> ## Documentation Index
> Fetch the complete documentation index at: https://docs.snowdoughnut.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Angle brief
> A fill-in workbook for one angle: the proposition, the person it speaks to, the belief it works against, the proof behind it, and the hooks it can produce. Fill one in before an angle goes into production.
*This page is a workbook. Fill in each section, and expand a concept review if you want the reasoning behind a section before you answer. You can complete the whole brief without opening one.*
*Grayed* `E.g.` *text is an example to delete. Every example uses one invented company, **Doughnut Labs**, a SaaS company that sells disruptive Doughnut Technology, working on a single angle so you can see one brief taking shape.*
One filled copy of this brief is one angle. An angle is not one ad and not one campaign: it is an argument that several ads, on several channels, will each execute differently.
Once the sections here are familiar, the [angle template](/angle-template) is the same structure with the explanations stripped out.
One sentence someone could agree or disagree with. Everything else is checked against it.
Who it speaks to, and what they currently think that the angle has to work against.
The proof that makes it credible, and what it must never be stretched into claiming.
Three or more ways into the same argument, then a row in the library.
## Setup
| Field | Fill in |
| :--------------------------------------------------------------- | :---------------------------------------------------- |
| [Owner](/glossary#owner) | *E.g. Aisha Bello* |
| Angle name | *E.g. Monday morning back* |
| Short form | *E.g.* `MONDAYBACK` |
| [Angle type](/ad-name-conventions#message-angle) | *E.g. Outcome,* `OUTC` |
| Brief name ([file naming conventions](/file-naming-conventions)) | *E.g.* `DOC_angle-monday-morning-back_2026-08-13_v01` |
| Date created | *E.g. 13 Aug 2026* |
| Status | *E.g. Testing* |
| \[add setup field] | |
The angle name and the short form do different jobs. The name is how the team refers to the argument in a conversation, so it wants to be memorable rather than descriptive. The short form is what travels into ad names and reports, where a report joins on exact strings and a name typed slightly differently the second time becomes a second angle nobody meant to create.
The type row connects this brief to the [dictionary](/ad-name-conventions#message-angle) that already exists in the ad naming conventions. That dictionary owns the closed list of argument kinds, because a report needs a small set to group by. This brief owns the argument itself. Keeping the two separate is what lets a library hold forty angles while a report still groups them into six columns.
## 1. The proposition
**In one sentence, what is the argument? Write it so someone could agree or disagree with it:**
*E.g. The four hours you spend rebuilding the Monday report could be spent reading it.*
| Field | Fill in |
| :-------------------------------- | :---------------------------------------------------------------------------- |
| What it asks the reader to accept | *E.g. That the rebuilding, not the reading, is the part eating their morning* |
| Why it is worth arguing | *E.g. Nobody defends manual report building. They just have not costed it* |
| Funnel stage it suits | *E.g. Top and middle. Works before they know a tool exists* |
If the sentence could sit on any competitor's site unchanged, it is a category statement rather than an angle. "Save time on reporting" is something every tool in the market claims. An angle is specific enough that someone could read it and decide it is not about them.
The disagree test is doing more work than it looks like. A sentence nobody could disagree with is not persuading anyone, because it is not asking anything of the reader. "Reporting is important" passes unnoticed. "The rebuilding, not the reading, is what is eating your Monday" can be rejected, which means it can also land.
Writing it as a full sentence is what makes the rest of the brief checkable. A hook either delivers that sentence or it does not. A proof point either supports it or it supports something else. Without the sentence, every later decision gets argued from taste, and the angle quietly becomes whatever the last person to touch it thought it meant.
Angles also differ in how much they carry. Some are narrow and only work at one funnel stage; others sit underneath a whole quarter of creative. The stage line is where that gets recorded, so an angle built for cold traffic does not get quietly reused on a retargeting audience that has heard it four times.
