Predictive Dialing & It’s Nuances
Predictive Dialing & It’s Nuances
- June 28, 2020
- Posted By srini@trust.co.in
Definition Of Predictive Dialing
A predictive dialer dials a list of telephone numbers and connects only the answered calls to agents. Predictive dialers use statistical algorithms to minimize the time that agents spend waiting between calls and ringing time on a dialled call, while minimizing the occurrence of someone answering when no agent is available.
Live Scenario
Refer the image above of a 80 seat callcentre with 8 PRI Lines(240 active channels) as its telecom resource. The customer is KLR Global an Airtel outbound callcentre. They start calling from 10:30 and close at 6:30 with 30 minutes lunch for callers. This report was taken @ 4:56 PM with about 6 hrs of active calling on the given day. The data inference is as follows.
| Param | Value | Inference |
| Callcentre Setup | ||
| Agents | 80 | |
| Channels | 240 | Agents to Channel Ratio - 1:3 |
| Starting Time | 10:30 | |
| Closing Time | 18:30 | |
| Abandoned % Set | 20 | This is Campaign Setting of Max Abandoned % Allowed |
| Callcentre Data @ 16:56 | ||
| Dialed | 68150 | |
| Failed | 21917 | |
| Failed % | 32% | This is due to invalid data as well as insufficient channels. To curtail this we need to add another 80 to 90 Channels |
| Attempted | 46252 | |
| Answered | 18083 | |
| Answered % | 39% | This is a reflection of database quality. A good database will deliver between 40 to 50% |
| Abandoned | 3151 | |
| Abandoned % | 17% | As the customer has set 17%, Com1 will ensure that it is between 15 to 20% by ensuring maximum productivity |
| Top 5 Callers Count | 1631 | |
| Top 5 Avg Calls | 326 | With still 90 minutes to go, the productivity expected to cross 400 EOD |
To conclude, to achieve a productivity of 400 calls per agent in a 8 hour shift, you need to have a database with at least 40% call maturity and an agent to telecom channel ratio of minimum of 1:3
Return Of Investment
| Current Expenses | |
| Agent Salary Per Month | 10000 |
| Agent Salary Per Day (@25 Working Days Per Month) | 400 |
| Gsm Gateway Rental Per Channel Per Month | 500 |
| Gsm Gateway Rental Per Channel Per Day (@25 Working Days Per Month) | 20 |
| Manual Mode | |
| Productivity Per Day | 100 |
| Expenses Per Day (Salary) | 400 |
| Expenses Per Day (Gateway) | 0 |
| Expenses Per Day (Total = Salary + Gateway) | 400 |
| Cost Per Call (Expenses / Productivity) | 4.00 |
| Progressive Mode | |
| Productivity Per Day | 250 |
| Expenses Per Day (Salary) | 400 |
| Expenses Per Day (Gateway for 1 Sim) | 20 |
| Expenses Per Day (Total = Salary + Gateway) | 420 |
| Cost Per Call (Expenses / Productivity) | 1.68 |
| Predictive Mode | |
| Productivity Per Day | 400 |
| Expenses Per Day (Salary) | 400 |
| Expenses Per Day (Gateway for 3 Sim) | 60 |
| Expenses Per Day (Total = Salary + Gateway) | 460 |
| Cost Per Call (Expenses / Productivity) | 1.15 |
To conclude, it is clear that predictive mode of dialing achieves 4 times the productivity @ 15% additional investment per month over manual while progressive mode of dialing achieves 2.5 times the productivity @ 5% additional investment per month.
Cost Impact In Reality
| To Process 25000 Calls Per Month | |||
| Per Day Calls To Be Processed (25000/25) | 1000 | ||
| Manual | Progressive | Predective | |
| No Of Agents Needed(Per Month) | 10 | 4 | 2.5 |
| Total Cost Incurred per Month | 1,00,000 | 42,000 | 28,750 |
| To Process 50000 Calls Per Month | |||
| Per Day Calls To Be Processed (50000/25) | 2000 | ||
| Manual | Progressive | Predective | |
| No Of Agents Needed(Per Month) | 20 | 8 | 5 |
| Total Cost Incurred per Month | 2,00,000 | 84,000 | 57,500 |
To conclude, it is better to invest on dialer either on rental or outright to achieve the best per call rate. The difference is too large to ignore and to continue ignoring must have an element of insanity than any logical reasoning.