Hybrid Model: Domestic Management and Offshore Operations

Exploiting the price advantages of going offshore with your IT work doesn't always require working directly with a company in another country. A new kind of service provider has emerged that provides a US-based management core to act as a face to Indian-based operations. That's the hybrid model: onshore, offshore, as the project pieces fit together. What's it buy you? Try a closer relationship with your service provider.


Few understand that concept better than Sajan Pillai, COO of US Technology, which provides IT help to Safeway and Costco, as well as customers in financial services, healthcare, manufacturing, utilities and shipping and transportation. US Tech, a subsidiary of Nairobi, Kenya-based conglomerate The Comcraft Group, has offices in Aliso Viejo, CA and Trivandrum, Kerala and Chennai, India, as well as Kuala Lumpur, Malaysia.

In the case of Costco, it even means having US Tech people from India embedded in the Costco development teams and directing work back home from Washington state. (You'll find a link to a Sourcingmag.com case study about Costco at the end of the article.)

In this article, Mr. Pillai shares three pieces of advice for picking the right service provider, explains why people-oriented metrics are just as important as those execution metrics you're used to tracking, and spells out what skills you need to manage your vendors well.

I met somebody this morning who happens to be the CFO for a city back east and she is shopping for a company to do application development work. What kind of advice do you have for somebody like that about making the selection?

Three things.

One, ensure you pick the company with the right values. For long-term deals, check out the company and ask for their values. The attrition rates — not the size or the scale. Make sure you fit this.

Number two is going back to the size and scale. Make sure it is the right size. If you are planning to outsource 100 people, try to do business with people [for whom] 100 people would be a reasonable amount of business. It is very important. Right sizing.

The third thing really is to verify before you plunge into multi-year deals. Build the trust over a period of time. Ensure that you do pilots, even though it might be a little slower. Be sure that you do pilots with the selected vendor. Really understand the variables to integrate, because it is not the technology that matters — because most providers have "technology skills." It is the cultural integration with the organization that makes it successful or not.

Your company says it works at CMMI Level 5. As a service provider at that level, is it hard to work with American companies not at that level?

I have to say, US-based businesses still are the most productive businesses in the world, in terms of innovation and in terms of reaction times. There are some inherent deficiencies — just to be fair. But my observation is that US businesses are much more innovative — more chaotic, but much more innovative.

If you look at the providers in India, you can only do the kind of quality and consistency by processes. It is a must-have, because they lack the subject matter expertise in general. So it's not a choice. In the US, you have a choice. You can be innovative, because you know the business Because you have the choice, you tend to be that way. It is not an either/or story.

I think the key for service providers like us, if you talk to the CIO of Safeway, why do they like US Technology? One is, of course, we have SEI-CMMi Level 5 and P-CMM Level 5 and… and all that. But really, the success comes from taking this rigorous process and not letting it get in the way of business, but figuring out how to translate that process into easily understandable, executable terms.

American business moves very fast. Sometimes processes take a lot of time to fix. Businesses are not going to wait for these processes to fix before they do business. So one of the successful ways to enforce processes is to ensure that there is good education about the merits of doing it. Don't say, "Since we do it, you should do it." That is where a lot of companies fail. But if you start articulating and say, "You guys really need to look at this… Let me take it in smaller chunks," it is a very successful strategy.

You get the buy in from the middle managers to the benefit. It changes very rapidly and they begin to understand… We have been able to help our customers understand the process to a large extent. A lot of customers come undone, "We came to you because of your unique model and your competency and your cost; but what [we] really are getting is your process rate or your discipline, your predictability and your reliability."

…Reverse the equation. Service providers now have to innovate. We have the process rigor. You have to innovate, and you have to show that you can be agile, fast and flexible with the customer. Those guys win.

Even when the contract doesn't call for it? Because contracts can be kind of dumb.

You know, in this business because it is so innovative and so nascent, you really have to go beyond the level of the contract and really understand the spirit of it, because things are changing rapidly. The types of services that were rendered possible two years ago, now it is a completely different ballgame. Since it is changing so much, I also look at customers who have signed n-year deals — kind of doesn't make a lot of sense. I think a lot of customers went that route. So I think what you are going to see is shorter contracts, driving service windows to higher levels of performance that turn up their dial every year as what they expect in outcome. And those guys who can keep up with the customer's expectations will win the game.

Let's just talk about management by metrics, which is a page from your Website. What metrics are you monitoring?

Depends on the domain. It depends on whether you're production support or application development. The whole concept, three or four best practices in the industry. One was, of course, CMMI, CMM, traditional, hardcore, software development metrics. And then, on the production support, we look at standards like ITIL at the operational level. So that performs the fundamental operational metrics for execution.

