I run Boise Marketing Guy as a team of one. Always have. But in 2026, that “one” includes AI assistance in ways that would have seemed impossible three years ago. I’m not talking about automation in the old sense, scheduled posts and email sequences. I mean real-time collaborative tools that handle the grunt work while I focus on the parts that actually require human judgment. This post is a field report on how I actually work now, what changed, and what didn’t. If you’re a small business owner trying to decide between a solo consultant and an agency, this is the behind-the-scenes view.
What can AI actually do for a solo marketing consultant?
AI handles three categories of work exceptionally well: first drafts, pattern recognition, and research synthesis.
For drafts, I feed Claude or ChatGPT a brief and get back a structured outline or rough copy in seconds. I used this approach for a Meridian HVAC client last month. I needed five blog posts on heat pump efficiency. I gave the AI my outline, key points, and voice samples. It generated drafts in 20 minutes. I spent two hours editing, fact-checking, and adding the specific Idaho context (winter performance at 5 degrees, Treasure Valley utility rebates). Total time: 2.5 hours instead of 8.
For pattern recognition, AI audits are faster than I am. I can feed it a competitor’s last 50 social posts and get back a breakdown of themes, posting frequency, engagement patterns. I use this when a Boise client asks why their competitor is outranking them. The AI spots the patterns, I interpret what matters.
For research synthesis, AI is a reading machine. I can dump 10 industry reports into Claude and ask it to extract trends relevant to Idaho professional services. It surfaces the signal, I provide the local application.
What AI doesn’t do: It doesn’t know which insight matters to the client’s specific situation. It doesn’t feel the room on a Zoom call. It doesn’t recognize when a strategy is technically correct but culturally wrong for the business.
What is AI still bad at in marketing work?
Three things: judgment, relationships, and original strategic thinking.
Judgment is the hard part. AI can generate 10 headline options, but it can’t tell you which one will resonate with a Nampa construction audience versus a Boise tech startup. That requires pattern recognition from years of watching what works in specific verticals and geographies. I’ve tested this repeatedly. AI suggestions are often generic best-practice, not context-specific best-move.
Relationships are entirely human. Clients hire me because they trust my read on their situation. That trust comes from listening well, remembering details about their business, and occasionally telling them their idea won’t work (and why). AI can’t do the Zoom call where a client is frustrated and needs someone to help them think through a pivot. It can’t read the subtext when they say “we need more leads” but really mean “our sales process is broken.”
Original strategy is still human territory. AI is a remix engine. It can combine existing ideas in new ways, but it can’t create a genuinely novel approach based on a deep read of a client’s competitive position and market moment. Last year I advised a Twin Falls manufacturing client to completely ignore digital ads and focus on LinkedIn thought leadership. That call came from understanding their long sales cycle, relationship-based industry, and the specific gap in their competitor set. AI would have recommended a standard B2B funnel because that’s what the training data says.
I’ve also found AI struggles with negative space, knowing what NOT to do. It tends toward more tactics, more content, more channels. Good consulting often means simplifying, not expanding.
How do you structure your work week with AI assistance?
My week breaks into three buckets: client-facing time, production time, and strategic work. AI helps most in production, a little in strategic work, and not at all in client-facing.
Monday and Tuesday mornings: Client calls, check-ins, strategy sessions. No AI involved. I’m listening, advising, course-correcting. I take handwritten notes during these calls because typing on a keyboard changes how I listen.
Tuesday afternoons and Wednesdays: Production time. This is where AI earns its keep. I batch content drafts, ad copy, audit work. I use Claude for long-form drafts, ChatGPT for short-form social copy. I feed it my notes from client calls, give it context, and let it generate first passes. I’m editing, not writing from scratch. This is the time I reclaimed. I used to spend 12 hours a week writing. Now it’s 5-6 hours of editing.
Thursdays: Strategic work. Planning, research, competitive analysis. AI helps with the research gathering, I do the synthesis and recommendation building. I’ll use Perplexity to pull together industry trends, then I write the actual strategic brief myself.
