This paper describes a method of mean effort estimation for a class of incident tickets in the IT service domain. A two-stage approach is proposed, where the attention duration is first calculated for each ticket based on a metadata model and some ticket handling priority rules; then the maximum likelihood method is used to estimate the mean effort for a ticket class using such attention information under some statistical assumptions. To validate the feasibility of the proposed approach, we have carried out a simulation study. It yields some promising results. It shows that our approach can produce a better estimation for the parameters of assumed statistical model than the naive estimates, which in turn will give us a better estimate of the mean ticket effort. © 2013 IEEE.