With AI creative gaining traction, what does this mean for brands?
Tim Goodman of Media.Monks looks at how AI tools like MidJourney are starting to impact how brands are managing creativity.
One question I’ve been asked a lot recently from business leaders is how to improve their efficiency in maintaining their online presence. Specifically, demand is growing for fresh content that is personalised and delivered at scale. To meet these needs, more assets must be produced more frequently, which requires more agility, speed and cost efficiency. Creating digital assets is traditionally a labour-intensive process, and with the higher demands, organisations are looking more towards automation of the process.
What tools are available to marketers now and where does AI fit into all of this?
Like any other challenge, I start by breaking down the process into key areas and then looking at those discrete functions to see what can be improved or automated. Many organisations spend large amounts of money on asset production but do not understand where the money is going. Large organisations may engage many concurrent agencies to produce assets, and then also many other agencies which may further use these assets in campaigns.
Picture this scenario: a company invests multiple millions over the course of a year to produce photo shoots for seasonal catalogs. To manage this process, they keep a list of all the concurrent projects in a spreadsheet. For such a significant investment, that seems like a big risk if assets skip an approval or miss a deadline. The return on investment for each of the projects is not measured, so the company can’t make informed decisions on where to improve its spend.
As the investment grows, marketing resource management (MRM) is becoming more important for organisations. This is especially true where many agencies now operate remotely, often in different time zones. With more projects producing more assets, MRM allows a consolidated way to set and manage budgets across consuming business units. Each project can have its progress tracked in relation to deadlines. Each creative worker can manage their time and tasks, and their work might use an agile or waterfall methodology. Each asset may need multiple approvals by internal teams.
Just understanding the cost of the asset production doesn’t mean the process can’t be improved. Some of the most time-consuming processes in the production phase are highly repetitive. Think of the creation of scaled images and videos for use across the web, mobile, social and other channels. MRM platforms, for example, Adobe Workfront, have the ability to auto crop and scale now, saving labour time and cost, while still retaining the ability for the marketer to adjust if needed. Organisations should also look to online analytics to measure usage and capture the most successful images to leverage in the future.
Many assets can also be constructed in-house. For example, a product shot could be placed on any one of an organisation’s approved backgrounds. Within the asset, specific elements that may be custom to specific audiences or geographies—e.g. car number plates or restaurant items—could be utilised. This creates a great opportunity for automation. If elements are architected correctly, then the power can be given back to the marketer to construct assets from these reusable pieces. Adobe Photoshop has powerful APIs which can allow great integration to Experience Manager and could be automated for this purpose. This could reduce the overall effort in production and approval and therefore speed up the creation and delivery of assets. It also allows for clients to create assets that are personalised with visual content, which creates a closer relationship with their audience.
In the future, I can see artificial intelligence playing an even greater role in the generation of assets within organisations. Already there are AI services that will create entire images. There is the platform MidJourney, currently in open beta, which I recommend checking out. If you are not already aware, MidJourney has an interface where you type in a handful of words and it will generate a set of images from which you can create further variations or adjust.
I entered the terms “fast red car driving over the sunset in America photo-realistic” and chose the following variation.
The opportunity that this technology creates for the enterprise is huge. Imagine being able to create a scan of your product for assets to be generated in near-real-time. If you engaged a celebrity for endorsement, then photos of that celebrity in various poses could be created automatically, though of course with consideration of IP being taken into account. This technology will also raise questions about how to protect intellectual property or train models based on certain boundaries. Imagine an engine that measured success or popularity and created more assets with aspects of popular images, to be ingested and shared in real-time.
In summary, now is a great time to start looking at your existing processes with the idea of modernising. As the world is moving towards a greater reliance on digital platforms and catering to individual messaging, even small steps will put you in place to be ready when the future arrives.
Tim Goodman, chief technology officer, solutions, global team, Media.Monks
