Thoughts on the PTOTD
You cannot expect to get a solution if you’ve framed your problem wrong. Think about this situation – if the root problem is to “Get new customers from the source tables”, approaching the data engineer with a problem of “Can we get timestamps of updated customers?” will result in a binary answer. However, the previous statement might result in a better solution that spans the current domain.
Framing also requires you to understand who you’re talking to… You cannot present a high level business problem (like “Increase conversion rate by 10%”) to a dev team.