Outsourcing Offshore — The Inside Story
By Narendra Dev
November 7, 2002
Outsourcing has proven to be an effective strategy for many enterprises. In some cases, however, going a step further and taking an international approach can deliver even greater benefits.
Moving a contact center offshore, for example, can result in significant advantages such as cost reduction, coverage after hours and availability of technical labor. The key to reaping those benefits is setting clear objectives. In the example below, you'll see how a Fortune 100 company moved part of its contact center offshore and discover what it learned along the way.
Pick a Country
The countries normally viewed as attractive for offshore work are India, Phillipines, Israel and Ireland. The reason: It is easy to get people with English-speaking skills in these four countries. With time-zone differences with Pacific Standard Time ranging from eight to 16 hours, these countries are well-located for after hours coverage. Also, it is possible to achieve a 30-60 percent cost reduction aggregated over labor, telecomm, travel and training costs.
This Fortune 100 company had been experimenting with offshore support for more than a year, using two offshore partners in India. This decision to use two partners in India was made by the previous consultant managing this operation. The rationale was to split the business and achieve lower cost by having the partners compete.
The cost reduction achieved was significant (more than 50 percent) so it made sense to continue the offshore support from India. There were no service level agreements (SLAs) negotiated as part of this outsourcing arrangement. This had been overlooked as part of the original agreement. It became obvious that there was a lack of clear accountability with this arrangement and it needed to be addressed.
Beginning on July 2001, we outsourced three support activities to offshore companies — database administration (DBA) support for backoffice applications, Unix and NT system administrator (sysadmin) support and front-line (level 1) customer support. Coverage hours for the offshore partners were from 6 p.m. to 5 a.m. Pacific time and weekends.
After evaluating the results from the experimentation, we were pleased with the cost reduction and hours-of-coverage benefits that resulted from the offshore support and decided to formalize the arrangement. Here are some of our observations, based on an assessment of the first year:
The volume of the DBA support business — 17 DBA FTEs (Full-Time engineer Equivalents) was not enough to justify splitting the business. We were expecting 13x5 plus weekend coverage from two offshore partners, each having eight or nine FTEs. This was an unreasonable expectation since they were both getting spread too thin. We experienced situations when the partners were out of capacity and work had to be delayed.
We had no control over the partner's staffing and no visibility into how many engineers were actually deployed offshore. Sometimes activities would not take place on schedule because the partner was out of capacity. All the DBAs and sysadmins were using a single log-on so there was no traceability. They were working on our business and other businesses, too. It became increasingly apparent to us that either we needed an SLA or we needed better control over the partner.
There were significant issues transacting calls between the partners. They were located in different cities and the truck dialing facilities between cities in India (in this case, Chennai and Hyderabad) were worse than calling between the U.S. and India.
The systems infrastructure was a barrier to doing business. For security reasons, we were not able to implement the client version of the customized Vantive customer relationship management (CRM) system at the partner site. They were able to use only the Web version that customers' use, so they had to keep multiple windows open on each agents workstation for each of the customers supported.
The partners were outside the company's firewall, so we could not give them root password access. As a result, they had very limited sysadmin capability from offshore. We discontinued this sysadmin support once we discovered that it was not working.
Significant issues arose with the front-line customer support. There were problems with customers understanding the accent — the agents spoke too fast and they had a different phone etiquette than what U.S. customers expect. The partner did not want to hire two skill pools, so it used its DBAs as front-line customer support analysts even though these people had not been hired for their superior soft skills.
Based on these observations, we took the following actions, which we implemented over the course of three months:
Single-sourced the business to one partner to get a better deal. This would also make the hours of coverage issue easier since all the staffing could be applied at one location.
Signed an SLA with our Indian subsidiary. We had our Indian subsidiary sign an agreement with the partner for a defined level of headcount (17 FTEs, to be exact, that were selected after an interview process).
Increased the coverage period from 13x5 to 24x7. We started a U.S. day shift to cover 24x7 processing of proactive alerts and to schedule with U.S. customers their updates, patches and fixes to be applied after hours.