Fridays: Administrative, invoicing, following up on open loops. I also use Friday afternoons to test new AI tools and see what’s worth incorporating. Most aren’t.
The key shift: I can now handle 8-12 active clients instead of 4-6 because the production bottleneck loosened. But I’m not taking on more than 12 because the strategic and relationship work doesn’t compress.
What do clients actually see when working with you?
Clients see the same thing they always saw: a responsive consultant who knows their business and delivers useful work. They don’t see the AI assist, and they don’t need to.
Here’s what a typical client engagement looks like from their perspective: We have a kickoff call where I ask questions about their business, competitors, and goals. I send a strategic brief within a week outlining what I think will work and why. We agree on a plan. I deliver monthly reports, content, or ad management depending on the scope. I’m available on Slack or email for questions. We have a monthly check-in call to review what’s working.
Behind the scenes: The strategic brief is 100% me. The content drafts start with AI and get heavily edited by me. The monthly reports are formatted with AI help (I feed it raw data, it structures the narrative, I rewrite for clarity and add interpretation). The ad copy often starts as AI output that I punch up based on what I know converts for that client.
What matters to clients is quality of thinking and reliability of delivery. AI helps me deliver faster and handle more volume, but it doesn’t change the client experience. I’m still the one who answers the Slack message, takes the call, and makes the judgment calls.
I’m transparent about using AI when it makes sense. If a client asks how I turned around a draft so fast, I’ll say I used AI for the structure and first pass. No one has cared. They care that it’s good and it’s on time.
When does the solo-plus-AI model break down?
This model works for experienced consultants who already know what good looks like. It doesn’t work for beginners trying to scale before they have judgment.
I’ve seen newer marketers try to use AI as a substitute for experience. It doesn’t work. AI will confidently generate a content strategy that’s generic and misses the client’s actual needs. If you don’t have enough reps to recognize that, you’ll ship mediocre work and not know why it’s not landing.
The model also breaks when you try to push past 12-15 clients. At that point, the relationship management and strategic work becomes the bottleneck, and those don’t compress with AI help. You’re either underserving clients or burning out. I’ve tested this boundary. It’s real.
Another failure mode: over-reliance on the assist. I discipline myself to write from scratch at least once a week, usually a strategic brief or a tricky client email. This keeps my thinking sharp. If I let AI do all the drafting, my editing muscle gets lazy and I start missing things.
The model also assumes you’re working with clients who value strategy and judgment over pure execution speed. If your clients just want fast, cheap content, you’re competing with offshore teams and AI-native agencies that can undercut you on price. My clients pay for the thinking, not the typing. That’s the only way this works economically.
Finally, this isn’t a lifestyle business model. I work 45-50 hours a week. AI didn’t give me a four-hour workweek, it gave me a higher ceiling on client capacity at the same hours.
Should a small business hire a solo consultant using AI or go with an agency?
It depends on what you need and how you want to work.
Hire a solo consultant (like me) if:
– You want a single point of contact who knows your business deeply
– Your marketing needs are strategy-heavy, not just execution volume
– You value judgment and context over process and templates
– Your budget is $2,000 to $5,000 per month
– You’re OK with focused scope (3-5 active projects at a time)
The AI assist means I can deliver more per dollar than I could five years ago, but you’re still working with one person’s bandwidth and perspective.
Go with an agency if:
– You need a full-service team (design, dev, media buying, content, SEO all at once)
– Your project is large enough to justify account management overhead
– You’re spending $8,000+ per month
– You need 24/7 coverage or weekend support
– You want documented processes and team redundancy
Agencies have overhead I don’t have, but they also have specialist depth I can’t match. If you need a complex web dev project and simultaneous ad campaigns across four platforms, an agency is the right call.
For most Boise-area small businesses I work with, professional services, local retail, B2B companies under 50 employees, the solo model works better. They get senior-level strategy without paying for junior account coordinators and project managers. They get my cell number, not a ticketing system.
The AI assist just means I can serve them faster and handle a broader mix of tactical execution alongside the strategy. It’s not a different service, it’s a more efficient version of the same service.