.svg)
.svg)
.svg)
1. The detector is guessing
Every parallel dialer depends on something called answering machine detection, or AMD. It’s an algorithm that listens to the first moments of each call and tries to decide whether a real person, a voicemail, or a phone tree picked up. The trouble is that some voicemail greetings sound a lot like a person, and some people sound a lot like a voicemail greeting. When the algorithm guesses wrong, either your rep gets patched into a voicemail or a real person gets hung up on.
That guessing also takes time, and that’s where the pause comes from. The prospect says “Hello?” and hears nothing. They say it again. By the time your rep finally comes on the line, the prospect already knows exactly what kind of call this is.
Many parallel dialers use the same detection in their single-line mode too, so they can skip voicemails faster. If yours does, turn it off.
2. Two people answer
Say the dialer calls 5 numbers and 2 people pick up at the same moment. It can only connect your rep to one of them, so it hangs up on the other. That second person was reachable. They took the time to answer, and what they got was dead air and a dial tone.
Parallel dialer vendors will tell you they show you who that second person was, so you can call them back. Some will also tell you it doesn’t matter, because the prospect never knew who was calling. Neither of those is a good reason to hang up on a prospect who took the time to answer your call.
This isn’t even just about first impressions, though. Timing matters for another reason: it’s why parallel dialer vendors compete on how quickly they connect a rep once someone answers, often advertising figures well under a second. US telemarketing regulations use two seconds after the person finishes their greeting as the benchmark for when a call counts as abandoned, so every vendor has a reason to keep that number low.
If you use a parallel dialer today, these are fair questions to ask your vendor. When you ask, insist they give you answers from your own call data, not their marketing-approved claims number or an average of all customers:
- How often does more than one person answer at the same time?
- What happens to the people the dialer hangs up on, and how often do they answer when you call them back?
- On your team’s actual calls, how long is the gap between a prospect saying hello and your rep’s first word?
Of course, you’ll never have to ask those questions about a power dialer, because it only calls one person at a time.
3. The carriers are watching
Before a prospect’s phone ever rings, the major carriers run your call through analytics engines that score your number and decide whether to label it as spam. Those engines look for specific patterns: bursts of calls from one number, lots of very short calls, and low answer rates. That’s almost a word-for-word description of how a parallel dialer works.
We saw it in the call logs. Reps who parallel dialed placed 27.38% of their dials from spam-flagged numbers. Reps who power dialed or used a single line placed 1.47% of theirs from flagged numbers. That’s about 19 times the exposure.
We could also trace how it happens, one step at a time, using correlation scores. A correlation score shows how closely two things are related, and it can be anywhere between −1 and 1. A score of 0 means no relationship, a positive score means both go up together, and a negative score means one goes down as the other goes up.
The more dials a rep makes, the more phone numbers they burn through, with a correlation of 0.58 between dials and numbers used. The more numbers a team uses, the more it ends up sharing pools of numbers, and shared pools get flagged as spam more often. That correlation is weak, at 0.12, but it showed up consistently. Flagged numbers, in turn, connect less often: the correlation between a number being flagged as spam and its connect rate was −0.25.
Across the 34 teams with at least 5,000 dials where we could see which phone number each call came from, 27.4% of all dials went out on spam-flagged numbers. Nine of those teams sent more than 40% of their dials from flagged numbers, and the worst team placed a whopping 84.7% of their calls on numbers that had virtually no chance of connecting because they were already showing up as “spam likely.”
It is possible to make a lot of calls without getting spam-flagged, though. A few teams stayed under 5% of dials on spam-flagged numbers while still dialing with serious volume, and one team had no flagged dials at all. We controlled for timing and market too, so the differences came down entirely to how they dialed: in parallel or one line at a time.
Now for the part that surprised us most.
Remember the first test, where we removed every flagged number from both sides and parallel dialing still got only half the connect rate? That means spam flags are one of several ways parallel dialing hurts you.
But there are other ways parallel dialing is worse.
4. Your rep isn’t ready
With a power dialer, your rep knows who they’re calling before the phone starts ringing. They can see the person’s name, company, and title, and they have a few seconds to get their opening line ready.
With a parallel dialer, the rep finds out who answered at the same moment the prospect is hearing silence. It could be any one of five different people at five different companies. So the first thing out of the rep’s mouth is usually a scramble.
That’s the gap you saw in the connect-to-meeting numbers: 7.44% for power dialers versus 3.88% for parallel. Your rep got the conversation and wasn’t ready for it.
5. It spends effort in the wrong place
A parallel dialer is built to spread effort across as many names as possible, as fast as possible. But what actually gets people on the phone is calling the same person more than once, over several days. In our data, the share of prospects a rep reached kept climbing with every additional attempt, and it was still rising past the sixth. About 94% of all connects came from the first five attempts. Every additional call attempt for each contact connected less often than a first call to someone new. By the fifth attempt, average connect rates are less than half of what you’d expect from a first call attempt for a new contact. Everything past the sixth attempt used 8.3% of all dials to produce just 4.0% of all connects.
So that’s your point of diminishing returns: about 6 call attempts per prospect. (For what it’s worth, we make 7 attempts per prospect at TitanX.)
Even so, nearly half of the reps in our data averaged fewer than 2.5 calls per prospect. When you hand reps a tool designed to get through the list faster, you push them in exactly the wrong direction: more contacts called just once, and fewer people actually reached.
.svg)
