How Rare Bird Uses AI - Rare Bird, Inc.

For a client’s ebook I wrote a few years ago, I interviewed an expert who claimed that Microsoft Copilot was “already replacing marketing departments everywhere.”

He wasn’t a marketer, and his job generally required him to get enterprise clients excited about the future of AI, but his confidence in making such a bold declaration rested on some deeply ingrained (and incorrect) assumptions about what marketing is.

Meanwhile, the client’s marketing team remains cautious about the use of AI in their own work. Copilot certainly hasn’t replaced them, and they continue to hire us to strengthen their own efforts.

Rare Bird and AI: A Short History

Since the arrival of ChatGPT in late 2022, Rare Bird’s blog has occasionally featured AI-focused articles:

Early on, we were doing what most thoughtful people were doing: watching carefully, asking questions, and resisting the urge to either panic or evangelize. (Even still, I couldn’t resist referencing the killer robot trope a few times in that first AI-related post.)

A few of us began to dabble with AI. By 2024, we were ready to say plainly that businesses shouldn’t reject the use of AI outright. Our underlying questions or concerns hadn’t entirely disappeared, but we’d seen enough to know that fearing AI wasn’t a useful response.

The posts about AI since then have become more specific and, in some ways, more critical of AI and the discourse around it, in part because we began to encounter numerous assumptions about AI in client conversations. We pushed back on the rote recitation of AI-generated “best practices,” questioned the reliance on (and science behind) AI detection tools (spoiler alert: they don’t work), and took a hard look at what the rise of AI search might mean for SEO and content strategy.

Elevation, Not Replacement

What began as wonder has become a working familiarity with where AI earns its keep around here—and where it doesn’t.

An AI policy even made it into Rare Bird’s employee handbook, appearing right after the one about maintaining a non-smoking work environment, which tells you something about how seriously we take it, but also about our inability to properly alphabetize.

The gist of our approach is that AI should elevate our work, not replace it. That central idea shapes how we use it.

Now, AI has become at least a small part of how every discipline works within Rare Bird, but its role in our agency emerged organically as we individually found ways for AI to help us do our jobs a little better. There was never a top-down mandate from Jim saying we all “must” use AI, which is good; my research on the subject for various client assets has revealed that such organizational mandates tend to backfire.

Culture, Not Mandates

What does work? One approach is to create a culture where experimentation is not just permitted but modeled from the top. At Rare Bird, that starts with Jim. He hasn’t issued mandates or rolled out formal training programs. Instead, he’s demonstrated what a curiosity about AI looks like in practice. He built a custom morning briefing tool for himself that automatically scans his email, calendar, and Slack each day to surface what needs attention. He also used Claude Code to help integrate an API in Slack that re-orders office snacks from Amazon.

When the person running the agency is doing that kind of tinkering, it sends a clear signal that this is a place where trying things is OK, and where a learning curve is part of the job. Now we have a Slack channel specifically for sharing AI tools and articles with one another.

Here are a few examples of how the Flock uses AI in a given week:

Our senior videographer built a custom workflow that feeds auto-generated captions into Claude to check for grammatical errors, proper nouns, and the missing quotation marks that Adobe Premiere never adds automatically. What used to be a tedious manual process is now a fast, human-approved one.
The design team uses AI to extend or correct images when needed, research color palettes, and ideate on potential layouts. For them, AI may be a starting point, but it’s never the finish line. (Honestly, they’re still getting used to the idea that “AI” no longer means “Adobe Illustrator.”)
Our developers use AI to troubleshoot server issues, write complex database queries, and plan feature additions, including surfacing edge cases they might have otherwise missed. One developer runs finished code through AI to make it leaner and more modular before it ships.
Most of us now rely on AI assistance to draft or improve our communications, build spreadsheets, summarize long documents, and manage our schedules.
As Rare Bird’s writer, I first began using AI to try to find a way to carve out more time for the thinking and discernment good strategy demands. I rely on AI—mostly Claude—for help with research and synthesis across multiple documents, creating SEO metadata at scale, fact-checking on technical topics, and conducting content audits. I also use Otter’s AI to pull specific details from long calls, sometimes mid-conversation and sometimes weeks or months later to refresh my memory.

What’s Tedious, Not What Matters

In fact, the consistent pattern across our team is that we use AI to handle what’s repetitive or time-consuming, while the ~~Birds~~ humans handle what matters most. Nobody here ships AI output as finished work. Everyone remains in the driver’s seat. That’s not a rule we have to enforce, but it’s what responsible AI use at an agency looks like in practice.

Our AI policy also states: Employees are encouraged to stay informed about advancements in AI technologies and best practices for their responsible use. They should actively seek opportunities for learning and development to enhance their skills and knowledge in this area.

“Keep Learning” is one of Rare Bird’s five beliefs, so it’s no surprise to see it show up in our AI policy. What Jim wrote several years ago is now more important than ever:

We live and work in a world where change is constant. We need to stay abreast of those changes and ahead of the curve to remain valuable and relevant. There is only one way to future-proof yourself and Rare Bird, and that’s through continual improvement and constant education. Always keep learning.

AI may not replace entire departments just yet, but it has already changed where marketing teams focus their attention, shifting effort from production to strategy, from execution to judgment, from generating options to discerning which are the right options.

For ourselves and our clients, we want to make sure it’s changing things for the better. That means treating AI as a tool to amplify good judgment rather than a replacement for it, and holding ourselves accountable for the quality of what we put into the world.