Steve Player interviews Gary Crittenden, CFO of Citigroup, Inc.

Player: From a CFO point of view, you've had a varied career. You started in consulting with Bain and then you've been the CFO of some very different organizations, including Sears Roebuck, Monsanto, American Express, and now you're here at Citi. What have you done to enable yourself to transition across such diverse industries?

Crittenden: Well, I was in consulting for 12 years -- and one of the nice things in consulting is that you actually work in many different businesses and industries. When you're a strategy consultant, the primary skill is to understand what the levers are that will influence the financial performance of the company, because you have to go in within a relatively short period of time, usually within 30 to 60 days.

You have to understand what the key financial levers are so that you can sit down with the CEO and have a conversation with him and have him feel like you're actually adding value to his thinking. It's kind of an audacious thing really for a strategy consultant to try to accomplish, but this is what most strategy consultants are trying to do.

I had 12 years of immersion in consulting before I had my first CFO job. In each of these different businesses there were very different elements, but there was also a lot of commonality. The levers that drive financial performance were different. The key factors for success were different, but the finance was essentially common.

The complexity of finance increases as you move from one business to another, and I would argue that financial services is one of the most complex types of finance. This is particularly true when you're involved in the markets and banking portion of finance, with derivative trading and securitization and all that goes along with that.

It probably would have been less interesting to start with Citi and to go in the other direction to retail. So I started in retail and then moved from retail to manufacturing to credit cards, to now one of the largest financial institutions in the world. These roles have provided new insights to learn from. Each one of these roles became increasingly more complex, and I had to learn just what the initial complexity was.

Player: In certain companies, finance people are sometimes relegated to a down-and-dirty transaction processing function -- and not the kind of organization I think you were trying to aspire to. How do you design a program to lift up the finance organization and literally transform it into the forward-looking strategy support unit that you seemed to describe?

Crittenden: Well, the first step is to make sure that the way people spend their time is dedicated toward higher-value-added activities. If you take a traditional company -- and go back 10 years or so -- you will find that reporting was done in lots of different communities. Each one of these supported an individual business line. If you had different divisions, each of the divisions had a different practice in the way that they approached their reporting.

In order to roll all that up, you had to have an army of people, and basically all that army could get done was to pull that information together and make sure that the information accurately reflected where the financial results of the company were. So the first step is to make sure that you pull all that less-value-added information together and put it into a few locations.

At Citi today, for example, we're in the process of migrating out of our business lines and into a few centers around the world -- Costa Rica, Manila, Tampa, and Buffalo -- where we can do things in a very common way with common processes, where we can measure and control the quality around each of those processes.

Two things happen when you do that. One is that your costs go down because you're doing it with lower cost resources. As a result, you're able to do it with fewer people and you can now get scale effects that you didn't have before. The second thing is that the quality goes up at the same time -- because you're now doing things in a controlled environment, with process control measurements.

Player: So you're using standard processes, consistent definitions, data-driven decisions, etc. ...

Crittenden: Yes. It's more of a fact orientation. I've always been able to then persuade the CEO to allow me to take some of the dollars that were freed up by making that transition and invest a portion of the savings into developing a finance organization that was more forward-looking. You can't have one without the other. You can't just spend more money on finance. You have to justify your existence by spending less, and then you take the difference and you invest it back in some of the higher-value-added things. So that's what we have tried to do.

If you think about the things we're trying to focus on now ... we say all the time that our number one job is to drive the financial results of the company. We should be the ones who make our living by trying to look around corners and trying to anticipate what's going to happen in the future.

I say all the time to our team that if a train wreck happens, it's interesting to go back and report why it happened. The key, though, is to try to do things that avoid the train wreck altogether. This is where the real value is, so how do we do a better job of anticipating how the environment is going to change and making sure that the business is properly positioned for how the environment is going to change? Certainly, given the experience of the last year and a half -- which has truly been extraordinary -- there has been more pressure to anticipate the future with greater accuracy than ever before, and this has given a lot of life to that kind of activity. We focused hard on that.

The second thing we've worked on is to try to improve the processes that we have, and clearly there are lots of processes that exist within a finance function. You have tax compliance, treasury operations, all of the accounting and reporting, all of which are really end-to-end processes that can be co-located in one or two low-cost facilities around the world and operated as a low-cost process. We've had lots of opportunity to do that, and we still have a lot of opportunity ahead of us to try to improve. The results there can also be measured. You can measure the quality level of the process, the head count that you can take out, the productivity you are able to achieve, and so on.