The innovative thing is, we started introducing people-related metrics — that is, competency improvement, people satisfaction — which turned out to be leading indicators as opposed to lagging indicators of troublesome projects. So we started incorporating people-related metrics beyond just traditional execution metrics.

The important thing is, we have exposed everything we do to the customers. So, whatever we see as executive management, senior management, project management, the customers see. If there is going to be trouble on the project, they are going to know about it. We share the same project management infrastructure.

It has made customers participate in helping us solve problems before it develops into catastrophe.

One of the biggest concerns in "outsourcing" and offshoring is a lack of control. You really can't manage 200, 300, 400 people sitting in California. Some of them come from a different company, different culture. The level of metrics give them the comfort that, yes, it is, indeed, getting done in the right way. People are all comfortable about it and the qualities and control…

It really gives them a sense of control that they wouldn't have otherwise. Not only having metrics, but also helping customers see it on a real time base, however painful that could be. It really helps us improve our "processes" that we thought we though we were pretty darned good at.

And I have to say that a lot of the processes are customer involved-designed. Customers help us design processes, not just standard, so that is one good thing. They may say that is doesn't make any sense. It may make sense to me, but that doesn't make any sense to the business, so we go and change the processes and metrics. It also helps us in ensuring that as the salaries are going up — which is a big issue — the only way to keep up the margins and provide values for the customers is to increase competencies. If the competency improvement does not match the salaries, your business would increase in size, but it won't increase in scale.

So you need to have a good understanding of competency. If you spent x number of hours to get this [number of] function points last year, this year it has to have x percent improvement. It sounds simple, but is very difficult to implement. We spend a lot of time on this internally. So customers know exactly how much time we take. It is not a mystery to them.

How does the customer know that you are not just tucking away the bad news?

Most customers don't have the rigor of asking the right questions, just because they haven't done outsourcing or offshoring. So one [approach] is to teach the customers what questions they should be asking. It is a competitive advantage too. Next time a vendor comes into [the environment], it is very hard for the customer to do business because their level of expectation has been raised.

Second, ensuring that there are no two status reports, no two project management systems. We tell customers to come and visit our facilities all the time. Customers take us up on it, come there, get involved with the project — and we have mechanisms to help customer participation. We have ways of measuring this on a monthly basis. Not just that the project gets done, but it measures a number of things, and they are actively involved in not only helping projects succeed, but also helping us refine our processes. And we share them every six months… We learn together what went wrong and come up with plans every six months on how to improve it… Really it is a big reason for customers to stay with us.

Does the customer know whom they need to retain on their side for a successful relationship or is that something to go in and train them on?

[First,] you have to have very good partnering skills. Most people without the background would think of this as a vendor/management problem. It is not a vender/management issue. That is where the problems happen. It has to be partner/management skills. A big difference between these two.

Secondly, you have to have cross-cultural sensitivity. You have no choice, because it is an international marketplace. You have to get comfortable with traveling globally. You have to communicate your personality. That is very important.

The third thing is, you really have to have the ability to manage a distributed work force. People have to get comfortable in asking for the right metrics, right systems, right processes, so that you get comfortable in accepting large amounts of work with people you have never seen and be comfortable they will come back with exactly what you are thinking. So these are three aspects that IT managers should be vigilant about in making sure the engagements become successful.

What advice do you have for a manager on the client side that isn't getting the expected results?

Change the vender. [Laughs.]

How do they extract themselves from the bad deal or a bad relationship? It may be that the problems are on their side as much as on the provider's side.

Most of the time it happens because somebody ran into a contract situation, which was not performance bound or it was based on some sort of a five-year, 10-year deal, but they really didn't understand the ramifications to the service and service levels they were expecting. So understanding the types of services and expectations when the contract is structured is very important. That is number one.

Number two is to follow the simple common sense principle. One is too few, three is too many — or something like that — where you get a reasonable amount of competition, especially if it is large scale, so that you can really understand best practices. You know, you have a large vender and a medium sized vender, maybe a small vender. Compare and see which one has a better fit.

Number three is pilot, pilot, pilot before you engage. I really think that a lot of companies and customers make mistakes by just getting third parties to vouch for venders without really checking them out. Really, the best way to understand the level of service that you can get from a vender is to pilot them, see if they can really make it work, and make the venders work hard on your business. From a vender perspective, why am I saying it? Because those are the relationships that really blossom.

Once you get a bad deal, what is the way to get out? Change the vender. You don't have [any other] choices…

Useful Links:

US Technology

Case Study: How To Tighten Up Your Partnership: The Costco/US Technology Approach