ADVICE FROM THE EXPERT: -> Lesson #1: Don't underestimate the gap between the executives deciding to outsource and the people who need to do the work.
In your book you advise trying a first time outsourcing project that requires little day to day interaction or that's well documented. Are companies actually following that advice or are they seeking something with more dramatic results?
Ravi Kalakota: Well, there's a sort of a gap between the people who make the decision and the people who have to live with it. And the people who make the decisions are really looking at these paper analyses about how much the cost cuts are going to be or value delivery they are going to get. But the people who have to live with it day to day, they are experiencing a whole different set of issues that the people at the top don't see. So that’s the tradeoff.
When we normally talk about offshoring, and we start talking about execution-related issues, invariably we find that the executive level doesn't really care to think about those issues or even is worried about those issues. It's almost like they think it will get done. So that is really the dilemma…
The executives making these decisions, how in touch are they with what is being outsourced? Do they really understand the true outlay of IT expenditures or the strategic value of their investments in IT?
To answer that question truthfully, I would say no. Most people see IT as a cost center. Even after two decades of talking about the strategic value of IT, it's sad to say that the state of the world is not even close to any of that vision. People still see IT as a huge operative. And so they are really saying, "Look, I have all these costs. I have to spend almost — if I am a large company — five percent to six percent of my revenue on IT. (That obviously varies by industry.) How do I offset some of that expense?
By moving things offshore, the impact is pretty tremendous, very quickly. If you can take 300 people off your payroll and give it to somebody else, [that translates] into a quick bottom-line boost. And the stock market obviously rewards you for that. Immediately. So for the last two years, I would say that has really been the momentum. "What could we do this quarter that would shore up our balance sheet or shore up our bottom line?"
So a lot of companies are saying there is a logical way to do it, and there is a practical way to do it, or there is what you would call, an externally posed rationale for doing it. And more and more people choose externally posed rationale. Well, I guess there are three ways to look at it: There is the practical and logical way to do it, an emotional way to do it, and an externally forced way to do it — competitive or stock market price-driven.
-> Lesson #2: The structure of your outsourcing engagement will depend on your company's size.
You explain a multitude of ways to structure outsourcing arrangements and just a few of them are third party, joint venture, subsidiary, and with those you've got onsite, offsite domestic, offshore, shared services, co-sourcing, offshore developments centers… What structures do you see clients most gravitating to currently?
The big companies, which have had four or five years of experience doing this, are starting to migrate towards this concept of a captive center, an offshore captive center where they are no longer relying on third parties. They are doing it themselves because they have got the experience. So they are increasingly bringing it in house, and they are building their own captive centers there.
The medium-sized guys, who are just getting in it, they are normally talking to a large vender like an InfoSys or Tata Consulting and they are saying, "Let’s build an offshore development center that is dedicated to me. So you run it, but it's all about me." They have dedicated units, dedicated people, but they don't really own the operation.
Now the guys who are even further down the food chain, they normally choose, where they don't have 50 people that they need, they might choose a simple outsourcing arrangement. Let's say you only need 10 people. It makes no sense to have a dedicated center for you. So then they are really saying, "Can you give me 10 people who will work on a project for me?" So they take a more project by project view of the work.
For that mid sized company, what is the advantage of going the route you just described?
There are two types of mid-sized companies. There are companies where they are trying to grow quickly and go to another level — the mid-sized guys that are in the fast growth market. There are mid-sized companies that are in what you call a stable environment, say, manufacturing, where they are really focused on cost and cutting costs. Now the high growth sized firms have more projects than they can handle. So they have to rely on more resources that are cheaper than what they have right now in order to hit their target.
-> Lesson #3: There is no single best way to pay for outsourcing services.
How do those companies pay? Do they get like a monthly bill where the services are detailed?
Usually it's based on a milestone. When you a hit a certain pre-agreed milestone, then the payments get triggered — which is normally the right way to do it, because you don't ever want to pay up front. Or you could do it per head where you are going to have five employees dedicated to you and this is your fully loaded cost per employee.
Are there advantages to going with an InfoSys or a Tata vs. just some off shore provider you find Googling?
The biggest player won't even talk to you if you have less than 15 people, 20 people [working on IT endeavors].
-> Lesson #4: Don't assume that any single outsourcing service provider can tackle it all for you.
The days of the mega outsourcing deals where the service provider bought the IT assets and inherited a huge staff — are those days over?
I don't think so, because IT has become extraordinarily complicated in some of these large companies and I don't think the companies have the stomach to deal with some of these larger problems. So you have to separate IT into software based services and the hardware side.
When you look at the hardware side, the biggest projects we are seeing are all under consolidation projects, where you are taking all these scattered hardware assets and trying to consolidate them and bring them all together so they are going to go from 20 data centers to — let’s say — three data centers. Now that type of project requires a tremendous amount of skill, that only the big providers like IBM and Sun and HP might have. So there is a lot of interest in just outsourcing that part of the business.
On the software side, there is a lot going on with packaged applications, and right now we are in an environment where the packaged applications of the late 1990s are coming up for a major upgrade. So companies are now asking the question, "Do I do this upgrade, of SAP 3.1 to a 4.8 using local resources or do I use an outside resource, an offshore resource?" And that is what is driving people to go offshore. So why do you want to pay $90,000 a year here when maybe you can create a center in India where they are upgrading SAP for $20 per hour? Just makes logical sense.
So packaged applications are the area where the action is going to happen in the next couple of years. Then, of course, there is the services side, which is already happening.
But now you put all the three things together and the company might say, "I really don't want to deal with the headaches of any of these players because they are all interconnected, but I want to give it to an HP or an IBM and let them deal with it." And I think the big deals are still going to happen, and I don't think you will see that momentum slow down anytime soon. But again, once you go past the first thousand companies, the big deals are going to become harder to get.