The third thing that we work on is to strengthen the control environment. We always have to be concerned that the control environment is sound and that there is adequate process around everything we do so that there is not a risk of major fraud or loss to the company or inaccuracy in our financial statements. There are a lot of systems elements to ensure that you have the right control environment.

The final major objective we have is to ensure that we continue to make this a great place to work, that people rotate jobs, that they have the opportunity to get ahead within the finance function, and that we have processes in place that actually allow for these developmental experiences for people.

I think that I'm giving you a very long answer to what was a short question. Part of it is to have a really clear agenda for what you're trying to do and to have it be a pay-as-you-go kind of process. It creates a value that you need to have in order to pay for the activities that you do, and then to demonstrate that the finance team is actually making those contributions to the organization overall.

Player: How do you keep the strategic agenda on track in an environment where there have been a number of crises that the business has had to ride out? How do you keep everybody on track?

Crittenden: The honest answer is that you don't. I mean, you can't. You can't just say, "Gee, it's too bad that all of this stuff is happening in the real business, but we're just going to continue on with the finance agenda." You have to prioritize things.

But I've always carried around this construct in my head which says that if I put everything into a two-by-two matrix -- where on one axis are things that are of importance and on the other axis are things that are urgent -- there are an awful lot of things that are not important but act like they're very urgent, right? Some things appear to be very urgent.

Then there are other things that are important but are not urgent at all. Moving work to low-cost locations is very important to your long-term success, but it's not urgent because there are people who are doing that work today, and it's going to get done tomorrow whether you move it or not. It's important but not urgent.

There are some things that are important and urgent. If you have a business crisis -- and certainly I think what has happened in the financial services broadly over the last year probably could be termed a crisis -- then you have something that is important and urgent that you have to deal with. This then has to take the front seat.

At the same time, you have to make sure that your goals remain intact -- that you continue to make sure that the organization understands that these are the goals that you're heading toward and that you need at least the minimum staffing necessary to make some kind of progress relative to those goals. Now, in this last year, we've raised $50 billion in capital, which is more than the capitalization of many individual banks in the world. Meanwhile, we've obviously had a very difficult financial environment to cope with at the same time, so this has been a full-time job.

I believe that we have made good progress on the other agenda, not as fast as we would have made had we not had all of these things happen, but you know, you do what you have to do and try to push forward on all fronts. Part of it is that you can accomplish a lot as long as everybody maintains their own lane. There is a little bit of a tendency in a large company to have everybody run to the ball.

It turns out that if you have to raise $50 billion, not all 8,000 people in the finance organization need to be involved in raising $50 billion. There are a few people who have to be actively involved. The treasurer has to be actively involved and a few people on his team, but because that is happening, it doesn't necessarily have to distract other people who have other normal job activities.

This effort that we've had to migrate to low-cost locations has really been driven by our company controller. He's had a busy year, but then again he has a busy year every year. He has been able to balance the demands of both day-to-day activities as well as more critical events of the last year to advance the overall agenda.

Part of it is ensuring that you are reviewing the goals that people have, that you set goals with them at the beginning of the year in spite of the fact that other parts of the organization are a little bit distracted.

The example that I always use is from when I played football in high school. At the end of practice, the coach used to have us always line up and put our fingertips end-to-end, and we would walk down the field, and you were responsible for picking up any paper garbage that was in your lane and throwing it over into the garbage can.

There were two points that he was trying to teach. One is that you had to clean up the field and everybody had a responsibility to take care of their own thing, but secondly it was that everybody protected their lane in football and you never wanted to get caught by a reverse, right? So, if you had the responsibility of protecting that lane, you owned that lane no matter where the football was.

It's the same basic idea that applies to doing work in an environment that is really complex or where there is a crisis going on. Everybody can't run to the football. Some people have to be here to advance the game at this part of the field, so just reinforcing this idea of staying in your lane, of not letting things distract you, allows you to continue to make progress -- maybe not at the same rate you would be making it if there was no distraction, but it still allows you to continue to make progress.

Player: We've seen a lot of finance organizations drive out and reduce transaction processing costs and free up a lot of head count, but they don't get the second half of it completed. How do you convince the CEO that this is a good investment?

Crittenden: The first thing that you have to do is to sell it as a package.

