Why predictive dialers are the best choice for your call center
What is predictive dialing?
You might have already heard of predictive dialers, but not been sure what they were or how they differed from other automated calling services. The easiest way to understand any piece of technology is to know what problem it solves, so let’s consider the use-case for predictive dialers.
Let’s imagine an average sales representative in a call center makes around fifty calls a day. Let’s say, being generous, that before each call there’s twenty dial tones that each last five seconds. If you work out the maths, you’re spending over a minute waiting for every call, and that’s before you even know if anyone’s going to pick up.
This is the problem that predictive dialing solves.
Predictive dialing is a system that automatically dials phone numbers so that the line connects just as an agent in the call center becomes available. It goes down a list of numbers, much like an autodialer or a robodialer, then passes each call to an agent when its internal algorithm decides it’s the best time to do so.
Even better, it can filter out low-quality calls, such as answering machines or busy tones, leaving your agents only with high-quality leads.
Why is predictive dialing useful?
Predictive dialers remove the downtime between calls for your agents.
It directly follows from that that it increases agent efficiency by letting them call many more leads than they would otherwise in the same timespan. Multiply that boost in productivity by every agent in the center, and even small optimizations – five seconds here, three seconds there – can lead to hundreds of extra calls in a short amount of time.
Let’s consider some of the things that can waste your agents’ time.
- Dialing the phone numbers
- Manually dispositioning calls
- Listening to voicemails, busy tones and no-answers
- Leaving voicemails
- Scheduling call backs
All of these are either reduced or eliminated entirely by the proper use of a predictive dialer. As such, it’s no wonder that reports almost unanimously describe them as being major factors in increased productivity for call centers across the world.
Another major benefit to predictive dialers is that they commonly allow for integration with other software tools in your calling solution.
This is most common with Customer Relation Management (CRM) tools, as this lets you transfer analytics data from a call – client’s response, favourability, openness to future calls – directly into your customer database without taking your call agents away from the phones.
Who uses predictive dialers?
The history of predictive dialers is long and complex, stretching back to the 1980s. They were invented by a technician named Douglas A. Samuelson and saw their first use in banks, who found them useful mainly for debt collection.
Since then, of course, their benefits have become widely-known, and their usage has exploded accordingly. Nowadays, predictive dialers are routinely used in customer service, market research, telemarketing and more.
The first predictive dialers were hardware – physical devices attached to phones.
In the modern digital age, of course, software has replaced these old versions, and most predictive dialers live in the cloud. These hosted versions are by far the most popular option nowadays because they reduce the amount of capital needed up-front for installation, as well as offloading maintenance costs to the cloud service provider.
How are they any different from autodialers?
If you’re new to predictive dialers, you might be wondering what all the fuss is about.
“Aren’t these just robo dialers?”
And yes, in many ways, they’re similar. Both automate the process of calling large lists of numbers.
The big difference lies in when the dialer chooses to make each call. Predictive dialers use the information and metrics you feed it to predict when one of your agents will be ready to take the next call. This lets it neatly snip out all the downtime between calls, letting agents jump seamlessly from one to the next.
And if there aren’t any agents available? Predictive dialers are smart enough to either slow down its outreach or freeze it entirely until it can tell an agent is near the end of their call.
The other big thing that sets predictive dialers apart from robocallers is that they don’t just go down a list line by line. They actually dial multiple numbers at once – this is so that if you have five agents sitting free, it can spin up five calls for them all at once.
What does a predictive dialer need to be effective?
Sadly, computers aren’t yet good enough to see the future. In order to predict when to start a call, the dialer information, and the more you give it, the better a judgement it can make. Let’s breakdown which kinds of information it needs.
Pacing algorithm
The pacing algorithm is the mathematical formula the dialer uses to decide how many calls it should try to place, based on the number of agents available.
This is the fundamental factor that decides how effective the dialer ultimately is. It’s how it decides what balance to try and strike between the number of free agents you have available and the number of outbound calls to place.
Abandoned call rate
When a customer tries to call a support center, you only have a few precious seconds before they lose their patience and hang up. Each one of these abandoned calls is a failure of your dialing setup, and it’s something your predictive dialer needs to know about.
One of the primary goals of any calling campaign should be to reduce the number of abandoned calls, and as a result, idle time among agents.
By feeding the abandoned call rate back into the dialer, it can make better predictions about the right time to route calls through to your agents, reducing the chance of them happening.
Campaign size
In general, predictive dialers work better the larger your calling operations are.
This is simply due to the fact that bigger call centers produce more statistical data that you can feed back into the prediction algorithm. Data from five calls can only get you a very crude and rudimentary prediction compared to data from five hundred, or five thousand.
