University of Arkansas System
Type of paper: Thesis/Dissertation Chapter
1. See attached slide on final page
2. There are a number of problems facing Sof-Optics (detailed below).
The most severe problems are in the CSR department:
a. Major problems:
i. Going forward, Sof-Optics will be under pressure to lower prices since B&L (the industry leader) just slashed prices by 25%. This is especially troubling, given the point that follows…
ii. Loss of customers (and sales) due to long customer service wait times or lack of phone lines available (i.e., getting a busy signal). Given that this is a high fixed cost industry, volume is especially important so that Sof-Optics can remain competitive on price and cover their high fixed costs of production. iii. The company had higher than expected start-up costs and, as a result of that, as well as their inability to scale more quickly, they did not show a profit in their first three years of operation.
b. Other problems:
i. Low (but quickly improving) yields on turning “buttons” into lenses – which prevents their variable production costs from being lower.
ii. Cost of returns are high, primarily due to misfits of lenses. A full 16% of standard lenses were returned and 74% of those were due to misfits. iii. High CSR staff turnover due to increasing volume and stress. This is a driver of the high wait times issue noted above.
iv. Customer dissatisfaction with delivery times due to the Postal Service not meeting expectations.
3. There are a number of ways that Sof-Optics could improve the system – specifically the CSR department. These ideas generally fall into four categories: Improve CSR speed, reduce call volume or call requirements, better balance volume with capacity, and other customer experience ideas:
a. Improve CSR speed
i. Specialization: Have CSRs specialize in specific types of calls (for example, a “order taking” rep or “new account” rep). Given the high turnover that is typical of a call center, this specialization will allow new reps to move up the experience curve more quickly and help them to reduce their average call time. All reps should probably still be trained to take standard orders, so that, during high volume periods they can help with those call as well. In order to accomplish this, they should:
1. Use touch tone routing (if available in 1980): When customer calls, rather than going immediately to a rep they should hear a menu of options to choose from (“touch 1 to place an order, touch 2 to check order status”…etc.).
2. If touch tone routing is not available, then use a “tiering” system to answer calls. For 5 out of 6 call types, the customer must provide their name and account number or other basic information. Dedicate 1 or 2 people as “Tier 1 CSRs” to 1) determine the call type and 2) collect this basic info. These Tier 1 reps would then pass the individual on to a “Tier 2” specialist – to take the full order, check order status, or create an account. In this system, the goal would be that when people call they experience no (or very little) wait before talking to a Tier 1 rep. Once they talk to the Tier 1 rep, they would experience any wait before talking to a Tier 2 specialist. Ideally, this system improve speed while simultaneously making the customer feel like they’ve waited less since they talk to a live human being much earlier in the process. b. Reduce call volume and call requirements
i. Reduce call requirements for some CSRs: Although new account creation and status checks are a relatively small fraction of total calls, they take up about 42% of total CSR time. We’ll come back to status checks in the balancing capacity and volume section.
1. New account creation: Since this is a value added activity (you don’t want to tell customers to call back later, maybe they’ll call a competitor!), we need to continue to handle these calls on demand. There may be a way, though, to take down some basic information from the customer quickly (enough to start the order process) and then ask the customer to fax some of the information that is difficult to capture quickly over the phone (e.g., payment info). The DEOs could then enter this info directly from the customer, which would reduce time required for the reps.
ii. Reduce volume: Continue to pursue the consignment program. This will reduce the number of orders coming in, since re-orders tend to occur once per week once consignment is set up. This also reduces the high number of returns which result for misfit.
c. Better balance volume with capacity
i. Push volume to mornings and evenings: Since the worst wait times are in the middle of the day, they should try to push volume to the mornings or evenings in possible. They could do this either voluntary (ask customers to call back during certain time in the future) or involuntarily (restrict hours for certain calls) – but those should only do it involuntarily for non-value added calls (e.g., status checks).
1. Status checks: They could limit the times when they do status checks to 8-10 and 2-5 (or something like that). Using their automated routing system, customers selecting the option to check order status could receive an automated message that says to call back during these times. This would have to be weighed against the reduction is service that customers would experience as a result of the change.
ii. Hire more CSRs and add lines: This, of course, would only be a last resort if all other steps fail, but it’s possible that they simply need more capacity to handle their future volume. Given the low cost of CSRs relative to the cost of losing market share in this early stage of their growth, they should probably be conservative and consider adding lines and people – in other words, they should get ahead of demand, not chase it. Part-time employees would be preferred (to work in the middle of the day).
iii. Improve lunch scheduling: Shift lunch breaks around to better match high volume periods with capacity – especially from 1-1:30 when there are several breaks overlapping.
d. Other customer experience ideas
i. Play music or news while customers are waiting on the phone 4.
a. Capacity of CSR = 27.6 calls per hour
Note: We calculated theoretical capacity using hours worked and estimated length and mix of calls from Exhibit 3. This calculation assumes that any time between calls is negligible. The 27,000 seconds available per rep per day is based on 9 hours of work minus one hour for lunch and two 15 minute paid breaks. We have excluded the 10 min break that reps tend to take in the morning due to high stress, because we assume that could potentially go away in the future if the center is running optimally. The calls per hour capacity is calculated by dividing the number of average seconds available per hour by the weighted average length of a call.