Chime 如何用 Runway 制作最新全国电视广告

与传统制作相比的成本节约

Shayla Love 是 Chime 的执行制片人：https://www.chime.com/，负责这家拥有超过 1000 万会员的金融服务公司的创意制作和广告。Chime 多年来一直在制作以会员推荐为核心的活动，但希望更高效地制作。在这次对话中，Love 分享了品牌最新的全国性活动，活动中几乎有 100 位真实 Chime 会员，并详细介绍了团队如何使用 Runway 在极短的时间和成本下完成制作。

我们已经拍摄这些推荐广告片大约四年了。它们是直击人心、以业绩为导向的电视商业广告（TVC），以真实会员为特色（为他们的推荐付费），在工作室环境下制作。

AI 的突破在于片尾卡。我们希望展示大量真实的 Chime 会员，以类似“布雷迪家族”风格的网格呈现。领导层看到这个概念后说：“很好，但让我们大幅扩展，不仅展示九个人，而是 100 个人。”于是我给超过 100 位会员发了邮件，要求每人提供三张照片并签署授权书，以换取酬劳。我请求提供专业照片、中性服装以及尽可能干净的背景。收到的照片质量和构图参差不齐，有些我真的认为根本无法使用。但有了 Runway，这变得异常简单。

Runway 帮助我们建立了初步工作流程，并与他们的远程部署团队紧密合作。他们基本上给了我们初学者的工具来开始。一旦我掌握了它，我便开始自己渲染每个会员，调整提示和参数，直到得到合适的输出。不久之后，我对自己可以独立处理所有渲染工作充满信心，最终还教会了我的团队如何操作。

我们通过该工作流程处理了 85 位以上的会员。输出是将静态照片转为动态画面：稳定、电视级准备完成，并且尽管原始素材（UGC 照片）质量参差不齐，却保持一致。片尾卡以网格形式展示所有会员。它成为了展示真实会员的亮点，以这种规模通过传统拍摄几乎不可能实现，成本也会非常高昂。

说实话，这一切始于 Runway 团队与我们密切合作。他们为我们构建的第一个工作流程，仅仅是向我们展示了可能性，这正是让我们从“AI到底能做什么”跨越到“这里有我们实际在做的事情”的推动力。

技术很重要，但由于它进展如此迅速，对我们来说关键的是拥有一个懂得如何将技术落实到真正创意想法中的合作伙伴。

我算过这笔账。即使采用这种部分方法来完成——80 多名成员，同时保证我们需要的质量和数量——用传统方式的成本大约为 30 万美元。这包括差旅、摄影棚时间、工作人员、化妆、服装和制作开销。在全国范围内协调真实人员在同一日期集合，从物流上来说几乎不可能尝试。我们实际上根本不会去提出或制作这场景。

使用 Runway 后，同样的最终画面——真实成员、电视质量、可播出——成本不到 3 万美元，直接支付给我们的成员演员，几乎没有制作开销。

我们正在做一些事情。概念分镜图已经改变了我们在整合营销传播（IMC）活动中获得领导层批准的方式。我们最近使用 Runway 为一则 John Cena 的广告以及一则包含 Larry Bird 和 Cooper Flagg 的 NBA 活动制作分镜图，审批速度明显加快，因为领导层可以逐场景看到我们的创意。这个应用案例非常重要。

我不断在寻找使用 AI 的新机会，重新思考我们所做的一切。从使用我们品牌指南和视觉识别制作品牌照片生成器等基础工作，到大规模付费媒体调整尺寸，再到在不同格式间智能适配布局——不仅仅是拉伸图像然后算完成。

更大的转变是文化上的。每次一个团队用 AI 做了什么并展示出来，另外三个团队就想做同样的事情。我们正处在这样一个阶段，人们开始看到实际的可能性，并根据自己的需求去应用它。

