In November, none of my clients mentioned it. By the second week of January it was on every agenda, usually framed as either an existential threat or a way to cut the content budget. Both readings are wrong, and I want to be precise about which parts of B2B marketing this genuinely changes — because the honest answer is narrower and more interesting than the discourse suggests.
What it genuinely changes
The cost of a first draft has collapsed. So has the cost of summarising a long document, restructuring an argument, generating twenty subject-line variants, or turning a rough transcript into something readable. These are real, immediate savings on work that consumed a meaningful share of a marketing team's week and produced no strategic value. If your team spends hours reformatting the same case study for four channels, that's now a much smaller job.
What it doesn't change at all
Knowing what to say. The model has no access to your customer calls, your win-loss patterns, your pricing conversations, or the thing your best customer said last Tuesday that reframed the whole category for you. It can produce a competent article about demand generation in any style you ask for. It cannot tell you that your ICP is wrong, because it doesn't know who your customers are or why the good ones bought.
- Judgement about what matters — unchanged, and now more valuable by comparison
- Proprietary insight from your own customers and data — unchanged
- The decision about what not to do — unchanged, and still the hardest part
- Distribution — a piece nobody reads is worthless however fast you produced it
The trap I can already see forming
The obvious move is to use this to publish considerably more. That will work briefly and then stop, because everyone in your category has access to the same capability and will make the same move. When the volume of competent, generic content rises across an entire category, the value of being competent and generic falls to nothing. The teams that benefit will be the ones who use the time saved on the parts that don't scale — original research, customer interviews, genuine opinion.
How I'm actually using it in client work
As a research assistant and an editor, not a writer. Summarising twelve customer interviews into recurring themes, then reading the source material myself to check what it missed. Pressure-testing a positioning statement by asking it to argue the opposite. Turning a messy workshop transcript into a structured first draft I then rewrite substantially. In each case a human decides what matters and the model removes the mechanical work in between.
This lowers the cost of producing marketing and raises the value of knowing what's worth producing. Those are opposite effects, and which one dominates in your company is a choice you're making this quarter.
Want to apply this to your business?
Book a free 30-minute strategy call and we'll diagnose your specific growth blockers together.
Book a Free Call