The Content Contractor Crisis Isn’t AI. It’s Your Scope of Work.
By Lisa Peyton and her team of AI super heroes
I was in the middle of onboarding a new client when I came across a research paper that I couldn’t ignore.
The paper is called Payrolls to Prompts, and it analyzed real firm spending data from 2021 to 2025. The headline finding: companies didn’t just adopt AI — they used it to aggressively cut their freelance and contractor spend. The firms that had relied most heavily on platforms like Upwork and Fiverr shifted to AI the fastest. More than half of businesses using freelancers in 2022 have stopped entirely. The researchers called it “order-of-magnitude cost savings.”
My immediate reaction? That doesn’t match what just happened to me.
I had literally just landed a contractor gig — hired specifically to create AI-enhanced infographics for an edtech client. She didn’t go to an AI tool instead of me. She came to me because I use AI. I was faster than her usual designer, more affordable, and I brought over 20 years of strategic content experience to the work. AI didn’t replace me in this equation. It made me more competitive. It let me price the project in a way that worked for both of us, because what used to take days now takes hours.
So which story is true? The research, or my inbox?
Both, actually — and that nuance is exactly what most of the AI-and-freelancing conversation is getting wrong.
Here’s what the research is really measuring: the death of the task-based, commodity freelancer. The person selling “10 blog posts for $200” or “5 infographic slides for $50” on a gig platform. That work is getting automated away, and fast. But that’s not the only kind of freelancing that exists — it’s just the most visible and the easiest to track in a spending dataset.
What the data can’t see is the contractor who brings two decades of expertise, strategic judgment, and AI fluency to a client who doesn’t have any of those things in-house. That person isn’t getting replaced. That person is getting hired.
The question is whether your proposals and scopes of work are making that case clearly — or accidentally putting you in the commodity category even when you don’t belong there.
The Real Risk Isn’t AI. It’s How You’re Positioning Yourself.
The research is clear about what got replaced: discrete, well-scoped tasks. Writing a blog post. Designing a graphic. Fixing some code. The kind of work that fits neatly into a Fiverr listing — and, not coincidentally, the kind of work that shows up as a single line item in a poorly written SOW.
Here’s a real example. I’m currently working with an edtech client on a series of AI-generated infographics. On the surface, this looks like exactly the category of work the research flags as at risk. But my client didn’t go to AI tools like Canva or Midjourney herself. She hired me instead, because I can do it faster, smarter, and with better strategic judgment with the help of AI.
The AI isn’t replacing me here. I’m the AI-augmented layer she’s paying for.
But — and this is critical — if my SOW just said “10 infographics = $X,” I would have commoditized myself right back into the at-risk category. I would have made myself look like the thing the research says is getting cut.
Your SOW is either your best argument for your value, or it’s quietly making the case that you’re replaceable. Most freelancers are writing the latter.
What a Strategic SOW Actually Looks Like
Whether you’re doing AI-enabled content creation, building agentic workflows for customer service, or designing sales enablement systems — the fix is the same. Stop selling the output. Start selling the thinking.
Here’s how.
Unbundle your phases explicitly. If your SOW has one line that says “Project deliverables: X,” you’ve already lost. Break your work into named phases — Discovery, Strategic Architecture, Production, Optimization — each with its own description and fee. When a client can see that the deliverable is the last step, they understand they’re paying for everything that made it possible.
Write outcomes, not activities. “Review client materials” is an activity. “Stakeholder and workflow analysis to identify automation opportunities and failure risks before build” is a strategic outcome. Same work, completely different perceived value. Every line item in your SOW should answer the question: what problem does this solve for the client?
Add a Strategic Assumptions section. This is the most underused tool in a freelancer’s SOW arsenal. List every judgment call you’re making on the client’s behalf — audience analysis, tool selection rationale, risk assessment, integration considerations, escalation logic for agentic systems. When clients see that list, two things happen: they realize how much invisible thinking they’re getting, and they stop thinking they could’ve just done this themselves.
Document your AI decisions. This is counterintuitive, but transparency about your AI process actually signals expertise rather than undercutting it. Which tools you chose and why. What you prompted, what you rejected, and what you overrode. For agentic workflow builds especially, this curation and quality-control layer is high-skill judgment work — and it’s billable. Name it. Charge for it.
The Question Every Freelancer Should Ask Before Submitting a SOW
Read it back and ask yourself: after reading this, could my client reasonably think they could’ve just done this in ChatGPT or Canva?
If the answer is yes, rewrite it.
Your SOW should tell the story of your process so clearly that no one — not your client, not her boss, not a procurement person reviewing the invoice — could mistake what you do for something a $20/month subscription could replace.
The research shows companies are cutting freelancers who sell tasks. They’re keeping — and hiring — the ones who sell thinking.
Make sure your paperwork knows the difference.
Ready to Put This Into Practice? Build the Tool.
If you’re an advanced AI user — and if you’re reading this newsletter, you probably are — here’s your next move: don’t just read this framework, automate it.
Build yourself a Claude-powered SOW assistant that takes either a rough project description or an existing scope of work and transforms it using the principles above. Prompt it to unbundle phases, rewrite activity-based language into outcomes language, generate a Strategic Assumptions section, and flag anywhere the document undersells your expertise.
You already know how to do this. The prompt architecture isn’t complicated — what matters is that you build it around your practice, your typical project types, your client language. A generic SOW template helps no one. A tool that thinks the way you think, trained on the principles that reflect your actual value? That’s a competitive asset.
This is exactly the kind of upstream AI application that separates strategic practitioners from people who are just using AI to go faster. Use it to go smarter.
I’ll be sharing my own version of this tool in an upcoming issue — but I’d encourage you to build yours first. The act of defining what belongs in a strategic SOW, in your own words, is half the value.
Lisa Peyton is an AI Communications Strategist, Fractional Chief AI Officer, and faculty at the University of Oregon’s School of Journalism & Communication. She publishes the AI Marketing Brief and helps marketing teams move from AI-curious to AI-dominant.
