Agentic ads management, starting with honest lift.
A free, in-browser tool to plan matched-market tests, estimate incremental lift, and turn the result into a plain-language next step. Built for marketers who do not want to trust platform ROAS alone.
1. Plan the test
Upload daily or weekly KPI by market. The planner ranks control markets (correlation, scaled distance or DTW), shows pre-period fit, and estimates the minimum detectable lift and a recommended duration.
2. Read the lift
The analyzer builds a synthetic control (non-negative weights that sum to 1), then reports lift %, incremental units, a 90% interval, optional iROAS, and placebo tests that treat each control as a fake test.
3. Decide the next move
A rule-based panel (labelled honestly as beta, not an LLM) suggests scale, hold or cut, and which unused markets are good candidates to test next.
Your data does not leave the browser.
The CSV is parsed and modelled locally with JavaScript. There is no login and no backend in this version. If you email us from the contact form, that message uses your own email app.
A downloadable sample file is included. It is clearly marked as synthetic data, not client results.
What this version is not
- It does not connect to Google Ads or Meta yet
- It does not call an LLM — recommendations are rules
- It does not replace a carefully designed experiment on its own
- It is a beta: useful for planning and reading a geo test, not a certified causal claim
A path to full agentic ads management.
v0 is the measurement core, shipped free. Later versions can add saved tests, ad-platform connections and an optional language-model layer — only if they can stay on free infrastructure or you choose to go further.
v0 · now
In-browser planner, analyzer and rule-based next steps. No account.
v1 · saved tests
Optional accounts and stored designs on a free tier, if and when needed.
v2 · connections
Hooks for Google Ads and Meta geo data, plus an optional LLM narrative.
Always
Human approval on anything that changes live spend or customer-facing ads.