Could a Focus on Getting Every Call Right have the Wide-ranging Benefits for Call Centers that JIT had in Manufacturing

This is a post about how the commitment to one simple idea can launch a thousand ships of changes and a virtuous cycle of profit improvements. After reviewing the far-reaching effects Just-in-Time, a key element of the Toyota Production System, had in manufacturing, I will argue that a focus on getting calls right in call centers could also have a wide-ranging effect on organizational performance and profitability.

The Toyota Production System (TPS) really is The Machine that Changed the World (see the book of the same title by Womack, Jones and Roos that has sold over 600,000 copies). TPS has changed how every single product is manufactured in every part of the world. One key pillar of the TPS system is Just-in-Time.

Just-in-Time is a production strategy that seeks to improve a business’ Return on Invested Capital by reducing the carrying costs associated with work-in-process (WIP) inventory. It is a pretty simple idea: There is a lot of capital tied up in WIP inventory. If we can reduce it, our costs will go down and ROIC will improve. If we have less inventory, when market demand changes, there aren’t as many write-offs. If we don’t have as much WIP, we don’t need as much space. As big as all of these savings are, they are the tip of the iceberg in terms of the gains wrought by JIT.

What is interesting is that the commitment to that simple JIT idea resulted in a virtuous cycle of changes that produced financial benefits way beyond the reductions in working capital. For example, when you take away the safety stock, all the production abnormalities such as machine reliability, excessive changeover times, and production bottlenecks, start to rear their ugly heads. So implementing JIT unveils production issues that management may not have even been aware of and forces them to improve those inefficiencies to avoid production outages.

Similarly, with only enough parts on hand for the production scheduled for that day, assemblers no longer had a choice of which part to use. Every part available had to work perfectly. If it didn’t, the line would shut down. Just as the commitment to JIT unveiled production problems, it also created a quality assurance crisis. The result was a continuous improvement in the quality of parts that extended back through the entire supply chain.

Further, the improvements in the quality of parts were not always enough. Many companies were also forced to redesign the product to widen tolerances. This had the effect of further improving the quality of finished goods.

The biggest effects of JIT may have been the impact on the customer. The factories were increasingly able to predictably deliver goods on the promise date. They were also more responsive. You no longer had to, in the case of cars, “take what’s on the lot.” You could choose from a blizzard of options and get the exact car you wanted with remarkably little wait time. The quality improved, the choices improved, the responsiveness improved, and guess what? So did customer satisfaction, loyalty, and revenue.

The commitment to this simple idea of Just-in-Time drove dramatic improvements through an organization. Could a commitment to getting every call that comes into a call center correct have a similar, virtuous cycle of effects? I think it can.

Let’s conduct a simple thought experiment augmented by real results. What would happen if we defined what a correct call was, by call type, and our combination of people and technology guaranteed that almost every call would be exactly right?

For this thought experiment, we are going to start with a typical, say financial services, call center. Customers call for lots of reasons…lost credit card, transaction inquiry, credit limit increases, etc. We have some calls well documented, some not so much. We have some good agents, others not so good. Agents make mistakes sometimes: they skip steps or don’t follow the most updated procedure. Some are crisp, others take a lot of time and there are a couple minutes of after call work, where the agent has to clean up the call and, ahem, take a little break.

Then we will assume that we make some changes. We define what a correct call is for every single call type…what the agent must do in their system and say to customers to correctly resolve their issues. And we will assume we have figured out a relatively efficient way to handle that request that minimizes handle time and after call work.

To ensure the agents follow the correct process, efficiently, we will deploy agent-assisted automation. Agent-assisted automation involves the use of pre-programmed system actions and pre-recorded audio files that are directed by the agent. The agent is live on the call, following the pre-defined path as appropriate, but intervening with his/her live voice whenever what is happening on the call requires it.

What follows is a list of benefits that would accrue if we got the calls right. As you go through the list, actual results we have achieved using our automation and the degree of benefit we have seen are shown parenthetically.

1. If we are using some automation, there is less to teach the agents. If there is less to teach the agents we could reduce the lead time we need for hiring agents in advance of when they need to be on the phones. (Our client reduced training time by 30%.)
2. If the automation is ensuring the calls are correct, we do not have to pull the agents off the phone for training/coaching as often which means we don’t need as many agents to hit service levels. (Call volume is up due to the growth of the product, but the number of agents has been reduced.)
3. If we know the calls will be right, we don’t need to do as much monitoring, which means we need less monitors.
4. If the call is right, we will get less repeat calls, which means we will have less volume and need less agents. (same as #2)
5. Because we have engineered the calls, they are shorter and there is less After Call Work. This reduces the number of agents we need. (We have reduced AHT by 40% and ACW by 90%.)
6. If the automation is handling the boring and fatiguing parts of the job, the agents may be happier which will delay turnover. No one stays in the job forever, but delaying turnover is the same as reducing it. (We have dramatically improved agent satisfaction, but we do not have long term data on the effect on turnover.)
7. If we have less agents and less turnover, it means we
   • don’t need as much HR staff to hire, fire, and administer
   • don’t need as many managers/coaches
   • don’t need as many seat licenses for software and hardware
   • don’t need as much space
8. When agents make mistakes costs often occur outside the call center. There can be big groups handling improper warranty returns, calls get escalated to supervisors, letters get written to the CEO, legal gets involved to handle mistakes made by agents, fines have to be paid, etc. If the calls are handled correctly, all of this work goes down and may result in the need for less staff. (We have reduced legal fees and accent escalations.)
9. Often times the steps agents skip or don’t do correctly are cross-sells. If these are done every time, revenue goes up because the agent makes the right offer at the right time, every time. (We have increased cross-sell by 5X over agents not using the automation.)
10. Usually when calls are streamlined, it means less AHT, which is important. But when the calls you are streamlining are collections calls, handling more calls per day means collecting more revenue. (Record revenues have been achieved by agencies using our solution…up 25% over target)

In manufacturing, all the WIP lying around was pure waste that covered a lot of other problems. Similarly, there is a lot of waste in and around call centers that is a direct result of the inefficiency and ineffectiveness of call center agents.

I am not blaming the agents. Call handling work is poorly designed, the job is fatiguing, the pay is low, turnover is high, and humans make mistakes. The thought that sporadic coaching could overcome all of these problems and provide a way out of the wilderness is laughable. But combining the accuracy and reliability of software automation with the warmth and flexibility of human interaction has been shown to dramatically increase the percentage of correct calls.

I am guessing when companies started with JIT, no one imagined that a simple commitment to reduce the waste associated with excess inventory would have the far reaching effects that it had.

Could a simple focus on getting calls right hold the same potential?