If your goal is to keep your call center as busy as possible, and you have plenty of leads, lines and agents, then predictive dialing may be right for you. Many people confuse predictive dialing with power dialing, but they are different. Power dialing is usually the straightforward dialing of X number of calls for every agent currently waiting for a call. It’s an unsophisticated approach that’s easy to implement, so it is widely available.
Predictive dialing is power dialing evolved. Predictive dialing software automatically adjusts the dial rate live, based on a number of factors and your preconfigured settings. This means you don’t have to keep someone constantly tweaking the dial rate to keep in compliance or to ride through a slow patch. Instead the system automatically adjusts for you.
You can maximize efficiency for your agents with an effective predictive dialer. Usually the agent is just returned to the outbound queue after dispositioning their previous call. The addition of the agent is noted, and one or more new dials may be made to keep the connections flowing fast enough. When a call is connected, it goes straight to a waiting agent. A good predictive dialer will allow you to enter a number of settings for each campaign, such as:
- Hard abandon rate maximum – Usually set as the maximum abandon rate allowed by regulations.
- Soft abandon rate maximum – You don’t want to dial as fast as possible, then hit a wall when you hit the real maximum. Setting a soft abandon rate allows you to say that if the abandon rate gets this high, let’s slow down a bit so we don’t hit the maximum.
- Acceptable idle time – while everybody would prefer this to be as small as possible, you do have to be a bit realistic. If agents are waiting 10 seconds between calls, things are going pretty well. No need to panic. Realistically, much less than this and agents are going to dawdle on the wrapup to give their voice a quick break. If the acceptable idle time is exceeded, the dial rate will be increased for each waiting agent.
- Maximum dial time – the number of seconds the phone will be allowed to ring before giving up. Too long, and you get a lot of answering machines. Too short, and nobody has a chance to get to the phone.
Other things the predictive dialer h
as to keep track of do do its job properly are:
- Connect rate – sometimes if you have high quality leads, you could get a 50% connect rate, ie. for every two calls you dial, one answers. Other times you might get a connect rate of 2%. If your leads are poor quality or its a bad time in the local calling area (ie. after 6pm for a B2B campaign), you might have to dial a lot of leads to get a connection.
- Abandon rate – an abandon is when a caller answers the phone, but there is no agent available within the maximum wait period. In North America, this wait time is 2 seconds.
- Age of current dials – if there are 3 agents waiting, the connect rate is 50%, then the system might have 5 or 6 calls that are ringing. If 3 of those are currently 16 seconds old and the maximum dial time is 18 seconds, the dialer shouldn’t count those 3 calls as 3 full calls, as they are less likely to connect to someone.
Every method of dialing has advantages and disadvantages. These are magnified in predictive dialing. The primary advantages are:
- Keeping agents talking – predictive dialing is designed to connect your agents as quickly as possible. Since a number of calls are being dialed on your agent’s behalf, they don’t have to sit listening to a ringing line. You also don’t have to depend on them hanging up quickly on a busy, or wrapping up and dispositioning calls that never connected. That is sometimes the biggest time saving for agents.
- Less mis-labelling calls – if the call is busy, or no answer, or an answering machine, and never goes to an agent, it’s automatically marked correctly by the system.
- Answering machine detection – you can use answering machine detection (AMD) with predictive dialing. If you’ve got an effective method of doing AMD, you can save a lot of time by not giving those calls to agents. You can also configure the system to leave a message on the answering machine if you wish, without the agent.
With so many powerful advantages, predictive dialing looks pretty good. Predictive dialing has some limitations, however:
- Short dial times – in order to miss answering machines and keep lines free for dialing, short dial timeouts are the norm. There can be a wide difference in connect rates between an 18 second timeout and a 30 second timeout, depending on your campaign.
- Answering machine detection – if your AMD isn’t very accurate, you may be abandoning calls that could have gone to an agent. You also have to consider this when determining if you’re in compliance with regulations. Not connecting a person to an agent just because the system detected the person as an answering machine doesn’t let you bypass the abandoned call regulations.
- Tying up lots of lines – if you’re dialing at a 4:1 pace, then 10 waiting agents require 40 outgoing lines. That can add up.
- Regulations – there is a lot of regulatory oversight on automated dialing methods, including predictive dialing. Abandoned calls are a big thing, with the US government mandating no more than a 3% abandon rate, a time limit of 2 seconds to connect a call to an agent before abandoning, and the requirement to play a message with certain information if the call is abandoned. Some jurisdictions limit the number of lines used to auto dial unless other regulatory hurdles are met. Some jurisdictions prohibit auto dialing cellphones. It can be daunting.
In short, predictive dialing is a powerful tool for turning leads into contacts. Used in the right way, it can boost profitability tremendously.