-> Lesson #5: As hard as it is, be prepared to retrain your IT people in the new ways of working in an outsourced environment or to show them the door.
What is your sense today of the survival rate of IT staff members and managers?
I think we are in the middle of sort of a change of guard, in terms of the skills required — and the skills of the '90s are not the skills of the 2000s. What I mean by that is that in the '90s if you were very good at programming, you were almost assured a steady paycheck.
Now you get to 2000 and your programming art has become a commodity because some of the tools are so good — the Visual Studios for instance — are extraordinarily good tools, so what used to be an artistic endeavor has become more commoditized. And that happens in every industry. You go from art into sort of a commodity war eventually. And in the case of IT maybe that cycle happens a little faster.
Now the question to ask is what happens to the people who refuse to acknowledge that? Who are in the IT department who just refuse to acknowledge that as a fact of life and say, "I still want to stick to what I thought was a valid thing 10 years ago"? And if you are a manager, what do you do? You have no option but to get rid of them. This is the classic change management problem. What do you do with people who just refuse to change? Your only alternative is to get rid of them. If the people are not willing to change, you have to give them a way out…
People still have to do requirements. People still have to manage these projects, people still have to figure out the quality side of it, whether it's is good enough quality to meet whatever the customers is looking for. So what we are starting to see is the emphasis is moving to a lot more management emphasis rather than the doing emphasis. So people have to move from a doer- to a manager-type model. That is kind of tough in the IT department… Managing a team of five programmers in India when you are used to programming that thing yourself — sometimes you're not cut out to do that.
-> Lesson #6: The time it takes to get an outsourcing engagement formulated, signed and transferred is speeding up year by year.
How long are the deals taking? From the time executives decide, "OK, let's mandate outsourcing," to the time the services are actually turned over?
In late 2003 the deals were very, very fast — they were happening in six months. In 2004 a lot of them have been put on idle. I am not kidding! They don't want their names anywhere as the people who are outsourcing America or something like that. So what we have found is a lot of the companies have said, "Let's cool it until the election passes… And this is the scary part. Once the election is over and in January of 2005 this thing starts picking up again, how will that play out? Because 2005 could be the year of offshoring.
You can almost see that in 2002, the speed of the deal was probably, let's say, nine months. In 2003 that shrank on the average to six months. In the beginning of 2004 it shrank further down to four months. In 2004 we are seeing the deal thing go up again where people are sort of saying, "Why don't we wait until the election?" So now we are going to see for the latter part of the year, the speed of the deal will on average go up beyond eight months. Eight or nine months. It is a type of curve we are seeing. It came down and now it is going back up. Now the question is, what will happen after the election? In 2005 all the guys who have been waiting on the sidelines, are they going to rush in at one time? Is there sort of a pent-up demand that is building?
-> Lesson #7: Don't assume the talent you want is in India.
I tell you the thing that worries me the most is Russia. It is not even India that I am worried about; I am worried about Russia.
Because the Russians are really hungry and they are getting to a point where they are seeing the opportunities that the Indians are taking advantage of and they are saying, "Why can't we do it?"
Does Russia have the infrastructure and the other elements in place?
Oh yeah, Russia has a great education system, fantastic programmers. There are two trends we have to watch out for: the open source trend, because if open source really kicks into high gear, and a lot of the great open source programmers are in Estonia and Lithuania, and all the Eastern European countries… If IT embraces open source, then I have a feeling that India might not be the dominant place. That's my observation.
A fair amount of the development and the testing, that could potentially go to places like Romania, Lithuania, Estonia…
While everybody is focused on India, Russia is a dark horse.
How can companies do the due diligence and the research and enough planning and analysis that are required to do it right in just four months?
In the situations with the four-month cycle, half of the influence is already in the company. These [service providers] are sort of using the Trojan horse strategy where they are coming in and doing a project here and a project there. And those are the deals where the CEO has really bought into the whole offshore thing because influences have demonstrated that the whole thing could work and have built a level of credibility. The CEO is saying, "OK, let's do it." So those are the deals that tend to be short. So there is a dating dance that is happening already, sort of behind the scenes, and then the marriage is happening pretty quickly once people figure out the dating role.
-> Lesson #8: Two hidden expenses can hit you hard in calculating the management costs of IT outsourcing.
In your book, you reference reserving 5-10 percent of the total budget for internal management. What does that encompass?
Travel for one… there is a tremendous amount of project overhead that is required, because you need to double-check the work three or four times over. You need to look at every deliverable and crosscheck to make sure the vendor is telling you the truth. Because vendors want to get their payments, so they will try to do it as fast as they can and invariably, quality will suffer. And this is an issue that is going to come up in the U.S…. When employees are asked to get on conference calls at 9 p.m. at night, to talk to a team in India, do they get overtime? Because normally what they call international meeting hours are usually between East coast 11 p.m. and 2 a.m. and then the same thing in the morning. So it changes the rhythm of business. And that is an overhead because people get burned out. If you have a family, imagine trying to do a 10 o'clock call. You need to get the kids to bed.
Have you come across the perfect IT outsourcing project in you research, where everything was just the way it is supposed to be?
It can vary across projects. I have seen companies like GE vary in their whole offshore operations… The level of sophistication within a company tends to vary quite dramatically.
So even the biggest player in this can botch it.
And I have noticed a couple of projects besides GE that are miserable failures. The architectural projects are the ones that have a high degree of failure rate. If you let the vendor design your architecture, you are in big trouble.
It is pretty amazing. You would think that with all the media attention that is being placed on the value of management, the value of flow, that people would learn. But I am finding that people are not learning. They don't learn from other people's failures.