Ad Testing Framework · Volley.ai · June 2026
3 Ads. 3 Audiences. What to test, measure, and kill.
Sequence · metrics · kill rules — set in advance, not post-hoc.
WINDOW 6 weeks
BUDGET £500/angle to start
PLATFORM Meta Feed + Stories
/ 01Busy Professionals
Volley
Volley
Sponsored
···
You have 15 minutes. That's enough — if someone's actually paying attention.
Swedish · AI + Human Tutor
15minutes of Swedish · every day
A private tutor.
Fits your diary.
Not the other way.
Your Swedish tutor · onlinePlans around your schedule
getvolley.ai
Start speaking Swedish in weeks
$99/month. Cancel anytime.
Ad 01 · Test Plan
Busy Professionals — Time scarcity hook
Week 1–2
Test the hook
"15 minutes" vs benefit-led headline. Same ad otherwise.
Week 3–4
Test the button
"Free placement test" vs "Start for free."
Week 5–6
Test the audience
LinkedIn pros vs Duolingo users. Pick the winner.
Why it should work
Busy people care about time first, results second.
Saying "15 minutes" upfront removes the main objection before they even read the offer.
What could go wrong
Lots of clicks but few sign-ups means the price is the issue.
If that happens, fix the landing page — not the ad.
What we measure
MetricTargetKill if…
Click-through rate≥ 2.8%below 1.4%
Cost per sign-up≤ £28above £44
Sign-up → paid≥ 20%below 12%
When to stop
→
Kill the hook if clicks stay below 1.4% after a week.
→
Kill the audience if fewer than 12% of sign-ups become paid customers.
→
Don't judge on cost alone — a pricier sign-up that pays is better than a cheap one that doesn't.
/ 02App Dropout Survivors
Volley
Volley
Sponsored
···
You didn't quit. The app let you go. This time there's a real Swedish tutor who notices when you skip a day.
Swedish · This time is different.
You didn't quit.
The app let you go.
Duolingo ✕Babbel ✕Pimsleur ✕Volley ✓
Your Swedish tutor
Tuesday · 8:43am
"Missed you yesterday — no judgment. I've set up a shorter session today so coming back feels easy."
getvolley.ai
A tutor who notices when you stop
$99/month. Cancel anytime.
Ad 02 · Test Plan
App Dropout Survivors — Recognition + accountability hook
Week 1–2
Test the message
"The app let you go" vs "Downloaded Duolingo 6 times?" Pick the one that lands.
Week 3–4
Test the tone
Warm "no judgment" vs direct accountability.
Week 5–6
Cold vs retarget
Run both. Put 80% of budget on the winner.
Why it should work
Nobody else is telling app dropouts it wasn't their fault.
The crossed-out logos get recognition. The headline removes shame so they're willing to try again.
What could go wrong
This audience has been burned before — clicks don't equal trust.
If they click but don't sign up, the landing page needs to prove Volley is different.
What we measure
MetricTargetKill if…
Click-through rate≥ 3.2%below 1.8%
Cost per sign-up≤ £50above £50 (cold)
Sign-up → paid≥ 18%below 10%
When to stop
→
Kill the guilt version if it gets fewer clicks than the validating one.
→
Kill cold targeting if cost per sign-up goes above £50 after two weeks.
→
Keep retargeting even if cold fails — they perform differently.
/ 03Serious Learners
Volley
Volley
Sponsored
···
You know the grammar. Your Swedish tutor knows exactly what's stopping you from using it.
Your Swedish tutor · online now
For serious learners
Stop collecting grammar rules.
Start having conversations.
›
Marcus T.
Learning Swedish · 5 months
"My tutor noticed I kept avoiding past tense. Two focused weeks unlocked more than 6 months of apps combined."
4×
Live tutor sessions per month
∞
Daily AI conversation practice
‹
getvolley.ai
Break through the plateau
$99/month. Cancel anytime.
Ad 03 · Test Plan
Serious Learners — Identity challenge + carousel proof
Week 1–2
Test the headline
"Stop collecting grammar rules" vs "You're not a beginner anymore."
Week 3–4
Test the swipe
If under 35% swipe to card 2, fix card 1 — not card 2.
Week 5–6
Carousel vs single
Try card 1 alone. See if simpler wins.
Why it should work
A face builds trust. Proof closes the sale.
Card 1 is emotional. Card 2 is factual. Serious buyers want both.
What to watch
This ad should have the highest paid conversion.
If it doesn't beat Ad 02, we're targeting the wrong people.
What we measure
MetricTargetKill if…
Card 1 → 2 swipe rate≥ 35%below 25%
Click-through rate≥ 2.5%below 1.5%
Sign-up → paid≥ 25%below Ad 02
When to stop
→
Kill the carousel if fewer than 25% swipe to card 2 — test card 1 as a static ad.
→
Kill the headline if clicks stay under 1.5% — it may feel aggressive instead of motivating.
→
Kill the angle if paid conversion stays below Ad 02 after three weeks.
Ad Testing Framework · Volley.ai · v2.0 · June 2026
"Pre-register the kill threshold. Don't negotiate with the data."