you to embrace AI, view it as your assistant and make it work for you. As she points out, AI is here to stay, this is only the start, so use it to your advantage. If you are not, someone else will, and get ahead of you. Joanne agrees, saying that investing a bit of time now to get familiar with AI, will get a thank-you from future you, when you’re in the creative doldrums and have a heap of content to write urgently! Like Gavin, Joanne mentions how AI is helping overcome the dreaded writer’s block in creativity.
One note of caution, though. AI is here to enhance your job, not replace human creativity and intuition. Always edit for tone of voice, always check facts for accuracy, and always add your own human spin to it! As Gavin mentions, don't expect everything outputted from the AI to be correct, you need to be the subject matter expert to ensure that everything aligns with the desired outcome.
Joanne agrees. As with using any new software or other tools, she found that there was a fair bit of trial and error! Setting aside some learning time without any deadline pressure was key to gaining confidence with the tools. For Adnan, the primary roadblock was education and the pace of adoption within their growing team. Investing time to upskill team members in AI tool usage is critical for fostering broader adoption and making the most out of these technologies.
As Dale puts it, it was so easy to get excited about “the thing that AI just made for me”, that learning to cope with Magpie Syndrome was something we all had to go through. Yes, it didn’t exist before. Yes, a computer made it for you from words. Yes, it’s amazing. But… No, it’s not good enough to use for anything.
How marketers are using AI on a practical level – podcasts, video scripts, EDMs, imagery and more
Sobia starts by saying that AI has had a profound impact on her role, particularly when it comes to researching specific topics. In the past, this could take days but now AI is like having a research assistant at her fingertips, which is such a time saver.
Rebecca uses AI for repetitive copywriting that follows a format or formula that allows her to train it to follow her tone of voice such as scripts for 90 second promotional videos.
Joanne benefits from using AI to summarise complex or longer form content into more snack-sized pieces of information, for idea generation and problem solving, to generate alternative copy for A/B testing, and as a starting point for website and EDM content. She now has more time to spend on editing and refining rather than originating.
To this, Moumita adds that AI helps record audio, transcribe, summarise meetings, add captions to videos, create animation videos, improve image quality and much more. What took hours, now only take minutes. She also uses it for brainstorming when there’s no one else around to bounce ideas of on.
Dale’s agency launched a podcast this year, and anyone who has gone through that journey will know that there’s a ton more work in it than meets the eye. AI decoded all their audio and video files into transcripts, built out chapter markers in each episode and turned long form audio into article-form text, for those who read rather than listen.
Dale, Adnan and Gale all point to the benefits of AI for agencies. Paired with the right thinking and human touch, tools like Mid-Journey can turn a conversation into a living concept in a fraction of the time it would take a person to start from a blank piece of paper. And what that means for agencies is that – like Dale – they can use AI to help clients see a use-case for creativity in their specific context, without them having to take an enormous financial leap to do so.
Adnan’s consultancy uses generative AI and LLMs to expedite data analysis, making it easier to spot patterns and develop actionable strategies for their clients. They also use AI technology as a springboard for more efficient and effective strategy development and ideation.
Gavin gives the example of briefing copywriters, designers and creatives - having AI consolidate the strategist's thoughts in a manner that is easily digested by the delivery team.
The AI tools marketers recommend
As Sobia points out the choice of AI tools depends on the specific marketing problem you want to solve. There are tools for text generated, visual, video, SEO, workflows etc but it's essential to assess your unique needs and explore tools that align with your goals.
Most marketers are likely to have tried ChatGPT by now but our contributors all agree that the key to making the most of this and other generative AI tools is to upskill on how to do prompts. Adnan would recommend ChatGPT 4.x for multi-modal input, generative AI and LLM functionality.
Dale is impressed watching his creative team, and finds the AI features built into the adobe suite just incredible. Moumita, too, is a fan of the in-built AI tools that come with BigTech. She’s tried a host of tools, signed up for 14-day trials but tend to stick to Canva or Adobe Firefly for design. She uses ChatGPT and Microsoft Bing to ideate and find these helpful in creating first drafts. She’s testing Claude now and proclaims herself a fan of Descript and Repurpose.
Our other contributors mention AI tools such as ChatSpot, Otto.ai, MidJourney, Fireflies.AI for notetaking, Meta's AI sandbox for the Advantage+ ads suite, and Google Gen AI ASA (automatically created assets) and Performance Max. Joanne thinks that Google’s new SGE sounds like an amazingly easy to use tool, one easy step for generating draft copy and images – and is keen to give it a try once it’s released.
On the analytics side, Adnan recommends Zapier’s Open AI integration for automated workflows, Tealium Predict ML for machine learning predictions in a CDP, Twilio Segment's Customer AI tool to predict, generate, and engage customers in a CDP, and Mixpanel's Spark AI driven by OpenAI (currently in beta) which allows you to have natural language conversations with your analytics dashboards.
Overcoming a few roadblocks with AI
AI doesn’t come without risks and challenges, and this was something all our contributors brought up. Rebecca recommends getting permission from leadership first before using any new AI tools and to run it past legal, as well.
Moumita mentions that at her work, they have blocked access to the public versions of ChatGPT and similar AI tools. There are risks and possible security exposures when using public versions of these tools, including the potential for exposure of commercially sensitive data. Instead, they have a partnership with Microsoft, where they have a dedicated, separate, and secure version of ChatGPT specific to the organisation. It is protected by numerous security controls – providing an environment where we can safely ‘play, test and learn’ with AI.
Sobia agrees, given the novelty of AI technology. She cautions about producing outputs from AI for external use as this could have an implication on your brand and company perception. Personally, she uses AI only for back-end work as there are risks that need to be taken into account.
There are also ethical considerations surrounding data privacy and AI usage that have been important to address. It's an ongoing journey to ensure responsible AI integration, but it's vital for trust, brand, and compliance.
Getting excited about what’s next in AI
Moumita puts into words the excitement that a lot of us are feeling right now. That we are getting to experience the start of something so big and be a part of it. AI will create so many breakthroughs and it can be applied in literally anything. She feels that by being an early adopter, it’ll allow her to take others along on this journey and be a sounding board.
Sobia agrees that the possibilities with AI are boundless and what excites her the most is how the marketing industry will rapidly evolve with this technology. Whether consumers will automatically expect AI in their day-to-day interactions or remain skeptical is a fascinating aspect to watch. Not everyone is aware of AI's full potential yet, so there's a lot of room for growth and innovation.
On that note, she can’t wait to see what educators and universities do to adapt their curriculum to gear up the students for the future of marketing with ever evolving tools and ethical considerations.
Adnan points to multimodal AI agents as a fascinating area to watch. These are automated AI systems designed to analyse data beyond just text inputs, to a myriad of inputs from multiple modalities, such as visual imagery, charts and even excel spreadsheets, to provide more comprehensive insights into intent and context offering an unparalleled level of personalisation. Logan Kilpatrick from OpenAI suggests that as these technologies mature, the need for complex prompt engineering will vanish. This will lower the entry barriers, enabling more businesses to leverage AI, which Adnan views as a significant positive development.
Dale admits that he’s most excited about AI outside of the marketing arena. As he says, it’s great that AI can halve the amount of time it takes us to repurpose content from one format to another, but if you compare that with some of the emerging applications in healthcare, the marketing use-case is worthy of a pat on the head and a “well done” sticker at best.