How to Build a Content Engine as a Team of One
How to Use AI Throughout Your Content Development Process
DEAR STAGE 2: I'm a team of one in marketing and struggling to keep up with content demands. How can I use AI effectively without sacrificing quality and authenticity? ~CONTENT OVERWHELMED
DEAR CONTENT OVERWHELMED: This week, I'm handing the baton to Erin Olsen, Head of Marketing at Stage 2 Capital, to answer your question. Erin has built content engines in her past roles and has developed an AI-native process here at Stage 2 that's worth sharing. She brings a unique perspective as someone who has mastered creating high-quality content while being a marketing team of one. I asked Erin how solo marketers can use AI to keep up without lowering the quality bar — and here’s her advice:
The Team-of-One Challenge
When you’re the sole marketer, or a founder wearing the marketing hat, you’re responsible for everything: event planning, content strategy, production, email marketing, social media, digital, and analytics. Writing high-quality content often isn’t the issue—it’s finding enough hours in the day to consistently create and deliver it.
Even when early-stage teams understand the power of storytelling, content creation often falls to the bottom of the list amid all of the other go-to-market priorities.
After working with numerous startups in this position (and living it myself), I’ve developed a process that weaves AI into every phase of content development—not as a shortcut, but as a way to enhance a strong strategic foundation and help you move faster without sacrificing quality.
Here’s what’s working:
Strategic Approach to AI-Enhanced Content
My first piece of advice is don’t skip the strategy. The foundation of effective content is understanding your audience, goals, and key messages - no AI tool can replace this critical thinking. It can, however, enhance the process. Taking a moment to ask yourself, “who am I writing this for and why?” shouldn’t be overlooked.
What I've learned is that AI works best when it shadows your existing content creation process (and if you don’t have one yet, that’s ok too, we’ll get to that), jumping in to accelerate various stages along the way. Rather than asking AI to create complete content from scratch, I've found success in letting it enhance specific steps - helping ideate concepts, develop outlines, generate interview questions, and refine drafts.
Here's my step-by-step workflow that you can adapt to your own needs:
1. Start with Strategy (Human-Led)
Begin with the fundamentals:
Define your target audience
Clarify your content goals and key messages
Identify subject matter experts who will contribute the unique insights or POV
2. Initial Subject Matter Expert Discussion (Human-Led)
Schedule 20-30 minutes with your subject matter expert(s) to:
Brainstorm potential content angles
Discuss delivery formats (blog, webinar, case study, postcast, etc. etc.)
Record this conversation for later reference and potential repurposing
3. Story Angle Development (AI-Enhanced)
Using the transcript from your initial discussion:
Prompt AI to generate 3 different story angles based on the conversation
Specify: "Come up with three different story angles based on this conversation for [target audience]" You can even remind it of your full content strategy.
Add: "Give me direct quotes from the transcript that support each story angle" (this prevents the AI from making up story ideas that have no substance behind them.)
Constrain: "Do not use any content beyond what's in this transcript.) This is a big one. Keep reminding the AI not to make up stories or content that goes outside of what you’re feeding it.
You can get more strategic by asking AI to research how these angles differentiate from existing market content or might appeal to specific audiences.
4. Abstract Feedback & Selection (Human-Led)
Share the AI-generated abstracts with your subject matter expert to:
Get feedback on which direction best aligns with business goals
Select the winning abstract
Refine messaging if needed
These abstracts should capture the essence of your content and give you a pithy snapshot of what the reader (or viewers) will take away. You may find that multiple story angles will work from a single conversation, giving you a series of content.
5. Interview Guide Creation (AI-Enhanced)
For your chosen story angle:
Next, prompt AI to "Develop an interview guide with questions that will guide a discussion aligned with this abstract.” This is a HUGE time saver. Now that you have defined the topic and the story angle, let AI come up with the open-ended questions that will build out your story.
If you have multiple subject matter experts that are contributing to your content, you can ask AI to customize for different stakeholders while still ensuring that it aligns with the overall goal and abstract of the content: "Give me questions for a product leader" or "Give me questions for a customer."
Review and edit these questions based on your knowledge of the subject and speakers
6. Conduct & Record Interviews (Human-Led)
Share questions ahead of time with interviewees
Schedule 30-minute calls rather than asking for written responses when possible. I find that once you get people talking they will share a lot more information and context than they would if you asked them to simply respond to questions.
Always record these conversations (with permission) for quote accuracy and potential repurposing as audio/video content. Many times these recordings can be used for social content and video content that gets embedded in your articles.
7. Content Outline Development (AI-Enhanced)
Compile all your materials:
The selected abstract
Interview transcripts
Original question guides
Then prompt AI:
"Read through all of this content; I'll provide next steps after" (I find that this ensures that the AI takes the time first to consume the information your feeding it before jumping into creation mode.)
"Create an outline based on this information that aligns with our goals (and I would remind the AI of what those goals are again)."
"Include direct anecdotal quotes from speakers that support each area of the outline.” (This ensures that the outline is tied to the human conversation and not making up its own storyline.)
"Do not deviate from the transcript materials provided."
8. First Draft Creation (AI-Enhanced)
Once you've approved or edited the outline (this usually goes through a couple of iterations to get it just right):
Feed the approved outline back to AI
Request a first draft that follows the structure
For multiple interview sources: "Weave these three perspectives into one cohesive article"
9. Feedback & Revision Cycles (AI-Enhanced)
When receiving feedback from stakeholders:
Use AI to help interpret unclear feedback: "Help me understand what this feedback means" or “Suggest three ways that I could incorporate this feedback into the existing article.”
Ask AI for suggestions on incorporating specific feedback: "What's the best way to address this comment?"
Have AI implement edits directly when appropriate
10. Distribution Planning (AI-Enhanced)
For promoting your finished piece:
Create custom AI profiles trained on your company's or executives' social media style
Generate social posts that match your brand voice
Produce "bite-sized" excerpts to tease the full content
Tips for AI Success
Break tasks into specific prompts. Think of AI as a team member who needs clear direction. The more specific your prompts, the better your results.
Try multiple AI tools. Different tools have different strengths. I find Claude produces more elegant writing, while ChatGPT excels in other areas like strategy, and creating outlines. When you hit a wall with one tool, try another for a fresh perspective. You can also use different AI to edit your article.
Record everything you can. These recordings give you raw material not just for your primary content piece, but for social clips, audio snippets, and other promotional assets.
Remember your goal. As a team of one competing in a noisy market, your objective is to get high-quality messages out consistently. AI is a means to that end, not the end itself.
Until next week!