As I reflect on 2023, I'm proud to share that it was another growth year for Spaceback marked by significant accomplishments on the product side. After years of development, we finally released our Social CTV product in September. Thanks to some seriously amazing support from new and existing partners, we have already provided Social CTV ads for literally hundreds of brands! One of our launch partners, CB2, received significant press coverage from Ad Age and Marketing Dive which helped tell our story and re-establish Spaceback as the leader in Social Display and Social CTV.
It was also a good reminder of how far we've come. When we invented Social Display ads in 2017, the concept was entirely new, so it was all about educating the market. Fast forward to today, we see dozens of companies attempting to replicate what we do. While it's incredible to see the Social Display category expand since we created it six years ago, this means we must differentiate by having a superior platform, and our CTV product proves that we are lightyears ahead of the competition. It also demonstrates our commitment to innovation. We are truly pushing the boundaries of video automation.
Despite the strong business performance and successful release of our CTV product, it's important to acknowledge that we're operating in a challenging environment. A few weeks ago, I had the pleasure of attending LUMA's Digital Marketing Summit West (thanks Terry Kawaja and Conor McKena for the invite). I had a great time catching up with some of my favorite partners, investors, and industry friends (I really do love all of you!).
But it was also a rare opportunity to learn more about how the bankers perceive our beloved category: There was no sugar coating that 2023 was a tough year for ad tech. Capital is hard to come by (I would hate to be fundraising right now!) and exits just aren't happening (at least not the good kind). The looming consensus that programmatic businesses will soon face even greater headwinds due to cookie deprecation doesn't exactly elevate the vibes.
There was one topic that everyone seemed to be genuinely excited about: Artificial Intelligence. AI is so hot right now that LUMA featured AI founder and investor, Jeremiah Owyang, as the keynote speaker. During his presentation, Jeremiah shared several predictions about how AI is going to change the world. He argued that in the not-so-distant future, 99% of content will be generated by AI.
I don't disagree with him. In fact, we're already beginning to see the impact of AI on advertising creative. This will be the beginning of one of the most significant changes in the history of advertising technology. With generative AI capabilities going mainstream, marketers are already seeing a need for purpose-built tools to help them better facilitate and deploy AI generated content at scale.
Before I continue, I want to be clear about something: I'm not suggesting that marketers are ready to completely turn their brand identity over to the machines. Humans will most certainly play a major role in brand strategy and designing messages that construct brand identity while driving key results. That said, humans will increasingly leverage AI applications to configure, distribute and optimize these messages at scale. This is where existing creative automation platforms are insufficient.
Another limitation has to do with how our current measurement and attribution systems work. Most measurement tools are focused on campaign performance, not creative performance. We might learn that creative "A" got more clicks than creative "B", but what we really need is structured understanding of the creative elements and how they contributed towards campaign objectives. The creative automation platform of the future will dynamically configure and measure individual creative elements to maximize performance across different ad experiences.
To dive deeper, let's look at how creative automation platforms work today. Most of the time, they are built on top of some kind of asset library which consists of images, videos, product descriptions, promotions, prices, etc. These assets are the building blocks used to automate ad experiences.
Our partners at Criteo, for example, have built a creative automation platform that uses a brand's product catalog as the asset library. Their system automatically builds ads using products that they think a user will be most interested in.
Waymark is a video automation company that uses a different asset library. Their platform automatically generates videos by scraping assets from a brand's website. Brands spend a lot of time and energy optimizing their websites to drive results, so this approach makes a ton of sense - a brand's website is an excellent asset library for creative automation. Without human editing, these videos aren't always 100% production-ready on first pass, but they are making rapid progress. Most notably, Waymark videos tend to be on-brand because they mimic the tone and stylings of the brand's website.
At Spaceback, we have our own unique approach to creative automation. We see the brand's social presence as a highly-curated, living-breathing, asset library, complete with all kinds of structured, real-time engagement data. Most brands have become social-first when it comes to content creation. As a result, a brand's social presence is extremely relevant, a real-time representation of the brand, and communicates so much more than just the products they sell.
Marketers already know that their very best performing assets are often social posts, so when it comes to building advertising creative, it makes sense to start with content that is already proven to generate strong engagement with their target audience. Therefore, the brand's social media presence represents the perfect asset library for creative automation. Brands tend to agree as over 5,000 of them have used Spaceback to automatically build and launch Social Display and CTV ads.
But where does AI fit into all of this? While Spaceback isn't exactly a generative AI platform, we are already incorporating quite a bit of AI in our platform today. Here's one example where the AI is behind the scenes, but fundamental to how we build Social CTV ads. Our system uses AI to automatically enhance the resolution of low-quality assets so that they can be executed as part of a premium CTV experience (we call this "uprezing"). This approach is less about using AI to generate entirely new content and more about enhancing existing assets to be optimally delivered in a new context.
While "uprezing" is definitely an AI-powered feature, this example doesn't fall cleanly into the dichotomy of human-created vs. AI-created content. In this use case, a human-created asset was modified using AI to make it a more versatile building block for creative automation.
Would you consider an "uprezed" ad experience to be human-generated or AI-generated? If you say this is still human-generated content, I'm curious what you think about this next example.
Another new AI-powered feature in Spaceback is "uncropping". "Uncropping" uses generative AI to expand the width and/or height of a media asset to accommodate a larger aspect ratio. Where cropping involves eliminating pixels, "uncropping" generates new pixels to expand the frame around the original media asset.
Here is a visual example of a 300x250 Social Display ad built from a vertically-oriented Instagram post:
Notice the black bars on the sides of the image? While this would probably still out-perform a standard banner ad, it's not an ideal Social Display execution because it doesn't utilize all of the real estate available in the ad placement. Now, let's look at an example of this ad after applying Spaceback's "uncropping" feature.
Presto! The black bars are gone, replaced by a more complete content experience. The resulting "uncropped" Social Display ad is a more effective ad. In addition to improving the user-experience, "uncropping" allows Spaceback to turn assets into different aspect ratios, making our asset library significantly more dynamic compared to how traditional DCO systems typically work.
Would you consider the "uncropped" ad experience to be human-generated or AI-generated? This time, it's a bit harder to answer. The content experience is fundamentally the same, but aspects of the creative were compiled by generative AI. This is a perfect example of how AI can be used to make assets more modular for creative automation. Automation still starts with an asset from a library, but that asset can do so much more when it becomes as dynamic as the template.
"Uprezing" and "uncropping" are just two examples of how Spaceback is already using AI in our platform today, and we will be releasing a lot more AI-powered features soon. While we remain bullish on generative AI, our current roadmap is focused on adding more functionality that enables us to make assets more dynamic for creative automation. We believe this is where AI can be most impactful for marketers in the immediate future, and an important building block towards creating the automation platform of the future. We're looking forward to expanding on our CTV capabilities and helping marketers incorporate AI into their creative strategy in the most brand-safe and impactful ways possible. We anticipate 2024 to mark the beginning of a transformative phase, and Spaceback is excited to continue to lead the industry as we navigate together towards an AI-enabled future.
