The first version of OneSage solved a problem we thought businesses had: slow reporting. We built clean dashboards, automated charts and morning updates. The data was accurate. People tried it. Then many stopped opening it.
At first, that was hard to understand. The product did what we had designed it to do. But we had mistaken a symptom for the real problem. Business owners did not want another place to read a report. They needed to understand what was happening and decide what to do next.
The morning that changed the product
At a trading company in Dubai, an owner walked us through his routine. He opened accounting software, a CRM, a spreadsheet, email and a team conversation before he could piece together the day’s picture. It took roughly 45 minutes to reach the question that mattered to him: had anything gone wrong yesterday that needed attention today?
That question reframed our work. The information existed, but it was scattered across tools. The owner had to combine it manually before making a decision. A polished dashboard in one more tab would not remove that work.
The data was there. The answer was hard to reach.
We spoke with about 40 more business owners across distribution, manufacturing, services and trading. Many already had an ERP, accounting system or CRM. They could find individual records, yet cross-system questions still required a report request or a long manual search.
People were asking business questions. We had been building a better way to display data.
Even a finance team with a clear view of cash and receivables could struggle to judge whether a particular trend needed action. The missing piece was the connection between records, context and the next decision.
Building around the question
We changed course. Charts and reports still have a place, but the starting point is now a conversation. Connect the relevant systems, ask a question in everyday language, and follow the answer back to its sources. Someone can ask which invoices are overdue above a threshold, how this month compares with last month, or why a margin changed.
In one early use, a margin question surfaced a supplier price increase that had left a product line selling below cost. The records had been available for weeks. What changed was the ability to connect them in response to the question.
What we are still improving
We have more work to do. Onboarding should be simpler. Connector reliability varies by system. Agents work best when a task is specific and can be confusing when it is vague. We keep refining those parts with real teams and real workflows.
The lesson that has stayed with us is simple: a business does not need to think like a database to use its own information. The best interface starts with what a person needs to know, then helps them move forward.