The deadline dates should be sufficiently late to allow authors time to submit relevant, quality abstracts. They should, however, be sufficiently early so that there is time for reviewers to submit their reviews and for the chairs to arrange the final schedule significantly before the registration deadline.
Let's work backwards :
At least two months should be allowed before the conference starts to when the speaker and poster schedule is announced. The schedule attracts people to the conference and allows attendees to plan their attendance.
Approximately two weeks before the schedule announcements, the deadline for reviews should be planned to allow the PC chairs time to create the schedule.
Approximately three weeks prior, the abstracts should be distributed to reviewers to give them sufficient time to review.
Two weeks prior, the deadline for abstracts submissions should be place. This allows a week for a deadline extension and a week to sort abstracts to distribute to reviewers.
Approximately three months prior, the call for conference abstracts should be made to give submitters sufficient time to prepare their abstract.
For us, this means :
July 6 - two months = May 6 <- announce talk schedule
May 6 - two weeks = April 21 <- reviewer deadline
April 21 - three weeks = April 1 <- distribute abstracts to reviewers
April 1 - one week = March 21 <- extended abstract deadline
March 21 - one week = March 14 <- Initially announced abstract deadline?
In relation to the "Full Paper Deadline" we also need to pick a deadline that will allow reviewers sufficient time to review the papers before the proceedings are published. The determination of this date must be discussed with the proceedings chairs, but I recommend one month before the conference.
The deadline dates should be sufficiently late to allow authors time to submit relevant, quality abstracts. They should, however, be sufficiently early so that there is time for reviewers to submit their reviews and for the chairs to arrange the final schedule significantly before the registration deadline.
Let's work backwards :
For us, this means :
In relation to the "Full Paper Deadline" we also need to pick a deadline that will allow reviewers sufficient time to review the papers before the proceedings are published. The determination of this date must be discussed with the proceedings chairs, but I recommend one month before the conference.