Twelve months ago, a junior marketer at a typical SME might spend their week doing data entry, pulling basic reports, drafting social media posts, researching competitors on LinkedIn, and formatting email templates. That job, as it existed, is largely gone. Not because the intern was made redundant. Because a combination of AI tools now does those specific tasks faster, cheaper, and in most cases better.
This isn't a scary prediction. It's already happened. And for small businesses trying to compete with larger marketing budgets, it's genuinely good news.
Let me be specific about what AI handles well, because the conversation around this tends to be either breathlessly optimistic or apocalyptically grim. Neither is helpful.
Data entry and formatting. Loading contacts into a CRM, formatting spreadsheets, cleaning up import files, standardising job titles and company names. AI tools and automation handle this in minutes. A task that used to take someone an afternoon now takes a script thirty seconds. The quality is higher too, because machines don't misspell things at row 847 because they've lost concentration.
Basic research. "Find 50 companies in the West Midlands that offer accounting services and have between 10 and 50 employees." This used to mean an intern spending a day on LinkedIn and Companies House. Now it's a structured query that returns results in minutes. The research isn't better in every case, but it's dramatically faster for straightforward criteria-based searches.
First-draft copy. Social media posts, email first drafts, blog outlines, product descriptions. AI produces functional first drafts that get you 60% of the way there. I wrote about this earlier in the year regarding email marketing specifically. The output needs heavy editing, but the blank-page problem is eliminated. Instead of staring at an empty screen for twenty minutes, you're editing a draft within two minutes.
Scheduling and distribution. Posting social content at optimal times, scheduling email sequences, managing basic campaign logistics. All of this is now automated to the point where manual scheduling feels quaintly inefficient.
Report generation. Pulling data from multiple sources, compiling it into a structured format, identifying basic trends. AI and automation tools produce reports in minutes that used to take hours of spreadsheet work. The analysis is often shallow, but the data compilation is flawless.
Here's where the nuance matters. For every task AI handles brilliantly, there are others where it falls flat. And these are, without exception, the tasks that actually drive results.
Strategy. AI can tell you what happened. It can't tell you what to do about it. Deciding whether to target legal firms or financial services, whether to lead with price or quality, whether to expand into a new sector or double down on an existing one: these are judgment calls that require market understanding, client knowledge, and experience. AI provides data to inform these decisions. It doesn't make them.
Relationships. Nobody has ever built a commercial relationship with a chatbot. The trust that turns a prospect into a client comes from human interaction: a phone call where someone listened properly, a meeting where the salesperson understood the real problem, a proposal that felt personally crafted rather than template-generated. AI can help prepare for these moments. It can't replace them.
Understanding a specific market. AI knows about industries in general. It doesn't know that the procurement process at a particular NHS trust takes nine months, or that the MD of a target company just posted on LinkedIn about their expansion plans, or that a competitor quietly lost their biggest client last quarter. Market-specific intelligence comes from paying attention over time. AI processes information. Humans contextualise it.
Quality judgment. AI can write a blog post. It can't tell you whether that blog post is any good. It doesn't know that the tone is wrong for the audience, or that the example is outdated, or that the call to action feels pushy. Quality assessment requires taste, experience, and an understanding of the specific audience that AI simply doesn't have.
Creative direction. AI generates variations. It doesn't generate vision. The decision to reposition a client's messaging from "we're the cheapest" to "we're the most reliable" is a creative and strategic choice that comes from understanding the client's market, competitors, and customers. AI can execute that decision once it's made. It can't make it.
I'll be transparent about this, because I think the honest answer is more useful than the polished marketing version.
On a typical day, AI helps me with: drafting initial email copy that I then rewrite substantially; generating subject line variations for testing; cleaning and formatting data imports; producing report templates that I then populate with analysis; researching companies and contacts for prospect lists; and creating first-draft social content that gets edited before posting.
On a typical day, AI doesn't help me with: deciding which prospects to target; choosing the messaging strategy for a campaign; assessing whether a piece of content is good enough to send; understanding what a specific client actually needs; having conversations with prospects and clients; or making judgment calls about timing, tone, and approach.
The pattern is clear. AI handles the production work. Humans handle the thinking work. Both are necessary. Neither is sufficient alone.
This shift is disproportionately beneficial for small businesses. A large agency has always been able to throw bodies at production tasks: teams of juniors handling data entry, copy drafts, scheduling, and reporting. A small business couldn't afford that.
Now, a one-person or three-person operation can produce the same volume of output as a team of ten, because the production layer is handled by AI. The constraint has shifted from capacity to capability. The question is no longer "can we produce enough?" It's "do we know what to produce?"
That's a much better problem to have. And it's one that rewards experience, market knowledge, and strategic thinking over headcount.
The marketing intern isn't coming back. The tasks they did are permanently automated. But the skills that sit above those tasks, the strategy, the judgment, the relationships, the market knowledge: those are more valuable than ever.
If you're an SME spending money on marketing, the question isn't whether to use AI. You're already behind if you're not. The question is what to use AI for and what to keep human. Get that division right and you'll produce better work with less overhead. Get it wrong and you'll produce a lot of mediocre work very efficiently.
Speed without direction is just chaos with better tools.
Martin Dugan, AA2
