Every few months someone announces that AI has changed marketing forever and the old rules are dead.
I've read the decks. Under the gradients and the robot illustrations, it's always the same thing: product, price, place and promotion. The four Ps, from a textbook older than most marketing directors, wearing a hoodie.
That's not a complaint. It's good news. If the fundamentals still hold, you don't need a new strategy for AI. You need to know where AI makes your existing strategy faster, cheaper or sharper. So let's go through them.
Product: fix the thing before you sell the thing
Marketing loves to skip this one, because it's awkward to tell the product team their baby is ugly. AI makes that conversation easier because you can bring receipts.
- Cluster support tickets, reviews and churn reasons into the five things people actually complain about.
- Compare your feature list against how customers describe what they use you for. They are rarely the same.
- Draft the "what's new" messaging the minute a release note exists, not two weeks later.
What AI can't do here is decide what to build. That still takes someone who has talked to customers and is willing to say no.
Price: test the story, not just the number
Pricing is one of the few levers that moves revenue overnight, and most teams barely touch it.
AI won't set your price. It is good at the work around it: summarizing competitor pricing pages, drafting three versions of the pricing page copy so you can test the framing, and reading sales call transcripts for the moment the price comes up and the mood changes.
The number is a business decision. The way you explain the number is marketing, and that's where AI earns its keep.
Place: be where they already are
Place used to mean shelves and distributors. Now it mostly means channels, and increasingly it means AI search. When someone asks a chatbot for "the best property portal in Denmark" or "an agency that does HubSpot migrations," are you in the answer?
That's a real shift, and it's one I've been working on with clients. The old-school fix still applies, though. Be the clearest, most consistent, most cited answer to the questions your customers ask. Good content, structured properly, said the same way everywhere. Search engines rewarded that. So do the models.
Promotion: where everyone starts, and should finish
This is where 90% of AI marketing effort goes: more posts, more ad variants, more emails. It's also where it matters least if the other three Ps are a mess. Promoting a confusing product at the wrong price in the wrong place, just faster, is not a strategy. It's a more efficient way to lose.
Done in the right order, though, AI is great for promotion:
- Versioning one strong idea across every format and market.
- Generating test variants so your ad budget learns faster.
- Scoring and routing the leads all that promotion brings in, so sales calls the right people first.
The boring conclusion, which is also the useful one
If you want an AI marketing strategy, don't start with tools. Get out the four Ps, write down where each one is weak, and ask where AI shortens the path to fixing it.
You'll end up with a shorter tool list, a clearer plan and a deck with a lot fewer robot illustrations. Your CFO will love it. Your LinkedIn feed won't, but your LinkedIn feed was never going to sign the purchase order.