IESG and AD Statistics

“You have to know how you are doing to improve”
Document Stats IANA Stats IESG Stats Lifecycle Stats RFC Editor Stats

The Purpose of These Statistics

The author of these statistics created them in effort to understand how he or the IESG performs today in some of its tasks. How long does it take to get a document approved? Where are the bottlenecks?

The primary use of these statistics is to create awareness of what is happening in the measured processes. The AD or people who work with him may have their own perceptions of how the process is working. Often the perception may not match reality. For instance, small delays here and there add up to a considerable overall delay.

Another potential use of these statistics is to create goals for improvement. For instance, the author has a goal to push his overall document approval delay under three months. The intent is to track actual end-to-end delay from publication request to approval, rather than set arbitrary timelines for individual steps in the process. It is also expected that reaching these goals is realistic, given that the starting point is known from previous measurements, and because each AD can focus on their own situation; the size and situation of different areas can be quite different.

Another potential use is to be able to answer questions about what causes problems. Can a change in the number of document approvals from the IESG be explained by a particular reason? Where does most of the time go in my own document processing? Is waiting for last call to end or IESG telechat date significant factor in overall processing delay? How do the processing times between documents that did not have to be changed and other documents differ? This will also point out possible areas of improvement.

Finally, these numbers provide additional transparency to the actions of the IESG and ADs.

Details of the tool's measurement mechanisms are available in the tool description.


These statistics merely reflect what is easily measurable. They are not applicable as a simplistic model for deciding what is efficient or ineffecient because of the following reasons:

  • The ADs have tasks much beyond mere document processing. For instance, they are expected to guide the direction of the area. These tasks cannot be measured.
  • It is not clear that the goal is a large number of approved documents. The goal is produce high quality documents, and the right documents. Is an AD who producing a small number of documents doing the right thing by pushing back on bad ideas, or is he inefficient? The numbers do not provide an answer.
  • Areas and workloads differ. An AD may have anything from few WGs to 20+ WGs and BOFs.
  • There are a number of measurement difficulties. One is that there is great variation between documents. Many ADs have problematic documents that have taken three or more years to fix. How should such documents be represented in the numbers? Another issue is that it takes a while for new ADs to get up to speed, and ADs who are stepping down typically produce a significant number of approved documents right at the end of their term; its hard to account for systematic differences such as these.

As a result, it is inappropriate to use these numbers to rate different ADs or label ADs good or bad. It is also inappropriate to focus only on these numbers and attempt to improve them while forgetting the more important parts of the work that the ADs do.

The Actual Statistics

Are you sure you have read and understood the above limitations? If yes, please proceed to the actual statistics.

The IESG and AD measurements software, bandwidth, servers, and source data have been brought to you by:

Jari Arkko   Ericsson   IETF   Henrik Levkowetz