A dataquest is useful to understand what users find interesting or useful in a certain dataset or group of datasets. This then helps you design a channel to this dataset that matches user's needs.
Say you have a dataset of land ownership in the country and you want to understand what citizens and other user groups find the most interesting and usable about it.
It typically takes the form of a day event (although it could be shorter or longer). People arrive, usually with laptops or access to laptops and the internet, and either individually or in teams explore the data and produce outcomes by the end of the session. These outcomes depend on the type of dataquest:
A data easter egg hunt is where you have set questions based on the data and it is up to participants to see how many questions they can answer in the time. You could then create a form for participants to fill in about what they found the most interesting and/or surprising about the data.
An open-ended dataquest would rather require the participants to present a data story at the end of the session, and during the time they create insight out of engaging with the dataset, and combining it with other data and contextualising in a human story. This could be how land ownership intersects with gender, or the changing nature of land ownership over time, in our land ownership dataset example.
In both cases, what you are looking for is to assess what users of the data find interesting about it, and how they might use it. Did they need it to be geo-referenced? Did they need it in a certain format? Was there a very common need that a guide, toolkit or data visualisation could be made for?
Ultimately, this starts to inform you what value a dataset holds for users, and some good ways of sharing a certain piece of data and information with users.
Here is a video of a dataquest run as a part of the National Open Data Portal Development in 2015: