We were running paid media for brands. Somewhere between the fifth tab and the third spreadsheet, we started building GoMarble.
Clients had hired us to think about growth. Most mornings, we spent the first few hours piecing together what had happened in the account.
We started with Ads Manager, then an export, then a spreadsheet, plus the usual question of why yesterday's numbers no longer matched. We checked the launches, found the right creative, and rebuilt the report. Only then could anyone decide what to do next.
Product Hunt launch day.
We pasted in the data and asked what had happened. The answers could be useful. But the setup was tedious, and anything we forgot to include might as well not have existed. The next chat started from scratch.
We had made ourselves another assistant that needed a full briefing before it could help.
GoMarble started as a tool for our own accounts. We connected the data, added the brand and campaign history, and taught it to investigate the questions we asked every week. We no longer had to assemble the evidence first.
We're not crazy. This senseless whiteboard is AI generated
A LinkedIn post brought in more than 5,000 comments. Those conversations turned into a Slack community of over 1,000 marketers. Their questions sounded familiar because they came from people responsible for live ad spend.
Can it watch the account while I sleep?
Can it catch the problem before my client does?
Can it remember how our team works?
Can it make the change safely?
We had built GoMarble for our own accounts. The response made it clear that plenty of other marketers were stuck with the same busywork.
Software can explain a performance drop without putting money at risk. Letting it change a live campaign is different. We are cautious about that, and we think every team should be.
GoMarble shows the work, stays inside the permissions the team sets, and records every action. We give an agent more room only after it has earned that trust.
Product Philosophy
Growth teams judge the work and make consequential decisions. GoMarble Agents monitor assigned jobs and propose actions inside the team's rules.
Palkush and Aditya became roommates in their first year at IIT Kharagpur in 2007. Years later, Palkush and Swapnil worked together at ShareChat. So when the three of them started GoMarble, they already knew what it was like to work side by side.
Palkush has built two companies. Before GoMarble, he worked on monetisation at ShareChat and at ITC. At GoMarble, he leads the product and the company.
Aditya led growth at Rupeek after working at BCG and McKinsey. At GoMarble, he brings that experience to growth and operations.
Swapnil was a Staff Engineer at ShareChat and previously worked at Goldman Sachs. He has built systems used by products with 300 million monthly active users.
Connect your accounts, give every agent its scope, and let the work run where your team already is.
