Metricanic

What is Metricanic?​

Metricanic is a cloud tracker built around three things: Simplicity, Speed, and Modern AI-native workflows.
Track all your campaigns in one place, let AI analyze and optimize faster, and make sharper calls.

What can you do?​

Connect ChatGPT* or Claude straight to Metricanic

You already know how scary-smart AI Agents like Claude and ChatGPT* have gotten - these are the same models being pointed at curing diseases. Metricanic lets you point that brainpower straight at your campaigns.
You connect Claude or ChatGPT* to your tracker and just say something like:

Analyze the campaign “PopAds - US - Antivirus” and advise me on how to make it profitable. Please provide a full campaign report.

And it actually does the work. It pulls the reports it needs, figures out your whole flow, digs through the traffic, and either hands you a concrete plan to get the campaign into profit - or tells you straight that the math doesn't work and you need a new landing/offer, before you burn more budget proving it yourself.

BTW: Metricanic doesn’t force the AI into narrow canned scenarios like “check offers”, “check browsers”, “check pubids” and so on. Based on OpenAI and Anthropic best practices we just give the model instructions for how to query reports. Then it decides what data to pull, what to compare, and what the numbers actually mean.

First time you watch it do that, it’s pretty wild.



Flexible events & postback rules


Metricanic runs on a modern event system - you can pass up to 10 events into the tracker, mark any of them as a conversion, and set up individual rules for each one. So you're not locked into one rigid idea of what a "conversion" is, you decide what matters and how it's tracked.

Event Rules. For any event you can build conditional logic the same way you'd set up a rotator - stack as many conditions as you need and decide exactly what fires. For example, on event1 you could say:

• if revenue > 20 → don't send the postback to the traffic source
• if token1 == 1998-propellerPubID AND browser == Safari → send a postback for 60% revshare
• if revenue == 0 → ignore it
• in every other case → send a flat $2 postback

That kind of per-event control is built for advanced users who want their postbacks to behave exactly how their deals are actually structured - not whatever a simple threshold lets them get away with.




Automation rules

Automation rules help you optimize faster by letting Metricanic act on conditions you set, automatically.

For example:
If a pubid has more than 100 visits and EPV below 0.1 - Action: Mark as Bad

It's there for exactly those moments you don't want to keep opening the same reports by hand, hunting for the traffic segments that are obviously already underperforming. You set up simple rules around your own optimization logic, and Metricanic flags the matching segments as bad the moment the conditions are met.
Right now the actions are intentionally focused - for the moment you can Mark segments. Next on the roadmap: expanding actions to pause/start a pubid, pause/start a campaign, and more.

And here's where it ties back to the AI - Claude or ChatGPT* connected to Metricanic can also tell you which rules are worth setting up for your campaigns, so you spot weak spots faster and spend even less time eyeballing reports manually.




Region: Earth

Cloud-based, with servers spread around the globe, close to your users, landers, and offers.
Redirects fire from ~19ms anywhere and your regional data stays clean.

No servers to buy, set up, or scale yourself: no picking a box, no install, no resizing when volume grows, no spinning up a new server when you find a combo in another GEO.
Found a winning angle in a new region? Metricanic already runs there, and it scales with your volume for you. You get fast traffic and clean regional data without thinking about infrastructure. Benchmarks are on the site.

​

You can also reach us at [email protected],
and the site here: Metricanic — Ad Tracking Software for Affiliate Media Buyers

- Stan

*ChatGPT support currently runs through its ChatGPT Developer Mode - takes a minute to enable.