TikTok UA — Meeting Prep

Your Briefing for Houssem + TikTok

Everything pulled from Omni + what it means + what to ask. April 2026.

Open with this

"I pulled our TikTok data in Omni before this call. I have questions rather than answers — can you help me understand if I'm reading it correctly?"

01 What We've Spent and What Came Back

12-Month TikTok Spend vs Revenue — Fabulous + Clarify
App Spend Revenue d30 Revenue d90 Revenue d180 ROAS d180
Fabulous €784,940 €398,525 €524,082 €570,292 0.73
Clarify €23,529 €10,928 €15,960 €19,237 0.82
Shape €0 — — — —
Source: Omni / ua_performance topic / TikTok network filter / last 12 months
↗ View in Omni — Spend & Revenue by App
What this means in plain language: We spent €784K on Fabulous TikTok ads. After 6 months, those users had paid back €570K. We're still €214K short of breakeven at d180. The model only works if users keep subscribing past 6 months. Clarify is a tiny test but its users pay back faster (0.82 vs 0.73 at d180).
Key question for Houssem: What's our internal ROAS target and expected payback window? At what point (d365?) does Fabulous become profitable per cohort?

02 Where People Drop Off (The Funnel)

Impressions (12mo)
272M
People who saw the ad
Onboarding Starts
1.59M
€0.49 each
= €784,940 ÷ 1,593,415
Payment Page Shown
40,715
€19.27 each
= €784,940 ÷ 40,715
Onboarding → Payment
2.55%
Only 1 in 40 people who start the app ever see the price
The big drop: 1.59M people start the app. Only 40K ever see the paywall. That's the leak. Everything before the paywall is awareness. Everything after is where the money is made. Onboarding starts are useful for diagnosing problems — but they are not the goal.

03 Creative Performance — What's Working and What's Dying

Top Creatives — 90-Day Window (best historical performance)
CreativeSpendCPMCPA D7Status
Nick Hypercut (Jan 26)€23,599€2.57€28.90Fatiguing
Nick Hypercut variant€21,641€2.46€30.80Fatiguing
TweetStart (Jan 26)€17,680€2.73€27.00Fatiguing
Nick UGC-Note (Dec 25)€16,381€5.30€30.20Fatiguing
PovV2 (Aug 25)€14,492€2.61€32.80Declining
Same Creatives — Last 30 Days (fatigue showing)
CreativeSpendCPMCPA D7vs 90-day
Nick Hypercut (Jan 26)€3,307€2.23€38.50+€9.60
Nick Hypercut variant€2,786€2.43€48.00+€17.20
TweetStart variant€830€1.92€33.20+€6.20
Source: Omni AI pull — TikTok + TikTok AdQuantum / Fabulous / ad-level data
↗ View in Omni — Creative Performance
Creative fatigue explained: The algorithm has already shown these ads to everyone who was likely to convert. Now it's showing them to harder-to-reach people, so each purchase costs more. The fix is new creative. 568 creatives were tested in the last 90 days — this was a high-volume machine that has lost its senior operator.
Questions for Houssem: Can you show me the actual Nick Hypercut and TweetStart videos? What made them work? Who produces creatives and how do I brief them? Is new creative already in production?

04 Your UA Metrics Toolkit

CPA D7 Cost Per Acquisition at 7 Days
How much you spent to get one paying customer within their first week.
Your fastest signal that an ad is working. If CPA D7 is €28, you need each user to pay more than €28 over their lifetime to be profitable. It doesn't tell you if you're profitable — it tells you if the ad is doing its job.
Our numbers: €27–31 at best (90-day window) → degraded to €33–48 in last 30 days
ROAS Return On Ad Spend
Revenue earned ÷ money spent on ads. A ROAS of 1.0 = breakeven. Above 1 = profit. Below 1 = loss (so far).
In subscription apps, ROAS below 1 at d30 or d90 is normal — users pay back slowly over months. D7 CPA tells you if an ad works short-term. D90/D180 ROAS tells you if the business model works long-term. You need both.
Fabulous: 0.51 (d30) → 0.67 (d90) → 0.73 (d180). Clarify: 0.46 → 0.68 → 0.82. Neither has broken even at d180 yet.
CPM Cost Per 1,000 Impressions
How much it costs to show your ad to 1,000 people.
Low CPM = cheap reach. But CPM alone means nothing. A €10 CPM with €20 CPA is better than a €2 CPM with €50 CPA. Here's why: imagine ad A costs €10 to reach 1,000 people and 1 in 2 buys (CPA €20). Ad B costs €2 to reach 1,000 people but only 1 in 25 buys (CPA €50). Ad A is more expensive per view but far cheaper per customer. Always look at CPM in relation to CPA.
Our best creatives: €2–2.60 CPM. Industry context: €1–3 = good. €5–10 = expensive. €10+ = problem.
€1–3
Good
€3–5
Watch it
€5–10
Expensive
€10+
Problem
LTV Lifetime Value
Total revenue a user pays over their entire time as a subscriber.
The number that makes everything else make sense. If average LTV is €80, then a CPA of €40 is a great deal. If average LTV is €30, then even a CPA of €28 barely works. You cannot evaluate any campaign without knowing LTV. We haven't pulled this yet.
Unknown — ask Houssem: "What's our average LTV per Fabulous user and what CPA does that justify?"
COHORT A Group of Users Acquired at the Same Time
Everyone who came in during a specific period (week, month, campaign).
When Houssem says "the January cohort is at d90" he means: all users acquired in January, measured 90 days later. Cohorts let you compare campaigns fairly regardless of when they ran. You'll likely review weekly cohorts for active campaigns and monthly cohorts for ROAS/LTV analysis. Ask Houssem what the review cadence is for TikTok specifically.
CREATIVE FATIGUE
When an ad stops performing because the audience has already seen it too many times.
The algorithm has shown your ad to everyone likely to convert. Now it's reaching harder people. CPA rises. The only fix is new creative. This is happening to us right now — CPA moved from €28 → €48 on the same videos.

