this video is part of the process mining use case Series where we'll focus on different techniques to identify processing efficiencies non-conforming activities and Improvement opportunities now approvals are an important part of certain business processes but they can also disrupt the efficient execution of the process and extend task completion times so to to help mitigate these issues organizations should focus on optimizing approval processes within the platform setting clear slas for approvals automating routine decisions and improving communication and transparency among users and approvers process mining can provide visibility into situations where the approval step of the process is causing significant delays so we know where we might want to make some adjustments to the process so how are we going to do this analysis in this example the approval step we are using is at the requested item level so we've configured a multi-dimensional map using both the requested item table and the SE task table so we can focus on both an efficiencies that live in the approval step of the process as well as the task level where the actual fulfillment work gets done for the requested item table we've added both state and approval as activities so adding the approval field as an activity will allow us to isolate situations where there's a long time between the approval request and when it's actually approved and just so you know the accompanying blog post to this recording contains the steps to configure this project that you're looking at here on the screen so what you're looking at here this table on the left or the process on the left this is the requested item right where we're moving through both requests right and this you the request uh activities as well as the state activities and then on the right here we have the actual tasks aligned to these requested items so we can dig into the Fulfillment piece as we start to narrow down the data what we're specifically going to be looking at is this step from request to approved here and we can see if we open this up right we had the th approvals um the median duration is an hour but the average duration is one week and we can see we have some long Runners out here that could be skewing the data but specifically if we wanted to focus in on approvals that are taking a long time what we can do is we can use our transition filters again and if we click on transitions we can come here and we can say I want to look at the approval step and say where it is uh requested first then come in here and say followed by or directly followed by approval is approved now of course we want to look at the things that are taking a long am long time so what we can do is we can add this constraint I'm just going to add a constraint of things that take longer than five days from when the request was requested or the approval was requested and it moved to the approved State you can adjust this based on your use case and we'll hit apply and this is going to narrow things down now to just the items in which the approval took longer than 5 days and I can zoom in a little bit again here and you'll see for these we've got approvals that are taking let's say on average there was 163 of these that took longer than five days and their average duration is two months if I want to start digging into this a little bit deeper now I can start using my breakdowns on the leth hand side of the screen we set this up to look at items the items that were requested the assignment groups and the locations they're coming in but items is the specific one we want to focus on and the one at the top of the list here Tableau desktop license requests the average duration of these is six months and two weeks right that's crazy so maybe we want to focus in on these and we can hit applied and narrow it down to just these items and then when I narrow it down to just these Tableau desktop license requests I can see that my average time between requested and approval is four weeks right the median is one we though and you can see here using the histogram that I've got a whole bunch of these right up front right that are taking less than four weeks and there's just this one outlier here that's really skewing it a little bit so I could choose to dig in and just check out this one but I probably want to focus in on these the majority of them that fell into that bucket of uh four weeks or less so we'll narrow it down to just those 33 but you'll see I still have a five Monon and 3 we average fulfillment time or duration for these T Tableau desktop licenses I can see now at the requested item level one week of that time is packaged up in the approval step of the process certainly we can improve there but now that I've got this in a multi-dimensional map what I can do is I can start actually looking at the tasks associated with those 34 in this case Tableau desktop licenses and and I can start looking for where the bottlenecks are in this process so from open to five days it takes five days before it moves into the pending State and then one week for these specific ones uh to go to close con complete I have these 11 that go directly they don't even go into a pending State they go from open to close complete but on average they're taking uh one month and two weeks then I have these five that took eight months to go to close skipped so there's there's lots that I can dig into at the Fulfillment level but really the purpose of this video was to show you how we could use process mining to start analyzing potential bottlenecks in the the approval request step of the process and then tying that back to the Fulfillment tasks as I mentioned earlier the steps you need to configure this project or a project like this on your own they'll be in the accompanying blog post appreciate your time happy mining everyone