Agent awareness
There are two key factors for a dialer: active phone lines and active agents. It can manage the phone lines by itself, but it needs you to keep it updated with how many call agents are free at any one time.
This information needs to be precise – remember that these dialers work down to fractions of a second. Once again, this is an area where the more agents you have, the more sophisticated the dialer’s guesses can become.
What other features do they have?
Besides the core feature of cutting downtime for your agents, most predictive dialers these days come with a list of impressive features that can make your business even more efficient. However, it’s not always clear what these features actually do just by reading the names. Let’s break them down.
DNC (Do Not Call) management
A DNC list is, essentially, a list of numbers which your company isn’t allowed to call for one reason or another. This is most commonly a national registry which people can opt-in to to signal that they don’t want to receive advertising or marketing calls.
If you do end up calling someone on this list, you may be held liable for hefty fines. Thankfully, predictive dialers are able to filter DNC numbers out of their lists without a second thought, meaning that they will never even reach your agents.
Customized dial times
Depending on the size or location of your business, it’s possible that the list of numbers you want to call may include people living in different timezones.
If the difference is only one or two hours, that’s not a problem. But if your automated calls wake someone up at 2AM, it’s not very likely that they’ll form a positive impression of your company.
While you could manually filter out numbers for people in bad time zones, predictive dialers can do this for you. This saves effort and also lets you plan out when you want your calls to hit – early in the morning? After lunch? The choice is yours.
Answering machine detection (AMD)
This is another feature that automatically filters out low-quality calls, leaving your agents to deal only with the most valuable leads available.
As the name suggests, this feature analyzes whether or not the dialed number is being answered by a human or an answering machine. If it’s the latter, you can usually configure the dialer to either leave a message for the homeowner or simply hang up.
Call-back schedulers
Call-backs are a powerful way to keep customers engaged with your business, build familiarity and construct engaging consumer journeys.
Unfortunately, they’re also easy to do poorly.
If an agent makes a promise to call a lead back within half an hour, it’s all too common for them to get caught up in another call and forget their promise. Being stood up like this is an easy way to frustrate customers and make them question the competence of your agents.
Thankfully, if there’s one thing computers are good at, it’s reminding us to do things at specific times. Predictive dialers let you schedule follow-up calls in advance and will provide pop-up reminders to the relevant agents when it’s time, making sure they don’t forget.
Are there any potential drawbacks?
No technology is perfect. If you can use a tool, you can also abuse it – so it’s important to know the potential pitfalls of using predictive dialers.
They’re not right for every company
Predictive dialers try to cut downtime as much as possible – that’s their only purpose in life. This is often exactly what you want in a calling campaign, but not always.
Consider an organization that’s selling high-value items. They might only make four or five sales a day, but those sales are enough to put the company in the black. For a situation like that, you want downtime between calls so that your agents have time to research their next lead and plan a strategy for the call.
In other words, sometimes understanding the context behind a lead is the key to a productive conversation. Because predictive dialers remove downtime between calls, they make it much harder – though not impossible – to formulate detailed strategies for specific leads, as opposed to campaign-wide plans.
They can be wrong
It’s worth remembering that even though predictive dialers work on the basis of very advanced algorithms, they’re also dealing with time spans of less than a second. As such, they don’t have a 100% accuracy rate, and can activate calls before there are any agents available to take them.
They are legal aspects to consider
Predictive callers are fully legal.
There is, however, one big restriction with how you can use them. Since 1991, the Telephone Consumer Protection Act (TCPA) has made it illegal to use any auto-dialer to contact numbers that haven’t given explicit consent beforehand.
There are exceptions for things like emergency calls and purely informational calls, but this is an important restriction to be aware of before planning a calling campaign.
They work best at scale
Predictive dialers are a wonderful piece of technology, but they don’t work for free. Installing a dialer and training your agents how to use it properly will always incur some lost time and money. This means that the dialer has to put in extra work to pay off this debt before it can start actually making you a profit.
Unfortunately, it’s hard for this to happen in smaller companies. If you only have a call center with six agents, you’re very likely not dealing with the amount of traffic where the seconds saved by a predictive dialer will get you many extra calls.
There’s no set number for when predictive dialers become a no-brainer. Even if you set a rule of thumb at around fifty or more employees, all constantly placing calls, the dialer might only sometimes pay for itself in a reasonable timeframe.
The best course of action is to conduct internal research to find out what your company’s own margins would be. Of course, if you expect to grow – and what company doesn’t? – then pre-emptively installing a predictive dialer is an excellent way to make your business scale better with demand.