With artificial intelligence (AI) dominating a lot of attention in the advertising industry recently, Madhumita Chandrasekaran of In the Company of Huskies decided to test-drive one of leading creative AI platforms Daydrm.ai.
There’s been a lot of talk about AI based models like ChatGPT and rightly so. As a text-based AI model that is naturally human-like in its conversations, it represents just how powerful human intelligence can be and how far we have come.
ChatGPT can currently understand a wide range of queries and questions and is able to respond in a manner that seems factual, logical and human. It can make jokes, call out inappropriate queries and can even be honest when it doesn’t have the answer to things.
It’s important to realize that while ChatGPT may be the focus today, it’s not the only model with these capabilities. We have seen AI generated art explode recently with Meta exploring generative art (to produce ads for its online shops through AI), so a future with AI models with different capabilities is on the horizon.
Do AI based models pose a threat to planning?
Given all these developments, there has been a steady rise in conversations around the threats AI poses to various fields, from education to journalism to art and advertising. The fear is a common one- can AI produce unique content that could replicate or even improve on the current human output? So as a strategic planner in advertising, I asked myself that very question.
Can AI based models be good Planners?
Turns out I don’t have to answer this question theoretically because there is already an AI based model for advertising called daydrm.ai. This model created by Aaron Adler (ex-CD BBDO and Droga5 NY) and James Fox (ex-Global CSO Mullenlowe and MD, Global Head of Brand Strategy Goldman Sachs) has been trained on award-winning campaigns across Cannes and D&AD and can generate strategy briefs and creative outputs in 15 seconds.
AI generated strategy: An example
I tried generating both a brief and creative output for a new hypothetical Oreo variant (Marshmallows) and I am both happy and a bit shocked to say, it is pretty good! The UI is simple to use, and you can breeze through it in less than a minute. What it does very well, is force you to focus on the core message that needs to be communicated and produces both relevant strategic briefs and quite promising, creative outputs, in no time!
Now, of course, it’s not perfect. For example, it makes you choose only one age group to target and requires only one task in the brief. This is obviously the ideal scenario, but not the reality for most campaigns. It also seems there is little understanding or room for brand messaging as the model currently is built around the product benefit alone. However, overall it’s still pretty impressive.
Critical factors that make a good planner
Having played around with the AI model, I think a good way of answering the question of whether we can be replaced, is to evaluate how such models stack up against some of the Planners’ key skills or attributes.
A planner is the person who brings the ‘What’ to the table. ‘What is the problem?’, ‘What is the insight?’. And most importantly ‘What do we need to communicate?’.
They are the people who need to answer these questions and come up with a solution that allows for a creative leap. And they need to do all this in a way that is not logical, because as we know from behavioural economics and human behaviour since the beginning of time, logic doesn’t hold the answer to most of our problems. That magic trick requires a lot of reading, researching and most importantly synthesizing. Connecting the dots to find that ‘aha’ moment, the insight that will unlock the comms challenge.
In its current form, daydrm.ai still requires the user to input that insight to generate output. There is also no scope for developing brand centric campaigns which are more emotional and less product centric. This is still something that requires the Planners’ skill sets to execute.
What we as Planners, bring to the table in terms of our nuance, creativity and insight cannot be replaced. For now, at least.
The Personal Perspective
Every planner is unique; they have different backgrounds, interests and perspectives. That diversity impacts how each individual approaches a problem or a solution. No two planners will look at one brief and come up with the exact same solution. And that’s the magic sauce of advertising. Currently, it would seem almost impossible to account for the unique mix of perspectives in any Planner’s head, let alone replicate that.
We as an industry and even as individuals, really connect with the stories we tell through advertising. We are proud of the brilliant ones, we are moved by some, we laugh with others, and we absolutely detest the rest. And that’s because these stories can tap into something through deep nuanced cultural connections and human understanding.
Can AI based advertising stand up to that very tall order?
daydrm.ai is already charging users for the model (there is a 10 day free trial- which I highly recommend), it is priced at about $19 a month. While this may seem nominal now, it’s obvious that these models will be monetized much more heavily in the long run, so they will come at a cost. This raises the question of value. Would brands be willing to pay considerable money for AI based work as opposed to what’s available today?
It can be argued that it may be helpful for small businesses to churn out something quickly. However, the majority of clients are used to getting some bang for their buck when it comes to creative development and often rely on multiple rounds of tweaks and adjustments and research. Can they change their behaviour?
Not to mention that clients would still need resources to direct and use the model. You still need people who are clued into the task to frame queries and set boundary conditions for the model, and that doesn’t come for free.
A tool, not a replacement
The way daydrm.ai is currently designed, it requires someone who knows and understands the challenge to input the task. And like anything else, the output is only as good as the input. The Planner is still the one who must identify the insight or the core message that needs to be delivered. And therein lies the difference.
All this leads me to believe that models like daydrm.ai (at least in its current form) are good tools which can aid planning in some circumstances, but not replace them. At least, not today.
Madhumita Chandrasekaran joined the Strategy team at Huskies in 2021 after relocating to Dublin from India. She previously served as the Associate Vice President of Strategy at Lowe Lintas, Delhi (Part of the MullenLowe Group) and has over 10 years of experience in advertising and marketing across tech, FMCG, durables and automobiles. She’s built award-winning brand strategies and campaigns for Google, Nestle, Absolut, Tata Motors, and Whirlpool.