## 2. Customer profile and the belief
**Link the [customer profile](/customer-profile-library-overview) this angle speaks to, and name what they currently believe that it has to work against:**
| Customer profile | Link | The belief this angle works against |
| :---------------------------- | :----------------------------------------------------------- | :-------------------------------------------------------------------------------------------------------------------------- |
| *E.g. Mid-market RevOps lead* | *E.g. [CP\_midmarket-revops\_2026-03-02\_v02](https://link)* | *E.g. That the manual rebuild is the job, not overhead. They have done it so long it reads as diligence rather than waste.* |
| \[add profile] | | |
**Who this angle is not for:** *E.g. Teams with no recurring report. The argument has nothing to bite on.*
An angle is the sentence where a product truth meets a customer belief. Take either one away and it stops being an angle. The product truth on its own is a feature: accurate, and no reason to care. The belief on its own is an insight: it describes a person and asks nothing of them.
That is why this section comes before the proof. Proof selected with no belief in mind tends to be whatever number was easiest to find, and it ends up answering a question the reader was not asking.
The "not for" line earns its place because angles spread. An argument that works on one profile gets reused on a second because it is already made and the creative already exists, and the result underperforms in a way that looks like creative fatigue. Recording who it was never meant for makes that reuse a visible decision instead of a drift.
## 3. Proof
**What makes the proposition credible rather than asserted?**
| Proof | Type | Source | Cleared for use |
| :-------------------------------------------------------------- | :------------------ | :--------------------------------- | :---------------- |
| *E.g. Median 4.2 hours a week spent on recurring reports* | *E.g. Research* | *E.g. 2026 customer survey, n=310* | *E.g. Yes, 2 Aug* |
| *E.g. Setup completed in one afternoon by 8 of 10 new accounts* | *E.g. Product data* | *E.g. Onboarding telemetry, Q2* | *E.g. Yes, 2 Aug* |
| \[add proof] | | | |
Proof is what separates an angle from a slogan. The proposition makes a claim; this section is the reason someone should accept it. A number, a mechanism, a demonstration, or a customer's own words all do more than an adjective.
The cleared column exists because proof ages differently from the argument it supports. A statistic from a survey two years old still reads as current in an ad, and the angle it supports can outlive the evidence underneath it by a year without anyone noticing. Recording when a proof point was last checked is what surfaces it at review, before a customer or a regulator surfaces it for you.
Angles vary in how much proof they need. An outcome angle claiming a specific result usually needs a number behind it. An identity angle, which works by recognition rather than by evidence, may need very little, and forcing a statistic into it can make it read as a sales pitch instead. What matters is that the proof matches the kind of claim being made.
## 4. Hooks this angle can produce
**Draft at least three ways into the same argument. These are starting points for writers, not final copy:**
| # | Hook | Format it suits | Used in |
| :---------- | :----------------------------------------------------- | :---------------------------- | :------------------------------------------------- |
| 1 | *E.g. "It's 9am Monday. Your report is already done."* | *E.g. Static, video open* | *E.g.* `DoughnutDesk-STATIC-OUTC-FREETRIAL-TRY-LP` |
| 2 | *E.g. "You didn't get hired to copy and paste."* | *E.g. Video, first 2 seconds* | |
| 3 | *E.g. "Four hours a week. Every week. Forever."* | *E.g. Static, text-led* | |
| \[add hook] | | | |
If you cannot write three hooks that differ by more than wording, the angle may be narrower than it looked. Better to know that before production starts.
A hook is one execution of the angle, and the angle is what survives when a hook stops working. Keeping several in the brief is what makes that survival practical: when a creative tires, the next one comes from this table rather than from a fresh strategy session.
The three-hook exercise doubles as a test of the angle. An argument broad enough to be worth production can usually be entered from several directions: the cost of the status quo, the moment of relief, the absurdity of the current process. An argument that only supports one phrasing is often a headline that has been mistaken for an angle.
What counts as a good hook is genuinely channel-dependent, which is why the format column is here rather than a rule. What earns attention in a cold scroll and what earns it in a search result are different problems, and the same angle has to enter differently in each.
