Marketing Angles: Find a New Reason to Buy Before Writing Another Hook

The account needs new creative, so somebody opens a doc and writes twenty hooks. And they’re good hooks, too. There’s a question, a bold claim, a “nobody talks about this” and one that starts with a guy dropping his phone in a lake.

Then you read past the first line of each ad, and it turns out you’ve built twenty front doors onto the same house. Every one of them goes on to say the same thing. (Yes, even the lake guy.)

Marketing angles are your golden ticket out of that loop.

An angle is the reason a particular buyer would care about your product in a particular situation.

A hook only opens that argument, which is why twenty new openers can leave you with the same single pitch you had this morning.

You’ll leave with a matrix for your own product that pairs each buying situation with a reason to buy, the evidence you need and an opening to test.

Let’s rock n’ roll.

IWhat’s the difference between an angle, a hook and an execution?

Marketers toss these three words into one bucket like loose change, and that’s how a team ends up “testing new angles” by changing the first sentence.

Simply put:

The angle is the reason to buy.

It names a buyer, the situation they’re in and why your product matters to them right then. Change the angle, and a whole different argument shows up to the party.

The hook is how you open that argument.

It’s the first line or the first frame, and its whole job is to earn the next few seconds so the reason gets heard. Change only the hook, and the same argument walks in through a different door.

The execution is how the whole thing shows up.

That covers the format, the person on camera, the setting, the edit and whether it’s a video or a static image. Change only the execution, and the same argument just got a new hat.

Now, of course, all three are worth testing. They just answer different questions, so it helps to know which one you changed before you read the results.

If you want the platform side of this, our Andromeda guide covers what Meta has said about creative diversity. Everything below is just how we brief at Coppica, so Meta’s system doesn’t come into it.

IIWhy can six hooks still be one ad?

Let’s make up a product so we can pick on it. (The product is invented, and so is every ad line for it in this guide.)

It’s an online Spanish tutoring service that sells live 25-minute lessons with a human tutor, and the team just wrote this batch:

1. “What if you could hold a real conversation in Spanish by summer?”
2. “I went from ‘hola’ to full conversations in 90 days.”
3. “Stop spending years on Spanish. Give it 12 weeks.”
4. “Nobody tells you how fast Spanish goes with a tutor.”
5. “It’s week twelve, and you just ordered dinner in Spanish.”
6. “The 90-day Spanish plan your old textbook never mentioned.”

That’s a question, a testimonial, a command, a secret, a little scene and a dig at a textbook. It looks like a full buffet, and it reads like a full week of work.

Now try the reason-to-buy test. Cover the first line of each ad and write one sentence that says why this person should buy. For this batch, you’ll write “you’ll be conversational fast” six times in a row.

So the team has six hooks and one angle.

The buffet has six trays, and every one of them is chicken.

One angle is fine... while it’s working... and a strong reason totally deserves more than one opener. It starts to fall apart when that angle... speed of learning... only matters to some of your buyers, because then you’re just fishing in one corner of the pond.

Everybody else has now heard about those 90 days six times and still hasn’t heard anything that would actually make them buy the thing.

IIIWhat can two old ads teach you about finding a reason to buy?

Before we fix the Spanish batch, look at two real ads that The Persuasion Taxonomy keeps on file. Both of them started from the buyer’s situation, and neither one started from the product.

Schaefer picked an evening, and left “best beer” alone. Its 1950s line was “Schaefer is the one beer to have when you’re having more than one.”

The Taxonomy files it under Occasion Framing (PT-RFM-9722).

The situation is a long night with several rounds, and the reason to buy is that this is the beer for that kind of night. It’s also a pretty fun pun... the one to have when you’re having more than one... Schaefer never has to climb into the ring and fight every other beer for the title, because it only asks for one moment the drinker already knows.

Postum went looking for the person who already blamed coffee. Postum was sold as a coffee substitute, and its 1940s copy said, “So if you’re nervous or irritable, if you can’t sleep, or have nervous indigestion, and you think coffee-nerves may be responsible”.

The Taxonomy files that under Switching Trigger Identification (PT-IDF-9014).

The situation is somebody lying awake who has started to suspect the coffee, and the reason to buy is that they want a replacement right now. Anybody who feels fine about coffee just keeps reading the paper, which is exactly how the ad was built to work. It’s a key cut for one lock.

