It’s nearly three years after ChatGPT launched, and about two since it’s become a household name. Hundreds of other tools have come on the scene since, and we’re moving beyond AI experimentation into practical application (yet we’re still cautious).
The question isn't whether AI can help marketers. It's how to use it strategically to enhance (not replace) human creativity and strategic thinking. Here's how our team at CID actually uses AI day-to-day, including the specific prompts, preferred platforms, and techniques that help us create better work.
Mackenzie (Account Executive): "I put my intake meeting notes into Claude to ensure our scope matches what the client actually asked for."
Specific prompt: "Here are notes from a call with [Client Name]. Please identify: 1) Their specific pain points mentioned, 2) Success metrics they defined, 3) Any constraints or requirements I should flag for the team, 4) Questions we still need to answer before scoping. Then summarize this in bullets I can share with our team."
Why it works: It keeps us focused on developing an accurate scope and approach by catching details that might get lost in conversation, and ensures we reach the right level of detail to create the thoughtful proposals we’re known for.
Heather (wearing her social media hat): "OpusClip has been a game-changer for repurposing our That's Notable Podcast content. I upload the full video and get 20+ short-form clips to choose from."
Process: "The AI identifies the most engaging moments, but I can adjust the cuts if they're not quite right, smooth out pauses and filler words, add captions, and even overlay our brand elements. Instead of going back to an editor with timestamps of clips I’d like, the tool helps me find them faster, and we save hours of manual video editing."
Platform preference: OpusClip specifically for video
Jade (Senior Digital Marketing Strategist): "I don't use AI heavily, but it's great for breaking through creative blocks. I use it more to rev up my brain than to generate final copy."
Specific prompt: "Write five meta descriptions for a [landing page/blog] about [topic] for a [type of company] in under 160 characters, including spaces."
Real example: "Write five meta descriptions for a landing page about brand strategy services for a mid-market B2B technology company in under 160 characters including spaces."
Process: "I'll look at AI's suggestions to get ideas flowing, then fully rewrite the content myself. For blogs, I sometimes paste in the first paragraph or full copy for context. But I'm always rewriting—never using AI output directly."
Why it works: Speeds up the process when you hit a creative rut. The AI suggestions spark different angles or emotional hooks that help you write better original content, without the risk of sounding obviously AI-generated.
Heather (wearing her content creation hat): "I consult AI as an editing assistant for tone and persuasion, especially when I've been staring at copy too long to see it objectively."
Specific prompt: "Read this landing page copy for our brand strategy service. The audience is marketing directors at B2B companies who've been asked by leadership to 'work on the brand.' I want them to book a discovery call. Identify places where the copy could be more direct, more persuasive, or better address their likely concerns about budget/timeline/buy-in."
Process: Similar to Jade, "I'm never using AI output directly as is. I'll look at each suggested edit line by line to determine if it makes sense to incorporate, and always rewriting myself."
Why it works: Gets an outside perspective on copy when you're too close to see weak spots, and helps identify emotional triggers you might have missed.
Neil (Senior UX Designer): "When developing personas or customer journey maps, I use AI to help synthesize research and identify gaps in my understanding."
Specific prompt: "Based on this customer interview data [paste transcripts], help me identify: 1) Emotional triggers at each stage, 2) Information needs we haven't addressed, 3) Potential friction points in the current experience, 4) Questions we should ask in our next research phase."
Platform preference: Claude for longer research synthesis, but also trying Perplexity to research industry-specific customer behavior patterns.
[Note: Here's a great resource on using AI for customer research from Andy Crestodina: How to Create an AI Marketing Persona: 8 Prompts for Deep Insights]
Heather (AGAIN): "I use AI to create subject line variations, especially when I’m creatively spent and need fresh perspectives."
Specific prompt: "Create 5 subject lines for our bimonthly (every other month) newsletter. This issue features an intro about [topic]. The audience is marketing leaders, primarily in the B2B space. Mix urgency, curiosity, and value-driven approaches. Keep under 50 characters for mobile optimization."
