A copy framework is not a creativity shortcut. It is a running order. It decides which sentence goes first so that the reader gets the reason to care before they get the reason to buy, which is the sequence almost every failing ad has backwards.
Three frameworks cover nearly all of it: PAS, AIDA and BAB. This page gives each one as a fill-in template, a worked example on a real Pakistani offer, and the conditions under which it beats the other two.
One thing to settle first. None of these frameworks makes a weak offer work. If the price is wrong, the product is wrong, or nobody wants the thing, better sentence order changes nothing. Frameworks make a good offer legible.
The three frameworks, compared
| Framework | Sequence | Leads with | Strongest on |
|---|---|---|---|
| PAS | Problem, Agitate, Solution | A frustration the reader already has | Cold traffic, impulse purchases, pain-led products |
| AIDA | Attention, Interest, Desire, Action | A hook or a claim | Considered purchases, services, higher price points |
| BAB | Before, After, Bridge | The reader’s current situation | Visible transformations, courses, skincare, fitness |
The practical difference is where the reader is when they meet the ad. Cold traffic has no reason to read a second line, so PAS earns attention by naming something already annoying them. Warmer traffic will tolerate a longer argument, which is what AIDA is for. BAB is the odd one out: it works on both, provided the outcome is something you can show.
PAS: Problem, Agitate, Solution
Name the problem, make it concrete, then resolve it. Three moves, no preamble.
The mistake in PAS is skipping the agitation. Writers name the problem and go straight to the product, which reads as a sales pitch. The agitation line is what turns a general problem into the reader’s own problem, usually by adding a specific consequence rather than more adjectives.
“Your ads are getting orders. Half of them are coming back.”
“Every COD return costs you the courier both ways plus the handling, so two returns wipe out the profit on ten delivered orders.”
“We photograph the stock you actually hold, in daylight, so the shade the customer sees is the shade that arrives.”
“Send us one product. We will shoot it and you can compare the return rate yourself.”
Notice that the agitation line does the arithmetic for the reader. “Returns are expensive” is a claim. “Two returns wipe out ten orders” is a calculation the reader can check against their own numbers, and it is the line that makes them read the third one.
PAS also compresses well. If you only have one line of visible text before the More link, PAS gives you the strongest possible opener, because the problem statement is the part that stops the scroll.
AIDA: Attention, Interest, Desire, Action
The oldest of the three, and the right one when the reader needs to understand something before they can want it.
AIDA gets criticised as generic, usually because people write the Interest section as features. Interest is not a spec list. It is the part where you explain how the thing works well enough that the claim becomes believable, and Desire is where the working mechanism turns into an outcome the reader wants.
“Sixteen weeks. One real website. Not a single slide about theory.”
“You pick a site in week one and work on it every week: keyword research, on-page fixes, technical audit, then reporting in Search Console.”
“By the end you have a live site with measurable movement, which is the only thing a client or an employer asks to see.”
“Next batch starts 1 October. Fifteen seats.”
The deadline in the Action line is doing real work here, and it is honest: the batch genuinely starts on a date and the seats are genuinely limited by classroom size. Manufactured scarcity is a different thing and audiences in this market have seen enough of it to discount it automatically.
AIDA is the framework to reach for when the price is high enough that the reader will not decide from the ad alone. Its job is not to close, it is to make the click informed.
BAB: Before, After, Bridge
Show where the reader is, show where they could be, then explain the crossing.
BAB is the most visual of the three, which is why it dominates categories where the result photographs: fitness, skincare, home renovation, education. The copy and the creative can carry the same structure, before on the left, after on the right, and the bridge in the text.
“Someone searches for your shop by name and gets a map pin with no photos, no hours and one old review.”
“Same search, three months later: photos of the shop, correct hours, twenty reviews, and directions that actually work.”
“We fix the profile, get the categories right, and set up a review process your staff can follow without being reminded.”
“Message us your business name and we will tell you what is missing.”
The Before line has to be recognisable rather than insulting. “Your profile is a mess” puts the reader on the defensive; describing what a customer sees puts them on your side of the problem. That distinction is most of the skill in BAB.
