You can use AIDA to write a draft or to check one. Checking is the better use. As a template, the four steps (attention, interest, desire, action) pull every draft toward the same shape: a hook, a build, a payoff and an ask. As a checklist, they show you which part of a draft is missing, which is padded and which is already doing the work.
What AIDA is
AIDA is a four-step model of how a persuasive message works on a reader: win attention, hold it by being relevant (interest), build the wish to act (desire), then prompt one specific action. It dates to the late 1890s and began as a description of selling, not a formula for writing posts.
Elias St. Elmo Lewis is usually credited with the earliest form, around 1898, and Edward Strong revised and promoted it in the 1920s. The aim was to help marketers capture a likely buyer’s attention and lead them toward a purchase.
That origin matters for B2B content. A buyer in a sales conversation expects to be asked for something. Someone scrolling a working feed isn’t there to be sold to, so the last step needs more care than the model gives it.
A model of advertising effects, not a law of reading
AIDA belongs to a family called hierarchy-of-effects models, which propose that audiences move through thinking, feeling and doing stages in order. Researchers have argued over the evidence for that sequence for decades. Barry and Howard reviewed the literature in 1990 and critiqued the case for it as an advertising model.
A later study, summarised in an abstract from the Erasmus Research Institute of Management, compared 26 proposed models using brand data from fifteen categories over seven years. It reports that sequence does matter, but the order varies by brand. Feeling before thinking was the most common.
Those studies look at advertising and brand perception. They don’t test LinkedIn posts, so they can’t say whether a post should run attention, interest, desire, action in that order. Our reading is narrower: readers don’t reliably move through four tidy stages, so the framework works better as a set of questions than as a script.
Run four questions on a finished draft
Write the post the way you’d explain it to a colleague. Then check each step.
Attention: does the first line carry information?
A post’s preview cuts off after the opening lines, and where it cuts depends on the device. So read your first two lines on their own. If they only promise something (“Here’s why most teams get this wrong:”), the reader has nothing yet. A line that states the observation gives them a reason to open the rest. Attention won by withholding has to be paid for later, and many drafts never pay.
Interest: does the next paragraph show the reader their own week?
Interest is recognition more than information. A marketer reading about approvals should think “that’s us” within a few sentences. Cut anything that explains what they already know. If a paragraph could sit in any post about any team, it isn’t earning interest.
Desire: do you have the proof this step asks for?
In B2B content, desire rarely means wanting a product. It means wanting to try the idea. That takes evidence: an example, a before and after from real work, a number you can trace to a source. If you don’t have one, say less. A precise-looking statistic with nothing behind it is the most common filler here, and a reader who can’t trace it has little reason to believe the rest.
Action: what should the reader do, and have you earned it?
Sometimes the answer is nothing. Reading to the end is an action. When there is an ask, make it one thing, keep it in proportion to what the post gave, and put it after the value. An ask that arrives before the reader has received anything reads like a toll booth.
A weak opening, run through the four questions
Take a composite draft that opens: “Most teams get LinkedIn approvals wrong. Here’s why:”
That fails the first-line test. It announces an answer without giving one. A version that states the observation works harder: “One LinkedIn post can pass through three people and end up in four places: the draft, a chat thread, an email and a screenshot.”
That line also does the job of interest, because a marketer recognises the week it describes. The draft’s next paragraph, which defines what an approval is, can go. The reader does this every week.
Further down, the draft claims teams halve their approval time. Nothing supports it, so it comes out, or it’s replaced with an example from real work. The close, “Comment APPROVE and I’ll send the checklist,” asks for effort before delivering anything. Put the checklist in the post, or drop the ask.
When to leave AIDA alone
Under a few dozen words there’s no room for four steps, and forcing them makes a post sound like copy. Write the observation and stop.
An opinion piece needs a claim someone could disagree with, then the reasons the accepted view is wrong, the evidence and what follows. AIDA gives it a marketing arc with no argument inside.
And if your readers already know the problem well, a hook that describes it back to them wastes the first line. Open on what changes instead.
If the idea is finished in sixty words, stop at sixty.