Located the headcount in the local office of our company in India. This way, we had a clear view of the resources working on our account and their skillsets. Also, working inside the company infrastructure and firewall made security issues easier to deal with. This was reflected back in the partner price since we bore the infrastructure costs; their costs were basically only labor.
Moved the front-line level 1 support 24x7 back to the U.S. There were other business reasons for doing this besides customer dissatisfaction. We had some customer dissatisfaction on this decision since we moved from direct-connect to callbacks. But overall it was a positive move.
Simultaneously, we moved from our Vantive CRM system to a customized Remedy CRM system to mesh with the rest of the company business. This cutover was done in a phased manner by end-customers, over a period of one month. We migrated only the open cases, not the history. It went very smoothly. The only issue was that the customers liked the Vantive Web interface more than the Remedy Web interface, so there was some grumbling during this transition.
Made a joint commitment along with the partner on training, agreeing on a percentage split of costs. This covered soft skills and backoffice applications. The soft-skills training covered basic phone etiquette, accent modification, setting expectations and making commitments. This was distributed over a period of two months to allow the administrators to practice and internalize the skills learned. This soft-skills training is still in progress, so it is too early to evaluate if it met the objectives.
The application training covered the data architecture, application support tools and how to do upgrades, patches and fixes. This training, too, is in progress.
Reaping the Financial Rewards
There were significant savings (more than 50 percent) in moving activities offshore. We pay hourly rates of $15-20 for DBA resources, $10-15 for sysadmin resources and $7-10 for customer support analyst resources. The corresponding onsite rates are $80-90, $60-70 and $25-40, respectively. There were some increases in infrastucture costs to accommodate the 17 people, but the labor cost savings were much more significant.
We developed a project plan to implement this new support arrangement over a two-month period.
This new support model has been operational for about three months. Here were some of the issues we still face moving forward and our actions to address them.
Once we single-sourced the business, the partner selected had to add about 50 percent additional labor. The new DBAs were not experienced on the application, so there was some degradation in quality while the new people came up to speed. We spread the experienced people out across the shifts to ensure that the shift leads were experienced. This helped but there were still a few quality issues. We are also investing in application training.
The additional load stressed the local network at the Hyderabad office, which had been designed for a maximum capacity of 50 people using e-mail and personal productivity applications. It was not designed for operations support with 17 additional people putting heavy load on the network. We upgraded the local loop network from copper to fiber, which improved the reliability and bandwidth. We also added an ISDN backup network to Bangalore in case the fiber failed. There was still a single point of failure on the Bangalore to Singapore undersea link (and we did have one failure on this link). We are in the process of developing and implementing a disaster recovery plan to address this scenario.
Having the DBAs co-locate at the company site made the infrastructure issues easy to deal with. We were able to leverage the existing phone exchange, company e-mail, CRM system and company time-vouchering system for the 17 additional people. We also leveraged the standard process in place to setup a contract employee. This gave the partner's support personnel access to all the relevant systems, e-mail, time vouchering and logins to customer systems.
The bulk of the learning has been on process issues. We have a daily meeting from 8:30-9 a.m. Pacific time to go over issues with the offshore operation from the previous night and shift-transfer issues. We also discuss recommendations from incident reports, analyzing failures in delivery for root cause.
Unless there is strong management commitment to make the offshore arrangement work, it is easy to blame the offshore partners for all the problems. They are usually absent from management meetings (because of the time difference) and tend to get dumped on. As the champion for offshore support, I had to represent their case at management meetings and correct these perceptions if it was unwarranted.
Successful Outsourcing Requires a Commitment
Outsourcing contact centers offshore is a one- to two-year process of experimentation. It is often not obvious which activities will be successful. You need to keep an open mind, listen to the customers' feedback and make adjustments. Most important, there needs to be management clarity on the objectives and a commitment to address the problems as they are encountered.
Do you have a comment or question about this article or the ASP industry in general? Speak out in the ASP Discussion Forum.
Narendra Dev is an independent consultant. He has worked in operations since 1975, the last 16 years in a senior management capacity. His areas of focus are customer service, IT, hosting and network operations in enterprise companies and startups. He has an MBA & MS Physics. He would also like to thank Jacky Hood for helping to prepare this article.
IE 7
Firefox 2.0