Your analytics count function calls. The reading layer sees journeys: who arrived, what they tried, what they did next, and which of your “users” are users at all.
A quarter of all activity runs on someone else's gas. If you count addresses, you are counting budgets.
transactions — 9% of the whole chain — were one app's daily check-in button, pressed by 88,169 addresses. Every dashboard counts that as engagement.
Between every pair of steps: the exact place users stop, what stopped them, and whether it was your contract, their wallet, or a broken integration — with the cohort you could win back.
Where exactly do new users stop — and why?
Which integrations send us malformed transactions?
What do our best users do after they leave?
Run a protocol? Which of these would move a decision this quarter?
bartosz@research.tech