• The percentage of calls answered/completed/connected is nearly entirely dependent on your recipient list’s quality, freshness, source, etc.
• Roughly 50% of calls which are answered are answered by a human This is important for survey broadcasts because you can’t collect input from a voicemail box. After millions of calls, we have seen broadcasts where as few as 35% of completed calls were answered by a human, and broadcasts where as many as 65% of answered calls were answered by a human. Many factors influence this percentage (and are discussed in another article), but include the time of day of the broadcast, the day of the week, the quality, source, and freshness of the recipient list, etc.
• Of human answered calls, roughly 16% of those humans respond to the first question. Again, many factors might influence this percentage, including the length of the survey introduction, the length of the question, the eagerness of your recipients, whether you tell them how long the survey is, etc.
• Approximately 1%-4% of respondents will hang up after each question. Again, this is a wide range with multiple influences. How long is each question (one survey of note asked respondents at the start for a 1-5 rating of each option and then breezed through 20 options and saw 2% hang up after the first question, and about 0.3% hang up after each question, with 12.2% responding to all 20 questions).

After a voice broadcast call was answered by a machine/voicemail, no voicemail was left.

Possible Causes:

• Seen frequently when testing broadcasts, if the caller-ID used in the broadcast is the same as the number dialed, likely the call did not reach the voicemail box because it appears to the carrier that your phone is calling itself. Many carriers have configured their network such that when a phone calls itself (i.e. the calling-party is the same as the called-party), the call is redirected to the voicemail main menu where the caller is prompted to enter their password in order to check their messages. Since this is a different menu than the voicemailbox where a message would be deposited, no message is left in the voicemailbox.
• Some carriers have configured their voicemail systems to discard any voicemail messages which contain too much silence at the beginning of the message. Some possible reasons why too much silence might exist at the beginning of a voice message is discussed in another article, but the simplest solution is to remove any silence which exists at the beginning of the voice message which is sent on the broadcast.

There is a long silence between the time a voice broadcast call is answered and the time the message begins playing.

This is an age-old problem, detrimental to the quality of your broadcast and likely the receptiveness of your recipient toward the call. Of course, we want to reduce this silence as much as possible to give the listener the best experience.

Background:

In another article, we discuss how to determine whether a call is answered by a human or by a voicemail system by analyzing the initial greeting of the recipient. Gigabark also determines when the initial greeting has ended, so it can determine when to start playing the message.

One solution to this silence problem is to simply start the message as soon as the call is answered. However, this may not increase the recipient’s experience, and may actually decrease it. The message may play over the recipient’s greeting, or the message might be severely clipped by a voicemail greeting or the time it takes the listener to place the receiver to their ear.

But how does the analyzer know when the initial greeting is over? Of course, as long as there is voice activity coming from the recipient, the analyzer thinks the greeting is not yet over. At some point, the greeting will end, but the analyzer continues to listen for a short period of time for more voice activity before declaring the greeting ended. So some delay prior to beginning the playing of the voice message will always occur, but it is generally very small (less than one second, but typically less than half of a second).

Causes for Silence:

The following may be reasons for silence prior to the start of the message:

• The analysis of the initial greeting (discussed above)
• NO initial greeting is given. For example, when testing a broadcast you might not say “Hello” since you expect a recording. This throws off the analyzer, which assumes a voicemail system answered and waits longer.
• Silence in the message itself.
• Background noise on the recipient’s end may be the biggest cause. If the analyzer thinks that background noise is part of the recipient’s greeting, it will continue analyzing until there is silence. It will not start playing the voice message until it believes the initial greeting is complete.
• Particularly in messages left on voicemail, there is more delay in the analysis of the greeting due to the wide variety of voicemail systems. Since some systems play a tone to signal recording and some don’t, it isn’t sufficient to analyze for the tone, but for the end of voice activity. To make matters worse, since the systems may wait up to 3 seconds after the mailbox greeting to play the tone, we must analyze for much longer to verify the end of the voice activity prior to playing our voice message. We do all of this work to prevent playing the message prior to the time the voicemail system starts recording. It is also worth noting here that some voicemail systems listen for too much silence at the beginning of voicemail messages and may discard the voicemail messages if it contains too much initial silence. This may explain why a voicemail message may not be deposited at all.

One of the most difficult things for any automated calling system is to determine if the party which answers is  voicemail or an actual human.

Because of the many different types of carriers, handsets, voicemail vendors, and voice delivery technologies which make up today’s telephone networks, there is no reliable way to determine whether a call has gone to voicemail or if an actual person has answered your call (sometimes even a carrier will mistakenly think a call is answered when it is not, such as in the example of an old-fashioned telephone answering machine. These devices actually answer the call to play a recorded message). Even if we ignore this small percentage of people with old-fashioned answering machines, it’s unlikely carriers would open up their networks to share this information with a large number of outside businesses on a per call basis.

Decision Criteria:

Since we can’t be 100% sure that a call has been answered by a machine or a human based on the network technology, we must base our decision of Human vs Machine elsewhere.

It turns out, people exhibit strong tendencies to how they answer the telephone or how they record voicemail greetings. So strong, in fact, that estimates suggest we can be as high as 90-95% certain that a particular call was answered by a human or a machine by analyzing the way the call was answered. The following tendencies are the foundation for this analysis:

1. Generally speaking, humans tend to answer telephone with a short greeting such as “Hello”, “Hi”, or “This is Nathan”.
2. Generally speaking, voicemail messages tend to be much longer, such as “I’m unable to come to the phone right now, please leave your name and number…” or “I’m sorry to have missed your call, but I will be unavailable from 1PM to…”

So, if we have an analyzer listen to the initial greeting, we can do a pretty solid job of determining if a live person is on the other end… or if you’ve reached a machine.

Drawbacks and Limitations:

Of course, there are a number of nuances and exceptions that bring down our confidence in our determination. Here are a few:

1. Live person with long greetings. Retail businesses tend to require employees to answer the telephone with much longer greetings to announce their slogan or their specials, such as ”Thank you for calling Joe’s Pizza, would you like to try our triple-pepperoni and sausage special for $9.99?” or “Dave’s Car Sales, where everyday is a great day to buy a car! This is Pete.” Some people may also answer their phone in a long-winded way. These events will cause the analyzer to mistakenly judge these calls to be answered by a machine.
2. Background noise may be the biggest hindrance, especially in today’s cell-phone world. The analyzer may think that background noise is part of the greeting, and consider a short greeting to be a very long one. This is probably the biggest reason why a human-answered call might be judged to be a machine-answered call.
3. Delays prior to the initial greeting or slow speaking greetings may throw off the analyzer, since it may determine that a greeting was over when it actually wasn’t finished.

But in general, these are the exceptions rather than the rules. Especially if you’re less likely to be automating calls to the retail line for many types of businesses or calling home numbers instead of cell phones, these may not have a significant impact in your confidence levels in determining Human vs Machine.

Conclusion:

Overally, this method for determining Human vs Machine is a fairly reliable method, and its impact to your broadcast is likely small.

This method does have some bearing on the quality of the voice messaging being played. Specifically, the amount of silence that is played prior to the message might be a direct result of the greeting analyzer’s configuration. We cover this topic more detail in another discussion.