One look at Ipininder Singh, director-technology, Convergys, and you know little fazes the man. His lithe frame and steely gaze leave little doubt.
But then, maybe all he needed was for the right project to come along, a project his own size. And Convergys had something just in his range.
When Convergys launched operations in India in 2001, it--intentionally or not--created waves in the Indian BPO industry. It was the largest contact center in India at that time with two facilities in Gurgaon.
The Delhi NCR region, especially Gurgaon, was being developed as the hub for BPO operations, thanks to cheap real estate. But while Gurgaon allowed companies to lower cost on the property front, it created other costs. Limited connectivity from Delhi, poor infrastructure, and shifts in the middle of the night forced BPOs to introduce pick-ups and drops for employees. Little did they know that they were creating a Frankenstein.
Today, after employee salaries, transport is the biggest single cost for BPOs, at between 12-15 percent of total operational expenses.
That's a reality that Convergys cannot get away from. Over the last decade, Convergys has grown to six contact centers, with three facilities in Gurgaon and one each in Bangalore, Thane and Pune. With 10,000 employees across India, Convergys has to organize 20,000 pick ups and drops a day, that's over half a million a month.
The exercise gets harder if you try, as Convergys must, to create the most optimal route for its fleet of cabs; one that ensures employees have to travel the shortest distance--you won't want employees to arrive tired--yet one that requires the least number of cabs. Convergys sources over 1,000 cabs all over India from 40 vendors. Add to that the complexity of adjusting to 44 different shift timings and employees with different weekly holidays.
"The entire process of creating the routes, assigning cabs, slotting employees and monitoring the cabs was manual," says Singh "And like every manual process, the system was flawed, had loopholes, and was highly rigid and inaccurate."
But when Singh decided to do something about it, he was told that there was very little he could do. "Every technical solution provider I went to told me that this was how the transport system worked across the industry and that I would have to live with this necessary evil," recalls Singh.
With no ready software to fix the problem, Singh decided to roll up his sleeves. "I told my team that the only way we could clean up this mess was getting our hands dirty," he says.
Singh confesses that the sheer size of the project, with its multiple layers of intricacies, scared him. Breaking it down into smaller pieces helped. "Sometimes, we try to boil the ocean. When things are totally new, one should try to build small automation pieces, tie them together and then build on the innovation," he says.
He did just that and created was a model that the industry would emulate.
Where Does Geeta Go?
Among the most complicated steps in transport management is rostering--the act of creating a plan of who is going where, from which location, at what time, on which day. The rostering problem is so complicated it's classified as NP-Hard. Problems in this category are so complex that it's hard to tell how long it would take a computer to solve it.
At Convergys, they threw Excel sheets and man power at the problem. Every week, supervisors would send a worksheet to the transport department with the names of their team members, their shift timings, addresses, contact numbers, and holidays for the following week. "When we started, we had close to 100 supervisors, all of them sending different formats and versions of Excel. Just sorting some of them was complex enough to qualify as an analytical question in a GMAT exam," says Singh.
The transport team would then sort each employee by name, determine the area they lived in, and attach them to a cab route. The roster took three days to prepare--every week--and once done, it was frozen. That meant you couldn't change plans and that absenteeism lowered the efficiency of the transport fleet. Worse, if an employee changed address or phone number, or quit mid-week, cabs waited at the wrong address, while the Convergys help desk tried to contact an employee--on a wrong number. To fix the problem, Singh figured he needed to get the transport and the HR department on the same page, because HR was kept informed of all changes relating to employees. So with the help of the transport team, he tasked his team to create a HR-IS system and build their online rostering system on that platform.
"We realized that if we could link both systems together, the application could pull out employee details from the HR's records directly," says Singh.
When the team was done, their work cut the time it took for supervisors to fill a sheet, and eliminated the problem of employees who had quit or moved within Convergys.
The system also helped Convergys overcome another huge challenge: billing. Initially, after every trip, cab drivers filled up a trip sheet with the number of kilometers they had run. To decide how much a driver needed to be paid, the transport department needed to enter trip data into their Excel sheets, cross it against the cost of hiring that specific vehicle (Indica, Qualis, etcetera), and add on extra costs like over-time, or 24-hour rates.
"We automated that function and built an application that had all the parameters and metrics pre-loaded into the system. The system could give you costs by department and by employee. It brought down the billing time significantly," says Singh.
More importantly, the application cut roster-making time from three days to one. And allowed the transport to adjust rosters on demand, thereby improving cab utilization.
The Proper Route
The next step was to create the most optimal routes for cabbies. This isn't vital only because it lowers the cost of running cabs, which charge by the kilometer, but also because, in an industry that bills clients by time, it makes incredible sense. If a driver wasn't taking the most optimal route--not necessarily the shortest one--employees clocked in late and that meant money out of Convergys' top line.
According to a 2006 report by Financial Express, between 20 and 25 percent of India's BPO workforce arrive at their office 30 minutes late--everyday. Despite being a dated report, its findings give you an idea of the loss the industry incurs annually from staff delays: A whopping Rs 380 crore.
But how do you plan optimized routes for 10,000 employees?
Singh bought an off-the-shelf routing solution which had 10-12 algorithms to do basic routing. Then he roped in Eicher World Maps to get digital versions of their printed maps. "We got the algorithm to work on those digital maps. Also we got all employees to geo-code their locations," he says.
In addition to layering the maps and the algorithms, Singh's team added data including the availability of vehicles, cab capacities, and average speeds on various roads, etcetera. This, now, allows Convergys to plan ahead and create the most optimal route.
Importantly, it also allows the transport department to work in a more real-time fashion, increasing its efficiency. "Now when an employee's pick or drop is cancelled, we can reshuffle routes and fill in the empty space in a cab," says Singh. "This allows us to reduce the number of cabs by about 20 percent."
"The transport solution helped us reduce and optimize costs and provides a positive experience for our employees, something which is of so much importance to us. It enables us to manage the entire transport service chain from rostering, routing to billing, thereby creating a seamless end-user and vendor experience," says Gaurav Kalra, head of transport, Convergys.
Overall, Singh says that a modest estimate of how much the project lowered transport costs is 30 to 40 percent.
The Road Ahead
Revved up by their success, Singh's team began exploring ways to make the system more user-friendly. "Attrition and constant hiring means that the system has to be intuitive enough to not require any training at all," says Singh. So they introduced a host of self-service portals to reduce the burden on the transport team. "We have also started the concept of SMS-based workflow. By sending short codes such as "CP" for cancel pick up and "CD" for cancel drop, employees could notify the transport team without having to talk to the call center or help desk," says Singh.
"The high level of automation has set a precedent in the industry," says Kalra.
By building self-help portals where staffer can check their routes and pick-up time, Singh has managed to considerably lower dependence on the helpdesk.
Singh also experimented with barcodes. Earlier, a guard standing at the gate kept a manual record of when cabs arrived. This data was maintained just in case an employee came in late and wrongfully blamed a late cab.
Now, by assigning a barcode strip to each cab and trip sheet, Singh's made it possible for the guard to scan both with a handheld scanner. Also, the arrival of specific cabs is flashed on the respective supervisor's screen.
Singh is also exploring technologies like GPS to track cabs in real-time and, if possible, divert them from areas of heavy traffic. It will also allow them to send SMSes to employees, asking them to get ready when a cab reaches within a certain distance of a pickup point.
"There's so much to be done," says Singh.
