Here’s a small test you can run today.
Open ChatGPT, Claude, or Gemini. Type in “write a press release announcing our new product.” Give it a product name and three bullet points.
Thirty seconds later, you’ll have a press release.
Now ask yourself a different question: Should you actually send it?
That’s the gap we want to talk about here. Because the gap between “can produce a press release” and “can produce a press release that moves the business forward” is the entire reason agencies still exist.
Output vs. Outcome
Marketing produces two kinds of things: outputs and outcomes.
Outputs are the artifacts — the emails, the posts, the landing pages, the pitches, the decks, the releases. You can count them. You can stack them. You can point at them and say, “Look, we’re doing work.”
Outcomes are the business results the outputs are supposed to create — a qualified lead, a closed deal, a placed story, a shifted perception, a recovered reputation. You can’t just generate these. You have to engineer them.
AI is phenomenal at outputs. It is dramatically weaker at outcomes — because outcomes require judgment, not just language.
The Questions AI Can’t Answer
Take that press release. Before we’d ever send it, here are the questions a good agency asks:
- Is this newsworthy to the reporter we want to reach, or just to us?
- Is there a better angle hiding in here — a data point, a customer story, a counterintuitive claim — that would actually make it land?
- Who, specifically, should receive it? Not a list — a person, with a name, who covered something adjacent last month.
- What’s the timing? Is there a news cycle we can ride or a competing story we should avoid?
- What’s the headline that makes an overworked editor open the email instead of deleting it?
- If this gets picked up, what’s our follow-up? Who’s ready to comment? What’s the next story?
- And — honestly — should we even send this? Is this actually the best use of our credibility with these reporters right now?
Those are strategy questions. Not generation questions.
AI can help you write any of them faster. But it can’t decide which answer is the right one for your business this quarter, against this competitive backdrop, in this market.
Where the Real Value Shows Up
Here’s how this plays out in practice. We recently worked with a client who wanted to announce a leadership change. The obvious move was a press release. The AI-assisted first draft took about twenty minutes to generate.
The better move, it turned out, was to hold the press release and instead offer an exclusive to one trade publication that had been covering the company’s category closely. The result: a feature-length piece with a photo, a quote from a customer, and a headline that positioned the leadership change as part of a broader growth story — not just a personnel update.
Same underlying information. Radically different outcome.
No AI tool in the world was going to make that call. That came from a human who knew the reporter, knew the publication, knew the client’s strategic priorities, and had the relationship to pick up the phone and pitch the exclusive.
The “Just Use AI” Trap
There’s a version of this story we see a lot. A small business owner tries to save money by having a staffer (or an intern, or themselves) generate marketing content with AI. It works, sort of. The content goes out. Some of it performs. Most of it doesn’t. Nobody can quite tell why.
Six months later, they realize they’ve been busy producing content without actually moving the needle on anything. No pipeline lift. No brand recognition change. No meaningful media coverage. Just a lot of output.
That’s the trap. Output feels like progress. And when you can produce it infinitely for free, it’s very easy to confuse motion for momentum.
What to Ask Yourself
If you’re evaluating whether your marketing is actually working, or whether it’s just happening, here are three honest questions:
- Can you point to an outcome — a lead, a deal, a story, a change in perception — that directly traces back to your marketing in the last 90 days?
- If that outcome happened, was it because of the specific way the work was executed, or would it have happened anyway?
- If you took all the marketing content you produced last quarter and deleted it, what would actually be worse?
If the answers make you uncomfortable, you don’t have a content problem. You have a strategy problem. And more AI won’t fix it.
The Bottom Line
Generation is cheap. It has never been cheaper. That is exactly why strategy has never been more valuable.
An agency worth hiring today isn’t one that produces the most content. It’s one that knows which content is worth producing, why, for whom, and what it’s supposed to do for your business.
That’s the difference between output and outcome. And it’s the difference between spending money on marketing and actually getting something back for it.
Up next in the series: what three decades of client work taught us that no model can learn. In the meantime, if you want to talk about whether your marketing is producing outputs or outcomes, give us a call.
Frequently Asked Questions
What's the difference between marketing output and outcome?
Output is the artifact — the post, the email, the press release. Outcome is the business result it's meant to create: a lead, a placement, a closed deal. AI is excellent at producing outputs and weak at engineering outcomes, because outcomes require judgment, not just language.
Can AI decide what content my business should produce?
It can suggest options, but it can't make the strategic call about which story to tell, to whom, and when, against your specific market and competition. That decision depends on context AI doesn't have and consequences it doesn't bear.
How do I know if my marketing is producing outcomes?
Ask whether you can trace a lead, deal, story, or perception shift directly to your marketing in the last 90 days — and whether it happened because of how the work was executed. If you can't, you likely have a strategy problem, not a content problem.
Isn't cheap, fast content still worth it?
Only if it moves the business. Endless output can feel like progress while changing nothing. The value is in producing the right content, for the right reason, which is where experienced strategy earns its keep.