Since 2018, Stagwell has hosted an annual internal innovation competition, encouraging employees from within the network’s 70+ creative and marketing agencies to pitch ideas for new marketing technology to a panel not unlike the sharks on NBC’s Shark Tank. Winning products join the ranks among Stagwell Marketing Cloud’s SaaS tools for marketers.
2022 was no different, and we’re excited to announce this year’s winning product: An AI-powered asset management platform.
The soon-to-be-named SaaS tool (which is currently being referred to by its project name, “Smart Assets”) uses AI to analyze ad creative, suggest edits based on performance data, and enhance creative assets according to those suggestions at the click of a button. Imagine you have a suite of ads featuring a woman in a red dress: Smart Assets might suggest that ads highlighting the color green perform better with your core demographic. From there, the tool can turn the dress from red to green, using AI technology. Just like that, you can optimize your creative, while significantly cutting down on production time and resources.
The three creators of this product, Lindsay Hong, Vitaly Boitelet, and Eric Walzthöny, come from multilingual content agency Locaria and marketing agency Assembly. Last summer, Locaria acquired PEP Group, an omnichannel content creation and production company. The competition’s winners saw this as the perfect opportunity to bring together data and content through this new, first-of-its-kind product.
I sat down with the team behind the winning idea, and we had a conversation about why it’s the right time for Smart Assets, how they think about the relationship between performance and creative, and how the platform works.
Lindsay Hong (LH)
Currently: COO, Locaria
Background snapshot: “I’ve been at Locaria for the past eight years, structuring the business from a linguistics proposition to an end-to-end content scaling solution for any brand that wants to go international, from market research to activation and optimization of concept. Before this, I worked on a lot of digital businesses—I worked in e-commerce, and I was a management consultant before that. My experience lies in identifying opportunities to bring together siloed skillsets, transforming that idea into a profitable business. And that’s what I’ve done again and again working at Locaria.”
Eric Walzthöny (EW)
Currently: Data Scientist, Locaria
Background snapshot: “I lived in Argentina for a while, and there I had a digital marketing startup where we did everything from creative production to digital marketing campaigns. Soon after that, I began digging into machine learning, data science, and AI and fell in love with it. I decided I wanted to formalize my education in it, so I moved to Barcelona to pursue a masters in computer science specialized in AI.”
Vitaly Boitelet (VB)
Currently: Innovation Lead, Locaria and Global Account Director, Assembly
Background snapshot: “I had my first job in media after high school in France. And since then, I’ve been simultaneously studying and working. I studied psychology and corporate finance but always came back to media—I’ve worked in agencies of all sizes, from month old startups to established networks as well. I’ve also worked in a lot of startup environments in Paris. I worked at a SaaS company that scaled really quickly so I have a huge admiration and appreciation for the SaaS business model, and I’ve always been interested in going into that space and building something.”
LH: I think the thing about Smart Assets is, as with all things Locaria, that it is about bringing together previously siloed skillsets. There is value in this tool for brand managers, production managers, media managers…all of them get something out of it. The brand person wants to know you’re meeting their guidelines, the production person wants to know what they’re producing is meeting your guidelines, and the media person wants to know that what you produce is not going to hinder their spend plans.
And that’s why it’s a tool that can be used in-house or externally at agencies. Agencies can say to clients, “Hey, we have this tool, we can guarantee that we meet your brand guidelines and that this will fit on the media channels that you need.” But then also in-house teams can buy this and say, “Hey, media and creative agencies we work with, please use this tool to adhere to brand guidelines and to talk to each other and meet each other’s requirements.”
VB: First, there’s the market innovation of AI tools that are being democratized, lowering barriers of entry to automating tasks that were labor-intensive or even impossible to accomplish.
Secondly, the advertising market is evolving in a direction where brands are consolidating their business to fewer partners in an effort to limit costs and enable their partners to produce more valuable insights. And lastly, the media and analytics space is also experiencing change. Stemming from the same AI proliferation mentioned previously, most of the existing day-to-day campaign optimization methods in use are going to be automated.
I don’t know that the market would be ready for this tool had the technology existed earlier. Fortunately, we are now in this setting where we can observe both the push and pull signals to build it.
LH: I think the other thing that’s so timely about this is, you know, the last decade of language services has had to come to terms with Google Translate, and then neural machine translation. And then more recently, all these things like ChatGPT have come out and altered the traditional process of transcribing content and transcreation. Over the past ten years, there has been a huge transformation as an industry, bringing to the surface different roles and different skill sets that need to evolve.
And now, creative production is going through exactly what the translation industry was going through ten years ago. That makes this product extremely timely and also means we have a lot of experience around the pain points that this kind of industry-wide transformation brings.
EW: From a technical perspective, Smart Assets connects to your ad platforms and analyzes the images that you have, your creative assets. From there, the tool performs object detection to understand the contents of your images. We match that with performance data from your ads, creating a bridge between what components of an image drive performance.
What is very cool about this is that we can perform real time with tools like DALL-E. We do something called an in-painting where we can replace objects in an image, like changing the color of a specific object or including an object that wasn’t there. It provides a deeper understanding of how creatives work with their components and what types of edits or modifications need to be done to make the creative perform much better.
LH: We really care about understanding the ad, so this is not about prejudging your creative content, it’s about understanding what it is first and foremost. If you’ve got 10,000 brand assets across the globe, you want to know: What’s the best format? What channel should it go on? What actress is in it? What season does it apply to? What language is it for? What legal restrictions does it sit under?
There are all sorts of criteria that we can use Smart Assets for to help brand managers and media managers understand the assets before any judgment takes place. This tool is rooted in a deep sense of understanding and curiosity.
From there, it’s about informing decisions that drive engagement. There are some industry standards that no one’s going to disagree with, like the rule of three, the use of negative space, sharpness, platform formats—those are objective. You can measure that very easily. You can produce a dashboard where 80% of your assets meet the platform requirements, but then we can provide custom qualities that you can add as a brand.
VB: We are very careful about not wanting creative teams to sit around and wait for their data counterparts to guide decision making. On the contrary, this tool increases creatives’ involvement in the process by giving them a voice in the metrics conversation. Statistics have been used to optimize media for years now. What we’re doing is empowering creative teams with these tools, finally bringing these two disciplines together.