← Back to all lessons Lesson 02 of 09

Create & test · Lesson 02

Find an angle worth testing.

Most bad UGC starts with “we need a creator.” Start with what the customer is already trying to solve.

Find the tension first. Then give the creator something worth saying.

~5 min · leave with 3–5 UGC angles + a creator brief

TL;DR

Don't start with the creator. Start with the customer tension.

A good UGC angle usually comes from something your customer already thinks, wants, hates, worries about, or keeps failing to solve.

01

Find the tension

What is the customer already struggling with?

02

Turn it into an angle

Make the problem specific enough to build a video around.

03

Check if it's testable

Can different creators/hooks express the same idea?

04

Turn it into a brief

Give the creator the truth — not a script.

What to check in Airbridge

  • Trial starts by angle: did the idea attract users with real intent, no matter who delivered it?
  • Paid subscribers by angle: which idea actually converted? (Revenue and cost per install appear once you connect an ad account or upload what you paid your creators.)
  • Creators inside a winning angle: did one creator carry the idea, or does it work for everyone?

You'll compare angles, not just videos. Lesson 04 shows where the angle name goes.

01 Find the tension

The strongest UGC usually starts with something the customer already feels.

A benefit describes the product. A tension describes the person.

A product benefit tells people what the app does. A tension describes the situation that makes someone want the app in the first place.

Instead of: “Stay productive.”

Look for something more like: “I want to stay consistent, but every productivity app becomes another task.”

That gives you much more to work with. There's a person, a problem, and a contradiction you can actually dramatize.

Look where customers already talk naturally

You do not need a giant research project. Start with places where people describe the problem in their own words:

  • App Store / Play Store reviews
  • Competitor reviews
  • Reddit
  • TikTok / Instagram comments
  • Support conversations
  • Cancellation reasons
  • User interviews
  • Sales or onboarding calls
  • Search queries
  • Comments on your existing ads/content

What are you looking for?

Pay special attention to sentences containing things like:

“I wish…” “I hate that…” “I keep…” “I can never…” “but…” “even though…” “the problem is…”

Those often contain the tension almost verbatim.

Example

Imagine a workout app.

Generic benefit: “Personalized workouts.”

Possible customer tension: “I finally get myself to the gym, then spend 20 minutes figuring out what to do.”

That second sentence is already halfway to a UGC concept.

If the angle could describe almost any competing app, it probably isn't specific enough yet.

02 Turn the tension into an angle

Give the problem a moment, a promise, and proof.

A tension becomes testable once it has all four parts.

Use this simple structure:

Pain

What is frustrating them?

Moment

When does that frustration actually happen?

Promise

What changes with your product?

Proof

What can the creator show instead of merely claiming?

Example

Pain

“I never know what exercises to do.”

Moment

Standing in the gym scrolling through saved workouts.

Promise

Tell the app your goal and available equipment; it builds the workout.

Proof

Show the app generating today's workout, then show the actual exercises.

That gives you an angle:

“For people who finally make it to the gym… then have no idea what to do.”

Now you have something several creators can interpret in their own voice.

Build 3–5 angles before you make 10 videos.

Do not produce ten random UGC videos and hope one works. Start with a small set of meaningfully different ideas.

Angle A — confusion

“I get to the gym and still don't know what to do.”

Angle B — inconsistency

“I can follow a workout plan for three days. Then life happens.”

Angle C — intimidation

“I hate looking like I have no idea what I'm doing.”

Angle D — decision fatigue

“I don't need more exercises. I need someone to tell me which ones to do today.”

Those are actual hypotheses. “Creator A / Creator B / Creator C” is not.

Quick angle builder

Structure the idea. This isn't an AI writer — just a way to shape what you already know.

What does the customer say?

When does this happen?

What can the creator show?

Your angle

For people who want to something, but something gets in the way.

03 Check if it's testable

A good angle survives more than one creator.

Before you commit budget, ask three questions.

1. Is this one clear idea?

Good: “I get to the gym and don't know what to do.”

Too broad: “Fitness is hard and our app makes everything easier.”

2. Could three creators express it differently?

You want the idea to stay consistent while the delivery changes. One creator might act out the frustration. Another might tell a personal story. Another might screen-record the product.

That lets you learn whether the angle works — not just whether one creator is charismatic.

3. Can you tell if it worked?

If you run the angle, you should eventually be able to answer:

Did this idea bring more clicks, installs, trials, subscribers, or revenue?

If you cannot identify the angle later, you cannot learn from the test.

Name the angle before you publish it.

Keep the naming simple. For example:

angle-gym-confusion angle-no-time-to-plan angle-beginner-intimidation

Keep the creator and the video as separate labels instead of one long name:

angle-gym-confusion creator-maya hook-b

Later, each label becomes its own column. That lets you compare angles across creators, or creators inside one angle. You do not need a perfect taxonomy. You just need future-you to understand what was being tested.

Give every angle a name before the first creative goes live.

04 Turn the angle into a creator-ready brief

Give creators the truth to work with — not the lines to recite.

Protect the idea without making every creator sound identical.

Once the angle is clear, the brief should protect the idea without making every creator sound identical.

Do not send: “Say: I used to struggle with working out but then I discovered…”

That is how UGC starts sounding like an ad. Instead, give the creator five things.

Audience

Who is this for?

Problem

What is the tension?

Product truth

What can the app genuinely do?

Proof

What should they actually show?

CTA

What should the viewer do?

That is enough structure to keep the message consistent without writing the creator's personality out of the content.

Build your brief

Turn your angle into a creator brief. Fill in what you know — the brief updates as you go.

Your creator brief

Audience—
Angle—
Product truth—
Proof to show—
CTA—

Use your own words. Do not read this like a script.

Find an angle. Then find out if it actually grew the app.

Airbridge connects each creative to what happened after the click — install, trial, paid subscriber, and revenue.

Angle→ Creative→ Install→ Trial→ Paid→ Revenue

See which ideas deserve another creative — and which ones should disappear.

Start measuring in Airbridge →