## 5. Where it runs
| Channel | Formats | Copy template | Status |
| :------------------- | :--------------------------- | :----------------------------------------------------------- | :------------- |
| *E.g. Meta* | *E.g. Static, video* | *E.g.* [Meta static/video](/meta-static-video-copy-template) | *E.g. Live* |
| *E.g. Google Search* | *E.g. Responsive search ads* | *E.g.* [Search copy template](/google-search-copy-template) | *E.g. Planned* |
| \[add channel] | | | |
An angle is channel-agnostic in principle and rarely in practice. The same argument has to be made in three seconds of video, in a thirty-character headline, and in a subject line, and it does not always survive all three. An angle that depends on a slow build will not work in search; one that depends on a specific number may be the only kind that does.
Recording where an angle actually runs also turns the library into something a diversity check can be run against. An account can hold twelve angles and still be running one argument, if eleven of them only ever went live on one channel.
## 6. Guardrails
**What must this angle never be stretched into claiming?**
* *E.g. That reporting becomes fully hands-off. It does not, and support pays for the gap.*
* *E.g. A specific hours-saved number for an individual account. The 4.2 figure is a median.*
* \[add guardrail]
**Approvals this angle needs:** *E.g. [Legal review](/legal-review-process) on any time-saving claim carrying a number. Cleared 2 Aug.*
Angles drift under production pressure. Each new execution reaches slightly further than the last, because the previous one worked and a stronger version of it should work harder, and no single step looks like an overclaim. The gap between the approved argument and the running one opens gradually.
Writing the boundary at the angle level, rather than catching it per ad, is what makes that drift checkable. A reviewer looking at one ad sees a claim that seems defensible. A reviewer holding this list sees whether it is the claim that was approved.
This is also where an angle interacts with [legal review](/legal-review-process). An approval given on a specific claim does not extend to the argument in general, and where an approval carries a condition, the condition has to survive into formats that assemble themselves.
## 7. Measurement and review
| Field | Fill in |
| :-------------------------------- | :------------------------------------------------------------------------------------- |
| What would prove this angle works | *E.g. CTR above the account median and cost per trial within 15% of the control angle* |
| Where it is reported | *E.g. Paid media dashboard, grouped by* `OUTC` |
| Read date | *E.g. 15 Sep 2026* |
| Decision at read | *E.g. Scale, hold, or retire. Recorded in the library either way* |
An angle and a creative fail in different ways, and the response to each is different. A creative that underperforms while its angle performs elsewhere is an execution problem, answered with a new hook. An angle underperforming across every execution is an argument problem, and producing more creative against it spends money on a question already answered.
Separating the two requires that performance can be grouped by angle at all, which is what the type value in the ad name is for. Without it, every angle result is a creative result, and the account keeps relearning the same thing.
The read date matters because angles do not announce their own decline. Nothing in a platform ever suggests that an argument has stopped earning its place, and a scheduled read is the only thing that reliably catches an angle still running on a proposition the audience exhausted two months ago.
## 8. Status
| Item | Fill in |
| :------------------------------------------- | :-------------------------------- |
| Added to the [angle library](/angle-library) | *E.g. 13 Aug 2026* |
| First creative produced | *E.g. 20 Aug 2026* |
| First live date | *E.g. 25 Aug 2026* |
| Last refreshed | *E.g. New hook added 12 Sep 2026* |
| Outcome | (blank until the read date) |
## Build order recap
Not a category statement any competitor could also make.
Including who it is deliberately not for.
With a date on when each point was last cleared.
So the next creative does not need a new strategy session.
So the coverage check can see it.
## Related resources
* [**Angle concepts**](/angle-concepts) Why angles, hooks, and angle types are three different things.
* [**Angle template**](/angle-template) The same structure without the concept reviews.
* [**Angle library**](/angle-library) Where this brief gets filed, and the coverage check it feeds.
* [**Creative brief**](/creative-brief) and [**copy brief**](/copy-brief) The two documents this angle hands off to.
* [**Customer profile library**](/customer-profile-library-overview) Where the profile behind the angle comes from.
* [**Ad naming conventions**](/ad-name-conventions#message-angle) The type dictionary a report groups by.