How the Zaaz Collective Used AI to Create a Launch Video in Four Days
Our goal with the Zaaz weekly newsletter is to share our journey with AI in marketing to guide leaders from a cold start to understanding the huge potential in front of us. I’m excited to share the behind-the-scenes of how we’re building new AI-powered marketing models—what works, what doesn’t, and what we all need to learn together.
This is the most fun I’ve had at my job in years. Culture eats strategy for breakfast. Let’s go.
This week, we'll walk through the approach, workflows, and team that created our Zaaz launch video in a fraction of the time and cost of traditional methods.
Click below to see the result:
As we planned our announcement of the Zaaz Collective, we knew our launch artifacts had to reflect our belief in a distributed work model with AI at its core. We created our launch video with our new framework to show the potential of humans and machines working together.
Day 1: Writing the brief.
While we’ve written creative briefs for decades, we leaned into the spirit of this exercise and prompted both ChatGPT and Claude for creative brief templates. We checked our legacy templates, updated them with a few new fields from the AI results, and built a new brief for the video. We completed the brief with details about the new Zaaz Collective brand guidelines, sample social artifacts from our launch, our goals and must-haves for the video, and details about the final assets.
Day 2: ?Briefing the team.
We met early in the morning to brief our creative lead in Brazil on the approach and brand guidelines, underscoring our goal to create this video with a mix of human creativity and AI exploration. We walked through our brand, target customers, and other important details. We made ourselves available for questions but mostly let the creative lead, who is both a designer and AI specialist, feed the machine with the inputs and guide the narrative for the resulting video asset.
Day 2 continued: ?Human and AI creative exploration.
The creative lead oversaw multiple iterations and sent AI-fueled brand explorations as flat files for notes and guidance. Normally, this process can take an entire team of creatives weeks to turn around. With our approach, we were seeing concepts later that same day.
领英推荐
Day 3: ?Iteration.
We cycled through quick rounds of iteration, mostly over email, and saw the concept come to life in a matter of hours. During one review, our team was confused by a material deviation from the original concept, so we held a quick video conference to discuss. The creative lead explained the machine was hallucinating and placed guardrails into the next round to keep the creative in-bounds. We brought in a second designer to work on layouts for the opening copy to match the social posts we’d built the week before.
Day 4: ?Final review.
After a few more rapid rounds of tweaking the narrative, music, machine narrator, and more details, the team met to review the final video. Our art director sent over all assets produced in multiple formats for wide distribution and use.
What We Learned
When working with AI, you still need a trusted creative at the wheel.
For your distributed model to deliver on your creative vision, you can’t simply reach out to creatives on Fivr and hope for the best. Experts who have deep experience in both design and AI are crucial to making projects that will resonate with your audience. Tapping into a pool of these trusted resources by partnering with an agency like Zaaz is the best way to keep costs low while boosting creative output.
?
2024 is the time for rapid learning and iteration.
We went into this process with a range of expectations for output, timeline, and iterations. We are very happy with the results and turnaround time, but with AI, there’s a learning curve for not only makers but also marketing decision-makers. It’s crucial to partner and work with teams that understand how to feed the machine the right information while still controlling the narrative.
?
The machine has strengths and weaknesses.
AI creative work requires balancing knowing when to set limits and when to let the machine explore. Remember to leave room early in the process to review AI deviations while keeping an open mind to new concepts. Work with creatives who embrace AI and know how to identify hallucinations, iterate for a more diverse set of creative approaches, and ensure the final output falls within the bounds of the original brief.
?
The cost and time savings are massive.
In our big agency days, a project like this could easily take six weeks with a massive team and cost over $50,000 to produce. With our process, the final asset was created in under two weeks for under $4,000. In addition to the video, the AI-driven brand explorations delivered a library of new creative concepts, all within our brand standards, to explore for future campaigns.