Player: Don't unbundle ...

Crittenden: Don't unbundle, right. If you run down the head count and you take the expense out, then you've got a much tougher sell on the upside. So part of it is that you have to make sure that everybody understands what the strategy is overall. The strategy that we're following as a company for reengineering is very similar to what we have in finance.

There are other areas of the business where we're also removing head counts, but we're investing in sales force or technology or new products, and I view this as exactly the same thing. What we're essentially trying to do is to change the geography of the P&L.

The P&L today would have administrative costs and technology costs that would be higher than they need to be over the long haul. What we want to do is to bring those costs down and substitute in their place things that move the business forward and decision-support types of activities, more fluid planning processes, sales force technology -- all those kinds of things that are positioned to help the company go forward.

The best way to get people to reengineer is to make a commitment to them that if they do a good job of reengineering, they are going to be able to reinvest back into the things that they want to do. This is just one of many parts of the company that are going through this exact same process -- and although I said that you shouldn't unbundle, I was really joking. It really does require you to be extremely clear about what you're doing: This is how much cost you're taking out, this is what you're reinvesting back in -- and you believe that the things you're reinvesting back in will give a big return.

The thing that people don't understand is that people tend to manage their capital really well. Maybe you're at a manufacturing company and you're managing capital expenditures or whatever, but it turns out that there are many more dollars on the P&L. We don't scrutinize the P&L dollars. We don't call every P&L dollar into question in the same way that we do every capital dollar.

Player: So the normal operating costs, the day-to-day stuff ...

Crittenden: Right. Make it flow. What you have to do is change the mind-set and understand that it's not just the tenth salesperson you hire that counts, it's the first. Do you need the first one, and the second one, and the third one, and the fourth one, and the fifth one? When I've looked at this in situations before, I would have thought, "Maybe I have too many salespeople." But I analyzed it, and it turned out that I had way too few salespeople. I actually would have needed to triple the number of salespeople before I got to the point where I had diminishing returns.

If you have an analysis that focuses on the expense dollars in the same way that you scrutinized the capital dollars, you really get a lot more bang for the buck. If you do it well -- then you can pretty quickly convince the CEO and the division heads that the best money they can spend is to make every dollar that they spend more efficient. They get more bang for the buck out of every dollar that they invest, and that's what this is all about -- trying to improve their results.

Player: When you're talking about the processes that drive success, you mention different kinds of the budgeting practices. Certainly, you have overseen different approaches in the past. What are you guys trying to do in planning, budgeting, and forecasting?

Crittenden: It ties back into this primary thing, which is that we're trying to drive the performance of the business. There are several different elements to it.

One is to have a rolling forecasting process so that you're always looking further ahead than you normally would. The way this starts for us is with a strategic plan that looks at a couple of years. This then turns into an annual plan, and then the annual plan gets refreshed each quarter, out for that quarter plus then the additional six months on the end of that, and then we update this twice a month as part of our normal process.

We're always looking forward and doing a normal update to this. Now, if you're going to be updating this frequently, you can't do it from a bottom-up basis. You quickly conclude that you have to use the primary drivers of the business as opposed to the detailed, kind of bottom-up forecasting. It turns out, paradoxically, that doing this is just as accurate as the bottom-up forecasting. In fact, you probably get more accurate data because you're doing it a lot more frequently.

Player: Some people have claimed that it's more accurate because there are just fewer minutiae ...

Crittenden: Yes, and you're doing it all the time, so that driver-based forecasting is a big part of it. We then couple that with a reengineering effort to bring down expense, and then we think very carefully about every dollar that gets added back that offsets the reengineering.

This process is called investment optimization. We go through and look at the expense dollar optimization and ensure that we have a common definition for expenses across the company. I'm talking about the end-state now -- not where we are, but where we're headed.

In a company that is this large and this complex, even getting the definitions common is important so that somebody who is making an investment in a fixed income trader in, say, Brazil has the same financial metrics, terminal value, and present value calculations.

This is a big part of what has to happen as part of this process. But once you have this, you can then align those opportunities and say which one you prefer. You also know what's on the margin -- what the 20 last things were that I approved that would have the least impact if I had to cut them. You also know what the 20 next things are that you would approve if you had some dollars. And if you're constantly updating and reforecasting, then you're always able to ask yourself the question, "Do I need to cut or can I have the opportunity to add?"