Which framework for which situation
| If the situation is | Use | Because |
|---|---|---|
| Cold traffic, low price, impulse | PAS | You have one line to stop the scroll and a problem statement does it |
| Cold traffic, high price | AIDA | The reader needs the mechanism before the outcome is credible |
| Retargeting, cart abandon | PAS, shortened | They already know the product; name the hesitation instead |
| Visible transformation | BAB | Copy and creative can share one structure |
| Service business, local | BAB | The before state is something the client already suspects |
| B2B or long sales cycle | AIDA | Desire has to be built, not assumed |
| A genuine deadline or launch | AIDA | The Action slot is where a real date does the most work |
One caveat on that table: it describes where to start, not where to stay. Audiences tire of a structure the same way they tire of a creative, so a framework that won three months ago is worth re-testing rather than treated as settled. The ordering logic is durable; the specific winner is not.
Retargeting needs a fourth version
Someone who already saw the product does not need the problem explained. They need the hesitation named.
The standard mistake in retargeting is running the cold-traffic ad again with a discount stapled to it. The reader has already read that argument and declined it, so repeating it louder is a poor use of the impression. What changed between the first view and now is not their understanding of the problem, it is that something stopped them.
So the retargeting version is PAS with the Problem slot replaced by the objection. Instead of “your returns are costing you”, it becomes “you looked at this and did not order, and the reason is probably the price or the delivery time.” Naming the hesitation out loud is unusual enough in this market that it earns the read on its own.
There are usually only three or four real objections per offer: price, trust, timing, and fit. Write one retargeting ad per objection rather than one generic reminder. Price gets a payment-plan or comparison line, trust gets proof, timing gets a deadline that is true, and fit gets a specific “this is for you if” qualifier that also tells the wrong buyer to leave.
That last part matters more than it sounds. An ad that openly disqualifies people converts better on the ones who remain, because the reader who fits recognises themselves in the exclusion. Copy that tries to include everyone reads as addressed to nobody.
Mapping copy into the ad fields
A framework written as four lines has to survive being split across three fields with different limits.
On Meta, primary text shows roughly 125 characters before truncating, the headline shows around 27 on mobile feed, and the description often does not show at all. So the framework does not map one line per field. It maps like this:
- Primary text. The whole framework, first line carrying the Problem or the Before, because that is the visible portion.
- Headline. The Solution or the After, compressed to about 27 characters. This is the line beside the button, so it should read as the promise, not as your brand name.
- Description. Treat as optional. If it shows, use it for the qualifier: price, timeframe, or coverage area.
- Call to action button. Match it to the Action line. A “Send Message” button under copy that says “order now” loses people at the last step.
If the copy is going into Google Ads instead, the frameworks still order the argument but the fields are tighter: 30 characters per headline and 90 per description. In practice you get the Problem or the After into a headline and the mechanism into a description. There is more on that in our Google Ads service.
Writing for a Pakistani audience
Language choice changes performance more than framework choice does.
Roman Urdu tends to outperform English for price-led offers to a broad domestic audience, because it reads as someone speaking rather than a brand announcing. English holds better for B2B, software, and anything positioned as premium, where Urdu can accidentally signal cheapness.
Mixed copy works in one specific arrangement: English structure with the emotional line in Urdu. The Problem or the Before in Urdu, the mechanism in English. What does not work is half-translating everything, which produces copy that is fluent in neither and reads as machine output.
One more local specific. Price transparency outperforms “DM for price” on cold traffic almost every time. Hiding the price filters out buyers along with browsers, and it trains the audience to expect a negotiation, which slows every subsequent conversation.
Five mistakes
- Starting with the brand. “At [Company], we believe” spends the only line you are guaranteed on information the reader does not need yet.
- Skipping the agitation in PAS. Problem straight to Solution reads as a pitch. The middle line is what makes it personal.
- Features in the Interest slot. A list of specifications is not interest. Explain the mechanism instead.
- An After that is not measurable. “Grow your business” is not an After. “Twenty reviews and correct hours” is.
- A CTA that does not match the button. Copy and button have to ask for the same action, or the click hesitates.
How to test them
Three ads, three frameworks, one offer. That is the first test.
- Write the same offer three times, once as PAS, once as AIDA, once as BAB. Do not polish any of them yet.
- Use identical creative across all three so the copy is the only variable.
- Run them in one ad set, not three, and let the budget find the winner rather than splitting it evenly.
- Judge on cost per outcome, not on clicks or engagement. A framework that gets cheap clicks and no orders has been rewarded for being interesting.
- Once one framework wins, vary sentences inside it. That is when line-level testing starts earning something.
- Rewrite the losers against the winner rather than deleting them. A losing framework often wins on a different audience.
Creative and copy fail together more often than separately, so run this test on assets that are correctly sized and captioned to begin with. There is a full reference on that in the Meta ad sizes guide, and if the production side is the constraint, UGC ads is the service that covers it.