重新塑造我们生产方法的机会无处不在。有些机会显而易见，有些机会需要去发掘，有时你可能只需要问问什么是可能的。对我们来说，我们看到了一个机会，问了AI是否能解决这个问题，然后就付诸实践。一旦你看到它真正起作用——真的起作用，不只是测试，而是用于电视——你就会开始到处看到这样的时刻。

为数以百万计的普通美国人提供免手续费的数字银行平台

cost savings vs. traditional production

Shayla Love is Executive Producer at Chime：https://www.chime.com/, overseeing creative production and advertising for the financial services company with over 10M members. Chime has been producing campaigns focused on member testimonials for years, but wanted create them more efficiently. In this conversation, Love walks through the brand’s latest national campaign, which featured nearly 100 real Chime members, and details how the team used Runway to bring it to life for a fraction of the time and cost.

We've been shooting these testimonial spots for about four years. They're hard-hitting, performance-driven television commercial (TVC) spots featuring real members (paid for their testimonials), produced in a studio setting.

The AI unlock was in the end card. We wanted to show a large group of real Chime members in a kind of “Brady Bunch”-style grid. Leadership saw the concept and said, “great, but let’s expand this massively and show not just nine people but 100.” So I emailed more than 100 members, and asked each of them to send three photos and sign a release in exchange for a fee. I requested professional photos, a neutral wardrobe and clean backgrounds where possible. What came back was a mass of photos, wildly varying in quality and composition. Some of them I genuinely thought there was no way we'd be able to work with. But with Runway, it was amazingly simple.

Runway helped us build the initial workflow, partnering closely with their forward deployed team. They basically gave us training wheels to get started. Once I had it, I started rendering each member myself, tweaking prompts and adjusting things until we had the right outputs. It was not long before I felt confident handling all the renders myself and then eventually, teaching my team how it worked.

We processed over 85 members through that workflow. The output is a still photo brought into motion: stable, TV-ready and consistent despite wildly inconsistent source material (UGC photos). The end card animates with all of them in the grid. It became this showcase of real members, at a scale that would have been extremely difficult and costly to execute with a traditional shoot.

Honestly, it started with the Runway team partnering so closely with us. That first workflow they built for us, just showing us what was possible, was the nudge we needed to move past "what can AI actually do" and into "here's something we're actually doing."

The technology matters, but because it is advancing so fast, what's been key for us is having a partner who knows how to implement the tech in real creative ideas.

I did the math. Getting even this partial approach done — 80-plus members, with the quality and volume we needed — would have cost around $300,000 the traditional way. That includes travel, studio time, crew, hair, wardrobe and production overhead. Aligning schedules for real people all over the country for the same date, logistically, would have been unrealistic to attempt. We simply would not have pitched or produced this scene practically.

With Runway, that same end card — real members, TV-quality, broadcast-ready — came in under $30,000, paid directly to our member talent with virtually no production overhead.

We're working on a few things. Concept storyboards have already changed how we get leadership approvals on integrated marketing communications (IMC) campaigns. We recently used Runway to storyboard a John Cena spot and an NBA campaign featuring Larry Bird and Cooper Flagg, and the approvals moved so much faster because leadership could see our ideas scene by scene. That use case is huge.

I am constantly looking for new opportunities to use AI, rethinking everything we do. From basics like building a brand photo generator using our brand guidelines and visual identity to scaled paid media resizes to adapting layouts intelligently across formats — not just stretching an image and calling it done.

The bigger shift is cultural. Every time one team does something with AI and shows it, three other teams want to do the same thing. We're in that phase right now where people are starting to see what's actually possible and rolling with it for their needs.

Opportunities to reinvent our production approaches are everywhere. Sometimes they are obvious, some opportunities need to be figured out, and sometimes you might just need to ask what’s possible. For us, we saw an opportunity, asked if AI could solve the problem and then ran with it. Once you see it work once — really work, for TV, not just a test — you'll start seeing those moments everywhere.

Fee-free digital banking platform serving millions of everyday Americans