Stuck for a campaign idea? Use AI to find the angle
You are not out of ideas. You are too close to your own business to see it fresh. Here is how to use AI to find the angle you would never have reached alone.
You need a campaign. A promotion, a launch, a reason to email your list. And the harder you stare at the blank page, the more your brain offers up the same three tired ideas you have used before.
It is not that you are out of ideas. It is that you are too close to your own business to see it from a fresh angle.
You know your product so well that the interesting angles have gone invisible to you. That is a real and specific kind of stuck, and it is exactly where a bit of outside prompting helps.
The trap of asking for the answer
Most people open the model and ask for a campaign idea, get one bland suggestion, decide AI is not creative, and give up.
That is the wrong job to hand it. The model is not there to have your one good idea for you. It is there to flood the table with options so your own judgement has something to react to.
Reacting is far easier than inventing. You will know a good angle the second you see it, even when you could not summon one from nothing.
Ask for volume, not brilliance
So change the ask. Do not request a campaign. Request twenty angles.
Give it real material first: what you sell, who buys it, the season or moment you are aiming at, what you have tried before. Then say, give me twenty different angles for a campaign, and make them varied, some playful, some practical, some contrarian.
The first five will be obvious. Push past them. The interesting ones usually turn up further down the list, once the tired options are out of the way.
Then push on the good ones
Find two or three that make you sit up and ask the model to develop just those. Turn this angle into a full campaign, a hook, an offer, three emails, a couple of posts.
Now it is working for you properly, building out an idea you chose, rather than guessing at one for you. The taste stays yours. The legwork does not.
This is one of the few jobs where you want the opposite of caution. Temperature is the setting that controls how much a model wanders from the most obvious next word, and for facts you keep it low. For angles you turn it up, because a higher setting is what pushes it off the predictable path and towards the odd, surprising framing you would not have reached alone. Ask for fifteen or twenty angles rather than three, and the first few will be the tired ones everyone lands on. The gold usually sits from number eight onwards, once it has exhausted the obvious. You are not asking the model to be right, you are asking it to be plentiful, then you supply the taste.
Where it earns its keep
The point is not that the model is more creative than you. It is that it never gets tired, never gets precious, and never worries about looking silly.
It will cheerfully offer the daft angle that turns out to be the good one, the angle you would have talked yourself out of before saying it aloud.
What you get back
A campaign you are actually a bit excited about, found in twenty minutes instead of a fortnight of putting it off. And a habit you can repeat any time the well runs dry.
I work alongside these tools every day, and using AI to break a creative block is one of its most underrated jobs. It does not replace your taste. It gives it more to work with.
If you keep reaching for the same three tired campaigns, we can build you a repeatable way to find better angles fast. Book a quick chat and we will play with it.
Book a quick chat →Related: Better decisions from your own data.
Common questions
Can AI actually come up with good campaign ideas?
Not on its own, and that is the wrong way to use it. AI is poor at handing you one brilliant idea but excellent at generating twenty varied angles fast. You supply the taste and pick the good ones. Reacting to options is far easier than inventing from a blank page, which is where it helps.
Why should I ask for lots of angles instead of one?
Because the first few ideas any model offers are the obvious ones everyone lands on. Asking for fifteen or twenty forces it past the tired options, and the surprising, useful angles usually appear further down the list once the predictable ones are out of the way.
How do I get more unusual ideas rather than safe ones?
Give it real detail about your product, audience and moment, then explicitly ask for varied angles, some playful, some contrarian. If your tool lets you, raise the temperature or creativity setting, which pushes it off the predictable path towards the odd framing you would not reach alone.
What do I do once I have a shortlist of angles?
Pick the two or three that make you sit up and ask the model to build just those into full campaigns, with a hook, an offer, and a few emails and posts. That way it does the legwork on an idea you chose, rather than guessing at the whole thing for you.