Digital Transformation Needs New Metrics

Jim DuBois
Former CIO
Microsoft

In part three of our four-part series (part one, part two) dissecting former Microsoft CIO Jim DuBois’ book on IT leadership, Six-word Lessons to Think Like a Modern-Day CIO, and how it relates to digital transformation, we look into how traditional approaches to measuring success in today’s fast-moving, digitally-fueled business environment requires new KPIs that are as flexible as the outcomes they are measuring.

 

Thanks for making time for us today. Today we’re talking about chapter seven from your book, Six Word Lessons to Think Like a Modern-Day CIO, which is all about measuring to ensure success. To me, this seems kind of obvious. Peter Drucker and all that. Can you explain why you decided to include this concept in the book?

It’s obvious, and people do it, but I don’t think that we’re always measuring the right thing. So, part of the importance is to make sure that we’re measuring the right things. The example I typically use is, historically, the way IT is measured, how success is measured, how people are rewarded, is based on delivery — on-time, on-budget, on-scope — of something that will be highly-available, secure, and compliant. There’s a little more to it, but that’s basically the way that IT for decades has been measured.

But, at the pace that things are going now, by the time we actually deliver something that the business has ordered, what the business needs is actually different. The concept of measuring, “Did I deliver what I said I was gonna deliver on-time, on-budget, it is still an interesting measure, but it’s not success.”

Success is what we are trying to accomplish by doing this. So, what are the right measures for that? And then let’s not celebrate that we released something, let’s celebrate that what we are trying to accomplish was accomplished.

 

From a business point of view?

Yes, from a business point of view. Whatever measures we put in place on the business side, that we actually see those measures improve. That’s success. What we really need is to figure out is how the business is measuring success and measure IT success based off of the same thing.

When I got the chief marketing officer at Microsoft to work with me on rewarding my team based off of the same measures of success that his team was being measured on, it did a couple of things. First, it ended up building trust in the marketing organization because they saw that we were measured on the same things that they were. It also changed how my team worked because they could bring in new ideas and technologies to improve those business metrics that marketing didn’t think to ask about.

By changing the definition of success, we’re actually changing how decisions are made, how the hierarchical and political powers work within companies.

The model I’m advocating does take away some of the control from management. It substitutes the ability for everybody to contribute ideas. The control ends up coming more in how you prioritize, rather than how you dictate what to do. We get a better solution if we’re clear on what success looks like and we let everybody contribute ideas towards that solution.

 

The challenge for big companies like Microsoft is they have to hire to support their existing customer and product base. Often times, those roles require a certain type of person, someone who is more familiar with a more hierarchical way of working. That’s how they understand the world and they’re really good at it. That’s why they were hired.

This gets a little bit into what I talk about in the last chapter on people.

I made people the finale of the book because it really is all about the people. Your question here hits at the heart of being able to do a lot of this, so you have to have people that can both innovate for the future and perform for now.

 

How does this tie back to measuring for success?

There is a mindset that these people have. They deal with ambiguity better. They adjust better to the pace that things are going today. This ability to understand what we’re trying to get done and be adaptable to the best ways of getting there, that is a different mindset than just, “I understand my marching orders and I’m going to get that done.”

 

Is this well understood by IT organizations?

No. Some of it is just because things have been so drilled into people for years on how they are going to measure people in IT on success that it’s really hard for them to see there’s actually a better way to get people focused on real business success, as opposed to IT success.

 

But digital transformation goes well beyond IT, right?

Absolutely. When I was speaking at an event in London last summer, one of the other speakers was a headhunter for CIOs. She said every company she talks to that’s looking for a new CIO is in the middle of a digital transformation. The question she asks all of these companies how they’re measuring their digital transformation. Based off of that, she can find them a CIO that will actually fit what they were looking for.

If they defined their digital transformation with real business metrics, and real transformation of the whole company, and the company’s business model, and all of that, that was a particular type of a CIO. On the other extreme, if they described their digital transformation as fixing IT, and getting IT to be more available and faster, that was less of a digital transformation CIO and really more of a clean-things-up type of the person. If you put one of the really good digital transformation CIOs into that role, it’s not going to be a good fit.

 

How do you come up with the KPIs that you need to look at especially if you haven’t done it before?

It has to start with the vision of what you’re trying to get to and being able to come up with the measures of quantifying the vision. And then you would pull that back into a metric for where you are now versus where you want to get to. That becomes your measure. But you have to break that vision down into the component parts and then be able to measure those parts.

Then you have to decide how to prioritize things. That goes to the overall measure as opposed to the component measures. I saw many cases where we were trying to optimize something that would make a particular department in the company more efficient, but it didn’t help the overall metric because that department wasn’t the bottleneck.

You have to think about not optimizing just the components. It doesn’t always help the overall success. You have to make sure that you’re thinking about the impact of individual optimizations towards the overall success when you’re prioritizing what to do first.

You may find that you can improve things that you had originally targeted in the middle-term faster and move them into the short-term. And you may find that something that you thought was a short-term thing actually needs something else done first, so you have to move it out to the middle.

This goes back to one of those two big differences in the historic model versus where we need to go. The historic model would have said, “Here are the things we’re gonna work on first, and here are the things we’re gonna work on second, etc.” The Grand Plan model.

If people brought up things that were in the long-term improvement, they would just be dismissed as, “No, we’re gonna get to that later. Focus on these now.” Where in the model that we need to get to, we need to recognize that creating a plan like that, it’s going to be imperfect, and things are gonna change as we go, and we’re gonna learn new things.

We may find that what we thought we could do in three years, we can do in two years if we do some other things. So, don’t convince everybody that it’s a three-year plan. Convince everybody that the current view is that it may take three years, but we may find some ways to accelerate that or push something was out that wasn’t as important as we go.

 

»Read next, on Apptio:

  • Five cloud computing predictions for 2019
  • Forrester’s “Transformation is Pragmatic” report: What to expect in 2019
  • Five cybersecurity predictions for 2019
Article Contents

Categories

Tags

Additional Resources