Process: "After getting AI variations, I run the top options through SubjectLine.com's free testing tool to see how they score. I take the results with a grain of salt because the tool doesn't know the full context of my email or my audience relationship, but it helps flag potential spam triggers or clarity issues I might miss."
Why it works: The combination gives you creative options plus objective feedback, while keeping your judgment as the final filter based on your audience knowledge.
Dylan (Director of Client Services): "I use AI to help structure proposals and identify questions clients might ask that I haven't addressed."
Specific prompt: "Here's our draft proposal for [Client Name] who wants to rebrand and launch a new website. Play devil's advocate: What questions might their leadership team ask that could derail approval? What objections should we preemptively address?"
Process: "After getting the devil's advocate feedback, I'll also ask AI to help me strengthen weak sections: 'This timeline section feels rushed. How can I better explain our phased approach to make it feel more strategic and less arbitrary?' It's like having a proposal review session before the real review."
Follow-up prompt: "Based on this proposal, what are three ways we could position our expertise that would differentiate us from typical agencies pitching similar work?"
Why it works: Catches blind spots before they become client concerns, and helps me think through positioning from angles I might not naturally consider. Better to address objections proactively than reactively.
Platform preference: Claude for longer document analysis and strategic thinking.
Travis (Development Team Lead): "AI is like having a smart search engine for debugging. We don't expect perfect drop-in solutions, but it's great for finding issues buried in walls of code."
Specific debugging prompt: "This JavaScript function is supposed to validate form inputs before submission, but users are reporting validation errors even with correct data. Here's the code [paste function]. What edge cases might I be missing? What would you test first?"
Process: "Sometimes the bug is trivial—a spelling mistake or misplaced line buried in a component. AI spots those faster than human eyes. We also use it for 'talking to the duck'—that debugging technique where explaining your problem out loud often reveals the solution. Instead of pulling a colleague away from their work, I explain the issue to AI and let that process help me think through it."
Why it works: Like any search result, we vet what AI gives us against our experience, then refine or correct until we're satisfied. It's not about perfect code generation—it's about faster problem identification.
Platform preference: Claude for code analysis; Copilot for autocomplete/refactoring.
Here's what we've learned after two years of daily AI use: the technology will only get better from here. As Andy Crestodina perfectly puts it, AI equals Average Information and that average keeps rising. The AI we're using today will seem dumb compared to what's coming.
But here's the critical point: never accept your tool's first answer as gospel. The biggest mistake we see marketers make is treating AI like the old days of "Google Whack"—trying to get the perfect result in one shot and calling it done. That's exactly backward.
The power comes from the conversation. Push back. Ask follow-up questions. Challenge assumptions. Request different angles. This is how you bring your strategic thinking to AI rather than letting it make you intellectually lazy.
Because let's be honest—AI can absolutely make people dumber if you let it. Research is already showing that over-reliance on AI assistance can lead to reduced problem-solving skills and decreased creative thinking. The marketers who will thrive are those who use AI to amplify their thinking, not replace it.
Our rule of thumb: If you're not iterating and refining, you're probably not getting AI's best work. And you're definitely not doing your best work.
Want to dive deeper? The Marketing AI Institute is hands-down the best resource for marketers serious about AI. Their free trainings, weekly podcast, and practical case studies will keep you ahead of the curve without the hype.
The future belongs to marketers who can marry human creativity and strategic thinking with AI capabilities. The question isn't whether you should be using AI—it's whether you're using it to become a better marketer or just a busier one.
Ready to develop an AI strategy that actually makes your marketing stronger? Let's talk about how to integrate these tools into your team's workflow without losing the human insight that makes great marketing great.
As CID's Executive Director of Marketing & Strategy, Heather Vaughn blends strategic thinking with creative problem-solving to get $h1t done. When she's not leading her high-performing team or geeking out on marketing automation, you might find her binge-watching Gilmore Girls or cuddling with her bulldogs.