3 Questions Every IT Ops Person Hears (And How to Answer Them)


Hello, my name is Geoff Craig and I am an Automation Engineer here at LeanKit. In my previous life, I worked in IT Operations for numerous companies, large and small, on projects of varying complexity.

In every IT Operations group that I’ve worked with, there were always competing priorities from different parts of the business. We kept hearing the same questions about why work wasn’t getting done. We knew the answers, but we never had a way to communicate the problems in a way people would understand — so they never got solved.

One of the reasons I came to LeanKit is that I firmly believe that visualizing the work of IT Operations can change the way we work — and thus, how the value of IT Operations is perceived in the organization.

The Same 3 Questions

The nature of IT Operations work means that we’re always trying to strike a balance between the following conflicting priorities:

  • Providing value back to the business by deploying new software and software upgrades
  • Firefighting the existing systems
  • Keeping existing systems up to date; patch management and regular maintenance

These competing priorities always cause clashes between project work and the “keeping the lights on” work. We hear the same three questions:

  1. Why isn’t work getting completed as planned?
  2. Why does the IT Operations team always seem overworked?
  3. How do we justify an increase in headcount so that the first two questions are no longer asked?

Once I started at LeanKit and committed to managing work with Kanban, I realized that this was the tool that could answer the questions in a way that everyone could understand. It would give us a voice we’d never had before.

Spreadsheet Dysfunction

I’ve worked on several projects that included multiple teams working in multiple locations. We always spent a significant amount of time going back and forth with project managers, trying to track and report status in one of two ways: Microsoft Project, or a comically large Excel spreadsheet.

We would use these tools to update tasks so that they could report up the migration status of a location that was moving to a new messaging system. The problem with keeping track of tasks in this manner is there is no way of identifying or tracking bottlenecks in the process. There could be many:

  • Were we waiting for Outlook to get deployed?
  • Were we behind on copying mail?
  • Were we behind getting information back to begin the process?

If you’ve ever tried to answer those questions using Project or Excel, you have my sympathy. I’ve sat through too many meetings trying to explain where the bottlenecks were in a process and how to fix them. Visualizing the workflow would have easily illuminated our bottlenecks and allowed our team to adjust our work in progress (WIP) to resolve the issue.

Visualizing Work in LeanKit, at LeanKit

A good example of how our team does this at LeanKit is that we limit WIP when we deploy new systems at various stages in the deployment process. We also keep track of our firefighting in a special firefighting lane.  This allows us to easily communicate when firefighting has stalled progress in a deployment, and it allows us to limit our WIP so that we don’t front end a lot of development work and stall our deployment velocity. This allows our team to deploy quickly — and allows our product managers to easily find answers to those three questions.

Visualizing the IT Ops workflow - LeanKit

Beyond Visualization: Optimizing Workflow

The other area that visualization can help is in the area of Mergers and Acquisitions. Most of the companies that I have worked with in the past 15 years have acquired small or medium-sized businesses. The first IT task is typically to get those companies on the corporate standard platform. Those types of projects require coordination between multiples teams: networking, desktop support, application support, etc.

In those projects, we were always trying to figure out where the other teams were in the process. Meanwhile, managers and executives were trying to determine the state of the overall project. Trying to manage a multiple-team project with dependencies, due dates and unexpected issues without visualization was nearly impossible. It always led to:

  • Frustrations at every level, from executives to the implementation teams
  • Missed deadlines
  • Misaligned expectations
  • Difficulties working with and communicating between teams

On multi-team projects, handoffs between teams almost always slow down progress. I’ve worked in many projects where my team did all the back-end work getting systems prepared when moving an acquired facility over to a new email system. We were always having to go to a project manager or another team lead to see their status, so we never knew what effect our work was having on the system.

Visualization changes this paradigm. It shifts the conversation from, “When will you be done with that?” to, “How can we help you finish?”. When that shift happens, you can start to focus more on identifying ways to optimize the workflow via automation or general process improvement. Visualization naturally leads you to finding the bottlenecks and improving overall time to project completion.

Communicating the Value of IT Ops

There’s a great quote from Lean pioneer W. Edwards Deming: “You can’t manage what you don’t measure.” In IT Operations, you can extend this to: “You can’t manage what you don’t measure — and it is extremely difficult to measure what you can’t visualize.”

I’ve worked on many projects where we managed to get things done on time with multiple teams, but the overhead of trying to keep track of the status of all the moving parts caused unnecessary stress and burden on all the people involved. Using visualization to communicate how IT Operations works is the next huge leap forward in transforming not only how we work, but also how the business side of the house determines the value that IT Operations brings to the organization.

Geoff Craig

Geoff is a Senior Automation Engineer at LeanKit, where he works with development teams to continuously deploy new features. Working at LeanKit has helped him pursue his passion of automating all the things. Connect with Geoff on Twitter @GeoffCraig74.

6 thoughts on “3 Questions Every IT Ops Person Hears (And How to Answer Them)

  1. Thanks, these were questions that came up today. Pleased that your entry was only 2 days ago. Presumably its still valid and hasn’t aged into obsolescence

  2. This article is almost verbatim what I live with here in our IT department. Great check on reality and the common nature of these issues.

  3. Geoff,

    I like the approach with differentiating the OPs work and project work. The problem I run into is what about a group like helpdesk? In particular our helpdesk team averages around 1200 requests a month. I can’t imagine looking at a Kanban board with that type of volume on it. Who would you approach this?

  4. Hello Dan, I definitely agree that the volume of a HelpDesk can be very challenging to visualize. Our team used to have a very large backlog that we didn’t triage very well, so it had tons of cards. Since we were flowing so many cards through our regular process, the visualization was a bit tough — and it also threw off metrics.

    I think the first goal you must do to visualize correctly is to design your board so it represents the flow of your process. For a HelpDesk, that can be tough because the process can wind up like a choose-your-own-adventure type of flowchart. You almost have to have a holding lane for things like “waiting for customer feedback” or “waiting for external team interaction” (for example, escalating to a tier II team).

    The second part is how you triage incoming requests. All the cards should dump into a triage lane, and then a designated person(s) moves cards into a “Ready to Work” lane. The HelpDesk staff shouldn’t see the triage lane, though; they should just see the cards in Ready to Work. The good part of using a Triage lane is that you can quickly see how much tickets are backed up and can start visualizing the flow of tickets coming in. You could even try dumping all tickets into a Triage board and then moving the cards into a HelpDesk board Ready to Work lane. Finally, there should also be WIP limits in place on the Ready to Work lane that make sense for your organization.

    The great part about LeanKit is there are several ways to take large amounts of cards and use lanes and boards to visualize what is happening. If anyone else out there uses Kanban boards for a HelpDesk, how are you structuring your board?

  5. Thank you Geoff. What you are saying makes perfect sense to me. I have been looking for a company where the entire IT department has adopted Lean, not just Dev. Do you know of a company I can baseline?

  6. Thanks. I am working in a DevOps shop for an Agile dev group now and had issues understanding how to make a board work for us. This gives me some great ideas and I love the fires lane.

Leave a Reply