Control Test · EUR 2,000 one-time
10 variants against your best ad. 14-day test. One number.
For brands already running paid social: AI Vidia builds 10 variants to beat the ad you run best, they run in your own account, and one metric agreed up front decides the result.
- Price
- €2,000
- one-time, no retainer
- Build
- Days 1 to 3
- 10 variants in 9:16
- Test
- 14 days
- in your own ad account
- Readout
- Day 18
- one metric, yours either way
Already spending EUR 50,000+ a month? Start on the Performance Retainer.
What EUR 2,000 buys
- You name your current best-performing ad (the control)
- We build 10 variants to beat it
- Your team or ours runs them in your own ad account for 14 days
- One agreed metric up front: CTR or CPA against control
- Readout call on the numbers, yours either way
- No retainer, no lock-in, procurement never gets called
If the ten variants beat your control, the EUR 2,000 comes off month one of a Performance Retainer or Brand System.
Who it is for
DTC and consumer brands with a live paid-social account and a best-performing ad worth beating. A marketing lead can approve it without procurement.
How it runs
The Control Test: eighteen days, one number.
- 01
Day 0. You name the control
Send the brief with your brand URL. You pick the ad you run best today and the one metric we measure: CTR or CPA against that control.
- 02
Days 1 to 3. We build ten variants
Ten AI ad variants built to beat the control, delivered in 9:16 for Meta, TikTok and YouTube, ready to upload.
- 03
Days 4 to 17. The test runs
The variants run in your own ad account against the control for 14 days, launched by your team or by us, as agreed. Your data, your account, your spend.
- 04
Day 18. The readout
A call on the numbers, yours either way. If the ten beat the control, the EUR 2,000 comes off month one of a Performance Retainer or Brand System.
Proof
Tested weekly. 2.4x ROAS on winning cohorts.
IndianBites, an Indian restaurant and takeaway in Copenhagen. Read the case study →
Control Test FAQ
Control Test questions before the invoice.
What is a Control Test?
A EUR 2,000 one-time engagement: AI Vidia builds 10 AI ad variants against your current best-performing ad, your team or ours runs them in your own ad account for 14 days, and one metric agreed up front decides the result.
Which ad is the control?
The ad you run best today, by the metric you care about. You name it in the brief; we do not pick it for you. If you run several, choose the one with the most spend behind it.
Which metric is measured?
One, agreed before we build: CTR or CPA against the control. The test runs 14 days in your account, so the readout uses your own numbers, not a screenshot from ours.
What happens if the variants lose?
You keep the ten variants and the readout, and you owe nothing further. No retainer, no lock-in.
What happens if they win?
The EUR 2,000 comes off month one of a Performance Retainer (EUR 8,500 a month, 40 videos) or Brand System (EUR 18,000 a month, 90 videos). The credit does not apply to the Pilot Sprint.
Is the Control Test the same as the concept by application?
No. The concept is one speculative idea in 48 hours, built only for brands that qualify, no spend. The Control Test is a paid, measured test: ten variants, 14 days, one number.
Our best ad is wearing out. How do we turn what worked into genuinely new concepts?
Keep the buyer question your winner answered and tell it through another story, or test another documented buying reason. Note what the ad shows and the objection it answers; treat why it worked as a hypothesis. A new opening on the same script is the same concept.
If the offer or landing page changed, fix that first. To measure new work against that ad, a Control Test uses it as the control: we build 10 variants to beat it, and your team or ours runs them in your own ad account for 14 days.
Can an AI ad partner lower our CPA?
AI Vidia does not promise a lower CPA up front; the Control Test measures it in your own ad account.
For EUR 2,000 we build 10 variants against your best ad, your team or ours runs them for 14 days, and CPA can be the one metric agreed before the start. If the variants win, the EUR 2,000 comes off month one of a Performance Retainer or Brand System.
One named client result, in Singh Law's own words: "We lowered our CPL by 19%." That is their number, not a forecast for your account.
Next step
Book a Control Test. Name the control.
Ken replies the same working day and confirms the control and the metric before any invoice.