It’s easy to see how their “reason to buy” was different than what everyone else was saying.

IVWhat happens when a real page changes its reason to buy?

This next one comes with numbers, because our founder ran it. In 2015 he rewrote the headline for Smart Podcast Player, Pat Flynn’s podcast plugin, and Pat’s team split tested it.

The page opened with “Grow Your Podcast Engagement Like Never Before.” Its reason to buy was growth, meaning more traffic, more downloads and more subscribers. That’s a promise podcasters want, and it’s the one most of us would have guessed.

Then he read the customer survey and counted. He sorted the answers by what people said they liked about the player and what they hoped to gain from it, and the tally came out like this.

The page was selling the reason that came up three times. The other forty-two mentions were about a player that looked great and took no effort, and one survey answer just said, “To be cool.”

So the new headline argued that reason instead:

“The Easiest to Use, Best Looking Podcast Plugin Ever Made.”

Growth moved down into the line underneath.

Pat’s team stopped the test early because the result was so lopsided. Click-through on the header button went from 10.10 percent on the old page to 16.97 percent on the winner, which is 68.1 percent higher.

Why did it work so much better?

Nobody had to brainstorm the winning reason, because it was already sitting in the survey waiting for somebody to count.

So how do you apply this angle switcheroo?

VHow do you find five different reasons for the same product?

You go looking for five different buying situations, because the reasons are already sitting inside them, like the prize in a cereal box.

For each kind of buyer, ask what just happened in their life, what they’ve already tried, who else is involved and what date is coming. Reviews, sales calls and support tickets are a gold mine for this, and the nuggets usually sit in the first sentence, right where the customer explains why they showed up.

Ask those questions about our invented Spanish service, and five buying situations fall out pretty quickly.

Somebody booked a trip. The flight leaves in eight weeks, so they don’t need the whole language. They need the Spanish for this trip, and a tutor can build the lessons around it.

Somebody’s partner has a Spanish-speaking family. They aren’t chasing fluency either. They want to get through one dinner without their partner translating every sentence, and they’d like to practice that conversation with a person before it counts.

Somebody has a very long streak on a language app. They know a lot of words and still freeze like a deer in headlights when a real person talks back. Their reason is that speaking takes practice with a human. This is the Postum move, because the ad meets them at the moment the old thing stopped being enough.

Somebody started a job with Spanish-speaking clients. They need the words their work uses, and they need them soon. A general course is the scenic route, and they have a client call on Thursday.

Somebody grew up hearing Spanish at home. They understand nearly everything and answer in English, and they’d rather not sit through beginner lessons to fix it. They want speaking practice that starts from what they already know.

Put those into the matrix, and the batch stops looking like six versions of one ad.

Read down the last column, and you’ll see that each opening only makes sense for its own row. Try moving the taco truck line onto the trip row, and it fits like somebody else’s shoes. That’s a decent sign you’ve found different angles, because a hook that fits any row usually belongs to none of them.

The evidence column changes too. The trip ad needs a lesson plan on screen, and the streak ad needs a real student talking, so each angle sends your production team out to capture something different.

VIWhich reasons can your product truthfully support?

A matrix is cheap to fill in, and that’s the danger. A spreadsheet is the friend who agrees with everything, so it will happily hold a row your product can’t back up. So check every row three ways before anybody writes a script.

Is the reason true? Our Spanish service really does build lessons around a trip, so the first row stands. It can’t make anybody fluent in eight weeks, though, so it sells a believable and true claim instead.

Do you have the evidence today? Say the service has no tutors who’ve worked in sales, law or medicine. Then the job row is a check the operation can’t cash yet. Either the team recruits those tutors first, or the row waits. Running the ad and hoping would mean selling something the product doesn’t do.

Are there enough of these buyers to matter? Nobody can guess that from a conference room. Count how often each situation shows up in your reviews, your calls and your tickets. A situation that five customers brought up without being asked is a better bet than one somebody invented in a brainstorm.

Whatever survives all three checks is your next round of briefs.

And when the results come in, keep your claims small. If the in-laws ad sells better than the trip ad, you’ve learned that this version did better in this run. You haven’t yet learned that the angle is why, because the two ads also had different openings, different proof and probably different people watching. Our ad fatigue guide walks through reading a result like that without fooling yourself.

Download the blank angle matrix. It has a row for each angle and a column for each of the three checks, so the weak rows drop out before production starts.

