Early velocity measures how well you got attention. Repeat measures whether the product was any good. They are different questions and they call for different fixes, but they get reported in the same weekly update and get conflated constantly.
A launch with strong trial and weak repeat is not a marketing problem, and spending more on acquisition makes it worse — you are paying to introduce more people to a product that did not hold the last group.
The reverse pattern, weak trial and strong repeat, is the more encouraging one and the more commonly abandoned. People who find it keep it. Not enough people are finding it. That is a distribution and messaging problem, and it is more tractable than a product problem.
The practical implication is to wait for a repeat cycle before drawing conclusions, and to know what the cycle actually is for your category before launch, rather than discovering it while interpreting the data.