The test e-mail handed out a Google Docs share link. Not the first time a company chooses this option to share a test so it is not a problem. What was the (first) problem is that the Google Docs still displayed the number of people checking out the document. I believe I've counted 10 other people reading the same document at one time, being shown as 'Anonymous'. While it is understandable that companies will hand out tests to multiple candidates at a time, it shows a clear lack of professionalism to have this visible to the candidate and to not take steps to rectify this (this could be avoided by handing out the .PDF format but that would take additional work by exporting the document from Google Docs). The second problem was the test environment, it very much seemed like a complete product but we were only assigned a single area to test, relationship based connections between entities. The problem with this test structure is that all candidates were given a single test account, shared between them (at least 10 candidates) which was creating problem within the environment (test data disappearing and appearing randomly as multiple people were using the environment at a time). Again this wouldn't be a problem and would be a realistic setting, if not for the number of times my test data kept being lost due to erroneous deletes from other candidates (I don't believe this was on purpose but sadly this highlights the lack of professionalism). The third problem is the document itself, that gives contradicting functionality definitions to the testers. Two definitions are given for a relationship type, i.e. (not exact replication from the document) 1. relationships of X type must have a start date and an end date >=0 2. relationships of Y type must have a start date and end date = 0 The >=0, =0 overlap only served to confuse testers. I've given feedback on all of the above, but must emphasize again that no response was given.