Why we built Relay for real time, not reports

Before Relay, I ran growth at a subscription business spending about $400,000 a month on ads. Every Monday we pulled exports from six platforms into a spreadsheet, argued about attribution and made budget changes by Tuesday afternoon. By the time we acted, the auction had already moved on.

Reports answer the wrong question

A weekly report answers what happened. The question a growth team actually has is what to do next, and that question has a shelf life. A campaign that is fatiguing on Wednesday is expensive by Friday. A competitor who pulls out of an auction on Thursday creates an opening that closes by Monday.

We started Relay in 2024 with a simple idea: the useful unit of ad analytics is a decision, not a dashboard. Everything we build starts from a decision someone needs to make, then works backwards to the data needed to make it well.

What real time means to us

Real time does not mean refreshing charts every second. It means three things. Data arrives within minutes of an event, not overnight. Signals are ranked, so the most important change is at the top. And every signal comes with a recommended action and the reasoning behind it.

Trust over automation

We are careful about automation. Relay can move budget on its own, but only inside limits the team sets, and every change is logged and reversible. We would rather a customer approve a hundred good recommendations by hand than wake up to one bad automated decision.

Where we are going

Today Relay connects to more than 20 ad platforms and helps 2,000 teams decide where their next dollar goes. Next, we are bringing the same approach to creative, so teams know which ideas to scale before the results are obvious. If that sounds like a problem you have, we would love to hear from you.