Building two products that complement each other requires careful balancing. Inherent challenges and immediate needs easily tip the scales, and we find ourselves obsessing over one while the other waits. We're deep in Clint right now, yes, but Flint is cooking too.

In Flint's case, we're moving a bit slower, deliberately. As we mentioned repeatedly, most SaaS management tools assume you have API access. We assume our target audience, startups and scale-ups, doesn't have that luxury because we didn't for the most part in our journey before.

So instead, we're building algorithms that make sense of auth token patterns. When did someone last log in? How often? What's the usage pattern that signals 'essential tool' vs 'forgotten subscription'?

It's like being a detective with limited clues, but the insights have to be reliable enough to influence budget decisions and surface overlooked apps.

The reality check: Our beta algorithm sometimes showed a popular app, like Slack, as 'rarely used' while we used it 50+ times daily. Why? No repeat logins are needed; it stays logged in. We're now building smarter logic to understand these types of nuances. It's taking longer than expected, but rushing a half-baked solution would be worse than the spreadsheet chaos we're trying to solve. At this particular turn, we think that not moving fast is the key.

Perfect data is the enemy of useful insights. That's our way of thinking. Instead of pretending we know everything, we'll ask: 'Our data suggests low Slack usage, do you agree?'

We work towards providing the best interpretation we can from limited signals, then let users correct us if needed. The goal isn't perfect analytics, it's making SaaS chaos more manageable than spreadsheets.