All of this ties in with a flexible planning process through which you hold back on spending until you have the revenue. This is, I think, the hardest concept for most companies to adopt, yet we do it all day in our own budgets. Right? None of us at home spends money until we know that we've got the money. In fact, even if you're getting a bonus, you don't spend that bonus until that money is in your bank account.

Player: For instance, if you want to buy that big-screen TV, but you are waiting until the check comes in ...

Crittenden: You're waiting until the check is here because you know that regardless of what might happen, you're not going to be committed until you know that you've got the money. It works the same way in a company. You need to hold back on spending until you see that you have the revenue, and then when you see that you have the revenue, you trigger the spending.

People would say, "Gee, that's very inefficient." Well, it turns out that it is inefficient, but over the long haul it's the only way that you can meter out your expenses in a way that allows you to have some control over what you're spending.

Having expenses that are highly variable that can be turned on and turned off over time is a key component to it. What I've described is an overall financial management process, kind of end-to-end, and this is what we're working hard to make happen here.

We're coming from a place where we had a very traditional finance structure with finance embedded in each one of the businesses -- and where our primary role was to do reporting. We're working to get to a very different spot.

Player: The traditional annual budget process ... how long does it take?

Crittenden: You know, I've only been through one of them, but I understand that it can be a long process. We are focused on shortening the length of the process because we don't want to give people the time to wallow around in the budget.

We want people to have to come to a conclusion relatively quickly. Hopefully, we have a strategic plan already in place. We already have a pretty good idea of what we're trying to achieve in the next year, so the annual planning process hopefully dovetails into this nicely and then fits in with our normal kind of driver-based forecast in the semimonthly updates that we do.

Player: At what point will you think that you have your driver-based rolling forecast the way you want it to be?

Crittenden: It's going to be a couple of years. I mean, think about it this way: We have today about 350,000 employees with $100 billion in revenue. We have 8,000 finance people around the world, and we're talking about a very fundamental change in the way in which we approach our work.

So it certainly won't happen overnight, but there are lots of early wins. There are lots of things to give us encouragement that we're moving in the right direction, but for the organization to believe, you have to really work and work and work at it. Obviously, we have to save the money to be able to pay for some of the things that we're trying to accomplish, too, and it takes time to get the work to move to the right locations.

Player: In terms of the early wins, what are some of those? I think that you've touched on some of those. You've begun the consolidation, now you're working hard to pay for it ...

Crittenden: Yes, we're working hard to pay for it, and we have hired some key people who are going to run our investment optimizations process in the company. We've been fortunate to attract three or four people who have really good backgrounds in this.

What we're doing for this forecast cycle is that for the first time we're looking at what our incremental reengineering savings are and what the proposed incremental investments are. We'll be making a decision with this upcoming forecast about what we're going to spend of our incremental reengineering dollars on that basis. Now, that's a small step.

It's a far cry from challenging the first dollar invested in a sales force, but it's starting down the pathway of being able to judge the first dollar that's invested in a sales force.

Player: How long will it take to show people what it's going to look like?

Crittenden: A couple of years from now, we'll be there, and you know, people's confidence increases over time. When people see that they're adding value to the business in a different way, they start redefining the way that they do their job. The job is a whole lot more fun than that reporting job that used to involve a lot of reconciliations and a lot of late nights and that kind of thing. People begin to really get the flavor for the kind of value that they can add. It's a contagious thing. People want to do more of it.

Player: It sounds like it's more fun when you're working on a problem you find more interesting, more strategic to the business ...

Crittenden: And you are viewed differently in the business. You know, the function starts to be viewed differently, and everybody feels that it creates a certain sense of enthusiasm on its own. But we have a long, long way to go to get there. I wouldn't want to in any way give you the impression that we're even close to the finish line on this, but we know where we're trying to go.

We have the right people in place to try to make it happen, and, importantly, we're well on the pathway and starting to pay for it.

Player: Let me ask a closing question related to your finance career and where you've gone. You've had such interesting experiences. What leadership lessons have you learned along the way?

Crittenden: It's hard to choose one. I think that from my perspective I started thinking about finance because I followed on the heels of Alan Lacey at Sears, who was a really great CFO. When I took over after him, I thought, "These people have high expectations of me." He did a really good job. People will tend to rise to the level of expectation that you set for them, and Alan had set a very high bar for what he expected from that team. It was good for me to follow him because he could get people to actually aspire to a lot because he himself had high expectations.