Not your usual AI
Not your
usual AI
How the platform works
Advanced budget allocation
Budget management works at two levels and depends on how your accounts and goals are set up.
Shared budgets let multiple projects draw from the same pool — with minimum spends guaranteed and the remainder continuously allocated toward whatever is performing best. Requires a minimum budget of €10,000 and can hold up to 10 projects with 3 platforms in each.
Our cross-channel budget allocator decides how to split budget across platforms (but within a project). Connect your ad platforms, choose on one goal, and switch it on. Amanda AI optimizes the split from there.
Reports from the AI
Every optimization Amanda AI makes happens inside the engine. Reporting is the transparent layer — so you always know what’s running, what changed, and predictions of what’s to come.
Every week, Amanda AI generates an AI report — written by Claude, based on your latest data. Not a dashboard export. A document with chapters covering what changed, what was added or removed, how the competition moved, and what the numbers tell you about your account. Read it like a book or feed it to your AI.
Conversion lag gives you a more accurate picture of your conversions before they have been reported. Amanda AI tracks your account’s specific conversion patterns and predicts how many are still on their way. Budget decisions are then based on what actually happened — not just what’s been reported so far.
Performance reports gives you a combined view of your key metrics across campaigns and platforms with filter options — all in one place. And manual hours saved, because that’s a metric worth tracking too.
Audience optimization
The ad platforms’ AI are only as good as the signals they receive. We connect your first-party data and aligns optimization events with your actual goals — ensuring cleaner and better signals to all supported platforms. Audiences are continuously refined based on real performance data without manual segmentation.
AI assets
Great campaigns need a constant supply of fresh assets. Our platform handles that without external tools.
Generate images with Nano Banana directly inside the Amanda AI platform or let us import copy and images from your website straight into your asset library. You can also upload your own assets and put them to work.
On Meta, content groups let you organize assets for structured testing — making it easier to see what’s working.
How the platform works
Advanced budget allocation
Budget management works at two levels and depends on how your accounts and goals are set up.
Shared budgets let multiple projects draw from the same pool — with minimum spends guaranteed and the remainder continuously allocated toward whatever is performing best. Requires a minimum budget of €10,000 and can hold up to 10 projects with 3 platforms in each.
Our cross-channel budget allocator decides how to split budget across platforms (but within a project). Connect your ad platforms, choose on one goal, and switch it on. Amanda AI optimizes the split from there.
Reports from the AI
Every optimization Amanda AI makes happens inside the engine. Reporting is the transparent layer — so you always know what’s running, what changed, and predictions of what’s to come.
Every week, Amanda AI generates an AI report — written by Claude, based on your latest data. Not a dashboard export. A document with chapters covering what changed, what was added or removed, how the competition moved, and what the numbers tell you about your account. Read it like a book or feed it to your AI.
Conversion lag gives you a more accurate picture of your conversions before they have been reported. Amanda AI tracks your account’s specific conversion patterns and predicts how many are still on their way. Budget decisions are then based on what actually happened — not just what’s been reported so far.
Performance reports gives you a combined view of your key metrics across campaigns and platforms with filter options — all in one place. And manual hours saved, because that’s a metric worth tracking too.
Audience optimization
The ad platforms’ AI are only as good as the signals they receive. We connect your first-party data and aligns optimization events with your actual goals — ensuring cleaner and better signals to all supported platforms. Audiences are continuously refined based on real performance data without manual segmentation.
Asset generation with AI
Great campaigns need a constant supply of fresh assets. Our platform handles that without external tools.
Generate images with Nano Banana directly inside the Amanda AI platform or let us import copy and images from your website straight into your asset library. You can also upload your own assets and put them to work.
On Meta, content groups let you organize assets for structured testing — making it easier to see what’s working.
“With over 400,000 monthly optimizations, we secured Toyota's position as the market leader.”
Learn how Toyota Denmark increased car sales by 17% YoY,
increased conversions by 65%, and reduced ad spend by 20%.
“With over 400,000 monthly optimizations, we secured Toyota's position as the market leader.”
Learn how Toyota Denmark increased car sales by 17% YoY,
increased conversions by 65%, and reduced ad spend by 20%.
Other key features
Content groups
Shared budgets
AI image generation
AI generated copy
Weekly AI reports
Performance reporting
Floodlights conversion bridge
Future budgets
Enough of the theory. Let’s press start.
Frequently asked questions
Yes.
Brand and generic keywords are typically separated during onboarding to maintain structural clarity and reporting control.
This separation is not handled through Keyword Fencing but through campaign and ad group setup logic.
This ensures brand visibility and bidding strategy can be managed independently from generic acquisition keywords.
Yes.
Brand campaigns can be excluded from organic-based adjustments.
This allows advertisers to maintain full control over brand visibility even if strong organic rankings exist.
Brand protection and efficiency logic can operate independently.
Yes.
Existing campaigns can be taken over by Amanda AI. Historical performance data remains in your account and continues to inform optimization decisions. You can choose this in the setup.
No. Amanda AI always creates campaigns from scratch on Meta. Existing campaigns are not taken over or managed.
Yes.
You can update your project settings at any time through the different tabs in the Project Settings section.
This allows you to adjust goals, budgets, platforms, and configurations as your strategy evolves.
We recommend avoiding campaign experiments. They can interfere with how our system allocates budgets and optimizes performance across your campaigns.
Yes.
If a keyword consistently ranks in the top organic positions, paid bidding can be reduced or paused depending on configuration.
This helps avoid paying for traffic that may already be captured organically.
Rules are configurable, and not all top-ranking keywords are automatically paused.
Yes, we offer a 14-day free trial for all new users. The free trail automatically starts when you:
- Install the app on Shiopify App Store
- Create an account on your website
No.
Amanda AI works alongside Google’s Smart Bidding rather than replacing it.
Google continues to handle auction-level bid decisions through strategies such as:
- Maximize Conversions
- Maximize Conversion Value
- Target CPA (tCPA)
- Target ROAS (tROAS)
Amanda AI focuses on higher-level structural and budget decisions — determining where budget should flow and how campaigns should be organized to generate stronger signals.
Yes. Meta’s AI is still responsible for auction-level optimization. Amanda AI does not replace Meta’s learning phase — it structures campaigns so Meta can learn more effectively.
Amanda AI does not automatically downgrade optimization events just to exit learning. If an ad set performs well but shows “learning limited,” it may be allowed to continue rather than being reset unnecessarily.
In short: Meta handles auction learning. Amanda AI structures the system to improve learning.