VIIGive the first pass to your AI

Your AI can do a lot of the sorting and digging here, like a tireless intern with a highlighter, as long as it works from your real material and shows you where every idea came from.

The three prompts below run in order. If you start a fresh chat for a later step, paste the earlier output and your source material back in. Take customer names out of anything you paste, because none of this needs personal details.

1. Run the reason-to-buy test on your current ads

Paste the full text of the ads you’re running now, plus the scripts for any videos. If you’d like a second read on a single ad, the Taxonomy’s free copy analyzer names the persuasion moves in whatever copy you paste into it.

Help me find out how many different reasons to buy my current ads actually give.

Product and offer: [fill in]
Ads to review: [paste the full copy or script of each ad, numbered]

For each ad, ignore the opening line or first frame. In one plain sentence, state the reason to buy that the rest of the ad argues, and name the buyer and situation it assumes. If the ad assumes no particular situation, say so.

Then group the ads that give the same reason, even when their openings, formats or tones differ.

Return:
1. A table with one row per ad: the opening, the reason to buy in one sentence, the assumed buyer situation, and the evidence the ad shows.
2. The groups, with a count of how many distinct reasons this batch contains.
3. Any ad whose reason you could not determine, with the lines that made it unclear.

Work only from the copy I pasted. You cannot see my ad account, so do not comment on performance, and do not guess which ads are winning. Do not rewrite anything yet.

Check the groups yourself. If you disagree with one, ask which lines made it put those two ads together, because sometimes it’s caught a sameness you’ve stopped seeing.

2. Build the matrix from what your customers said

Now give it the raw material: reviews, survey answers, sales-call notes, support tickets and your real product facts. More of the customer’s own words means less for the AI to make up.

Using the analysis above, help me build an angle matrix for one product.

Product facts, offer and terms: [paste]
Proof we have today (demos, testimonials we may use, data, credentials): [list]
Customer material (reviews, survey answers, call notes, tickets): [paste, with names removed]
Claims we cannot make: [list]

First, pull out the buying situations that appear in the customer material: what had just happened, what the person had already tried, who else was involved, and any date they were working toward. Quote the line each situation comes from and count how many separate customers mention it. Do not add situations that are not in the material. If you think one is missing, list it separately as a guess for me to check.

Then build a table with up to five rows:
- Buying situation, with the supporting quote and the count.
- Reason to buy for a person in that situation, in one sentence.
- Evidence the ad would need to show.
- Whether that evidence is in my proof list: yes, partly or no.
- One possible opening that only makes sense for this row.

After the table, flag every row where the reason depends on something my product facts do not confirm, and every row that repeats a reason my current ads already give.

Do not invent testimonials, numbers or product capabilities. If the customer material is too thin to support five rows, give me fewer rows and tell me what to collect next.

Read the count column before you fall in love with a row. A great-sounding angle with one customer quote behind it is a guess in a nice suit, and you should brief it like one.

3. Draft one ad, then check that it gives the right reason

Pick a single row you’ve confirmed is true and can prove. This prompt also has your AI check Coppica’s free Persuasion Taxonomy MCP for ways to express the reason, so it helps to connect that first.

Draft one ad for the matrix row below.

Row: [paste the situation, reason, evidence and opening]
Format, length and placement: [fill in]
Approved product facts, offer and proof assets: [paste or reference the attached files]

Before drafting, if Coppica's Persuasion Taxonomy MCP is connected, use it to look for techniques that suit this situation and reason, and tell me which ones you used. If it is not connected, say so and continue with the supplied material without claiming you used it.

Write the opening, the body and the call to action. For video, add the first frame, the on-screen text and a shot list. Keep the whole ad about this one situation and this one reason.

Then audit your draft:
1. Cover the opening. Does the rest of the ad still argue this row's reason, and only this one?
2. Which line or shot carries the evidence, and does that asset exist in my proof list?
3. Would the opening still make sense on a different row? If so, make it more specific to this situation.
4. Which claims depend on a fact I did not supply?

Do not invent testimonials, demonstrations, deadlines, guarantees or results. Mark any missing asset as needed. Do not publish or launch anything.

The facts, the choice of angle and the decision to spend are still yours. The AI gets you to a first draft faster, but you’re still the one driving, because it can’t tell you whether your product really does what the row says.