January 2010 Archives

Ann's OIT email this week included a Kudos that I wanted to share here. From Dennis Skovsted to Steve Cawley:

Jac Campbell and Jim Hall now hold the record for fastest resolution of all essential findings in an audit report. In a single quarter they resolved all of the essential audit findings from the Data Management audit report. I really appreciate the quick turn around.

This was a fantastic quarter for resolution of outstanding technology related audit findings for the institution. This quarter 54% of all technology related audit findings could be marked as resolved. The great work of many of your folks helped make this possible.

I want to thank everyone on the storage/backup team for responding so swiftly to our audit actions! We view these audit reports as a very positive thing, helping us to identify areas where we can do better. And the backup/storage team did that, very quickly. Thank you!

There's an old adage in computing: "Good-fast-cheap, pick any two." It's a somewhat simplistic way to say "You can't have it all." You've all experienced this in some form: A vendor can quickly (fast) deliver a product that's high quality (good), but it will cost you. If you want it less expensive (cheap) then you either have to sacrifice some quality, or wait longer to have it delivered.

I like to think along the lines of another triangle: "Lead-manage-do."

Similar to "good-fast-cheap", the "lead-manage-do" triangle helps us to understand the focus we need to put in each of our areas. To be the most successful, one person really should concentrate on (at most) two of the legs of this triangle: "lead-manage", "lead-do", "manage-do". You've probably noticed that we already follow this practice in OIT: managers are expected to provide leadership within their teams, but don't have logins to the systems their teams manage, etc.

Looking outside the U of M, some example:

* Some large companies have a lead architect role who develops new technology, and provides leadership for using that technology effectively. This kind of lead architect is both "lead" and "do".

Typically, these architects are not responsible for a staff of engineers. They usually work alone, or partnered with another architect, so do not "manage".

* Several years ago, the idea of a working manager was popular. These managers would be responsible for running their department, but also provide some hands-on assistance with the systems (usually database administration, or systems administration.) These managers are in both "manage" and "do".

In my experience, the working manager is focused too heavily on the day-to-day running of the department, not to mention the systems, to provide much leadership for the "next generation" of what they do. They may push for more automation, or to make things easier, but rarely are able to focus on dramatic changes that take their organization to the next level.

Like any analogy, "lead-manage-do" is not perfect. I can't say it's impossible to provide effective leadership, manage efficiently, and still do the day-to-day hands-on work - but it's really hard to get the job done, and do it well. The best way to be successful is to practice one or two of these focus areas.

I encourage you to think about how to use "lead-manage-do", and apply it to your work.

Having recently met with each of you for mid-year performance evaluations, this article from the BBC seems apropos: Bosses can change if staff talk.

The article discusses a survey that explored the relationships between line managers and employees, which have been found to be the most commonly reported cause of stress in the workplace. I think I can summarize the findings in these two bullets:

• Managers who receive feedback from their staff can become more effective managers, including coping with their emotions, empathising with individuals and resolving conflict.
• Managers who do not receive feedback from staff are less likely to change their behavior.

Reading this article highlights that coaching needs to be a two-way street to have the most impact. I already do some level of coaching with each of you - sometimes as coaching sessions, other times as "coaching buttons". But I also need to hear back from you: how can I improve, what are the areas I should focus on, how can I help you reach your goals.

I'm pleased that, while going through the mid-year performance evaluations, we had very productive conversations. I learned about what you want to achieve, and I will try to find opportunities to help you reach your goals. But it doesn't end there, it's an evolving process. My door is open, always feel free to give me some feedback on how I'm doing.

I've previously discussed the importance of coaching and mentoring.** I'll take any opportunity to extol (at length) the benefits of coaching sessions.

A coaching session is time intentionally set aside to do coaching with someone. Asking probing questions, the coach helps the "coachee" to work through problems. These coaching sessions work best if you can find a quiet place to meet, away from distractions.

Where a coaching session is not possible, I recommend taking advantage of an available coaching opportunity. For example, during a 1-on-1 status meeting with someone, you may find an opening in the conversation to ask a few open-ended questions about a relevant issue, bringing you to an impromptu coaching session. A coaching opportunity can present itself anywhere; look for them.

You may also hear me talk about a concept I refer to as "coaching buttons". This is a variation on a different concept ("conversation buttons") for how to impart information to someone when you don't have a long time to spend with them (like, during an elevator rid.) The "conversation button" is something that sticks with your listener. Like the button on a shirt or coat, a "coaching button" doesn't do the whole job, but over time as you use more "coaching buttons" the whole picture comes together. They key is to make those "buttons" easily understood and memorable, able to stand on their own, but part of a larger story.

I've adapted this into "coaching buttons", or coaching of the moment where I take an available opportunity to do a brief coaching conversation with someone. For example, a manager might find him/herself early for a meeting, only one staff member is there, giving the manager a short time for a "coaching button". Never waste an opportunity for coaching, however brief. The "coaching button" might only cover one question without an opportunity for follow-up questions to delve deeper - but if you can find frequent opportunities for several "buttons", I find it can be helpful.

For example, I use "coaching buttons" when I meet with the fraternity that I advise. I show up a little early for each House meeting, and use the extra time to do brief coaching with the guys. Usually, they come to the meeting room in ones and twos, making it easy to have quiet conversations.

Closely related to "coaching buttons" is the concept of conversation "gifts". Just before Christmas, Roger Schwartz published "Giving and Receiving Gifts in Conversation" in his Fundamental Change newsletter. Jim Bruce (my mentor from ITLP) included this with permission in last week's Tuesday Readings email, but let me provide a summary. I think this is tied very closely to "coaching buttons":

• Read Conversation "gifts" on the ITLP Tuesday Readings blog.

In November, Donna Fitzgerald from Gartner gave a presentation at the Carlson School of Management. I wanted to share some thoughts from her slides. Donna's talk was titled "Herding Cats: Practical Approaches for the PMO" [Project Management Office].

It's a long presentation, but I think the following slide sums it up nicely:

The Problem

• When IT is a cost center, all IT resources are free
• Every business unit is entitled to 100% of the available resources
• Having too many demands means the common resource of IT is "over-grazed," leading to diminished throughput

The Solution

• Limit access to resources based on strategic importance
• Establish means to allocate resources based on needs of the organization as a whole

We've tackled this problem before. It's also somewhat related to my earlier post, Why it's hard to say "no".

A large portion of Fitzgerald's presentation addresses work planning - it's a great way to help an organization identify the work that needs to get done, versus the work that we'd like to get done. You may also have heard of this as "Critical v. important". And of course, we have been doing quarterly work planning since July. Hopefully, you have started to see the benefits to this level of prioritization.

We often view IT as having limitless resources, but really that's not the case. Our time is often our limit - there's only so much we can get done in a work day. So, we need to prioritize what we do, what we're going to focus on.

For more than a year now, I've made Simplify, Standardize, Automate a priority for the Operations and Infrastructure teams. This helps to address the very real problem of "how do we manage an ever-growing set of responsibilities with the same size group?" In other words, this helps us to put strategic importance on the things that we do, the services that we offer, and allows us to allocate our resources more effectively based on the needs of the university as a whole.

Following up from the earlier announcement about scheduled DR tests, John has gotten the OK from the CAB for conducting a DR test on the 1st Sunday of March (during the maintenance window 6:00am - noon.)

We will go ahead with our plans to test the e-Research application. Please begin to identify and allocate the appropriate resources in each of your groups. The DR team will schedule a meeting in the next week to lay out a planning schedule.