05 How to Review Creatives (When That Becomes Your Job)

1
Watch what worked and ask why

Before briefing anything new, watch Nick Hypercut and TweetStart. Was it the hook? The person? The editing pace? The emotion? You're looking for the variable that made people stop scrolling.

2
Change one thing at a time

Never test two variables at once. If you change the hook AND the person AND the style, you won't know what caused the result. Test: same video, different first 2 seconds. Same concept, different person. Same person, different edit.

3
Use this decision rule for every creative

Does it stop you scrolling in the first 2 seconds? Is the problem it solves immediately obvious? Is the call to action clear at the end? All three = test it. Any fail = send back with specific feedback on which one.

4
Let data kill things, not your opinion

You don't need to know if a creative is good before launching. You need to know after 50–100 purchases whether CPA is acceptable. Your opinion is a hypothesis. The data is the verdict. This is what makes UA more objective than most marketing.

06 All Questions for Houssem

Use this as your checklist. Tick them off as you go.

On the numbers
We spent €784K on TikTok for Fabulous and got €570K back after 6 months — is that good, bad, or normal? What number are we actually aiming for?
You need to know if you're inheriting a healthy machine or a struggling one. Without the target, you can't tell.
I found older data showing how much users paid after a full year — can we look at it together so I understand if TikTok is ever actually profitable?
The d365 data exists in Omni for Sep-Nov 2024 cohorts but I couldn't read it cleanly on my own. View in Omni
How much does an average Fabulous user pay us over their lifetime? And what does that mean for how much we can afford to spend to get one?
This is LTV. You can't judge whether a 33 or 48 euro CPA is acceptable until you know this number.
How come clicks and installs show zero in our data for TikTok — is that normal or is something broken?
Probably because all spend goes to web, not app installs. But you want to confirm nothing is actually broken in the tracking.
On the ads
Can you show me the actual videos that worked best? I want to watch them and understand why they performed.
Nick Hypercut and TweetStart are the top performers. Seeing them beats any amount of reading about them.
The ads have names like Fab_6774_Nick_Hypercut_01_26 — can you explain what that means so I can follow along?
You'll be reading these names in data every day. You need to decode them fluently.
The same ads that used to cost us 27 euros to get a customer are now costing 48 — should I be worried about that?
This looks like creative fatigue. You want Houssem to confirm it and tell you what the acceptable ceiling is before something gets paused.
Who actually makes the videos? And when I need a new one, what do I do — who do I talk to, what do I say?
568 creatives in 90 days is a fast operation. You need to know the process before you're expected to keep it running.
Do we have new videos coming, or is that something I need to sort out urgently?
If no new creative is in production and CPA is already rising, this is your week-one problem to flag.
On what I am actually supposed to do
Are we running ads for Fabulous only right now, or Clarify and Shape too?
The data shows basically zero TikTok spend on Clarify and Shape. You need to know if that is changing.
We barely tested Clarify on TikTok — is there a plan to do more with it?
Clarify had a better ROAS than Fabulous at d180 on a tiny budget. Worth understanding why it was never scaled.
How often will I check the numbers — every week? Every month?
You need a routine before you can build one. Ask Houssem what the cadence looks like in practice.
What can I decide on my own vs what do I need to check with you first?
Getting this boundary clear early saves a lot of awkward moments later.
What does a good first week look like from your side?
Lets Houssem set the expectation rather than you guessing at it.

07 Questions for the TikTok Account Manager

Don't volunteer that you're new. Let them lead the intro. Your job in this meeting is to listen and ask one or two sharp questions. They want you to spend more money — use that.

Looking at our recent campaigns, where do you see the biggest opportunity to improve CPA?
Why: they have visibility into how similar accounts perform. This gets them talking and reveals what they know about your account.
Our top creatives are showing fatigue — what refresh cadence do you recommend for subscription wellness apps?
Why: they see hundreds of accounts. They'll have a real answer to this. Also signals you're tracking performance closely.
What ad formats are performing best for subscription apps right now?
Why: TikTok rolls out new formats constantly. The rep will know what's getting algorithmic boost right now that we haven't tested.
Have you seen Spark Ads outperform standard in-feed for apps like ours?
Why: Spark Ads = boosting real creator content as paid ads. It's a natural bridge with your partnerships background. Worth knowing if it's worth pursuing.
What support can you offer us directly — creative reviews, performance audits, beta access to new formats?
Why: TikTok reps often have resources most advertisers don't ask for. Ask. The worst they say is no.