I objected information is not sufficient: why was implemented a new system? Which problems were faced in the previous one? Can I have the business case used to approve the implementation of Teradata? Business reasons should be there.
I won't say what I answered exactly but it was essentially building a data tool that would potentially improve the lives of people. I suggested using first some factor analysis with a cluster grouping to detect which factors to focus while building this tool. The product analysts I've met were not very science-oriented so they did not really understand this solution and instead suggested picking the factors arbitrarily. This was quite shocking because I think a product person with some data scientist background would understand that scientific approach is quicker and better. This was the feeling I got in most of the interviews; it's great that analysts are business-oriented but if you really want to make a difference, you should let the technology decide the approach rather than manual work.
A primary school question, but was actually asked. If a mailing first class cost 35 p per envelope, while third class cost 28 p per envelope, with a set up cost of 3,000 GBP, what is cheaper for 15,000 envelopes? How many envelopes make the two options cost the same?