Introduction

Call rescheduling, also known as callback, allows scheduling a new call for a specific date and time when a contact cannot continue the conversation at that moment. This document applies to the current operation of predictive and progressive campaigns and is aimed at operational and functional users to help them understand what to expect from the system when scheduling a callback. It is also important to note that this explanation covers the current behavior of the system but does not include technical configurations or future changes.

What Is Call Rescheduling?

It is a feature that allows scheduling a new call for a contact at a specific date and time when the advisor has already had a previous interaction and the customer cannot continue at that moment. This action is known as rescheduling or callback.

What Types of Callback Exist?

There are two types of callback:

• Automatic: The system makes the call at the scheduled time without advisor intervention.
• Manual: The system notifies the advisor, who then decides when to make the call.

What Is the Recommendation for Correctly Scheduling a Callback?

It is recommended to schedule it at least 30 minutes in advance of the current time. This gives the system enough time to detect, activate, and execute the call within its internal cycles.

Is the Callback Executed Immediately at the Scheduled Time?

No. The process is not immediate. Execution depends on two automated processes working together: first, the activation of the callback, and then the dialing of the call. Therefore, even if a call is scheduled for a specific time, its actual execution depends on the system's processing windows.

How Does Callback Activation Work?

There is a process that runs every minute but only checks callbacks at minutes that are multiples of 6. At that time, it identifies calls whose scheduled time is within the next 10 minutes and marks them as ready to be executed. This means a callback can take up to 6 minutes to be detected once it enters the execution window.

How Does Call Dialing Work Once the Callback Is Active?

Once the callback is active, it enters the dialing engine. This engine runs several times per minute, decides whether to generate calls, and defines how many to launch in each cycle. However, callbacks are not processed at any time; they are only considered when the current minute is a multiple of 5 and corresponds to the first execution cycle of that minute. This creates a very short execution window of approximately 15 seconds every 5 minutes. If conditions are not met at that moment, the callback must wait for the next cycle.

What Conditions Can Prevent a Callback From Being Dialed in a Cycle?

There are conditions that completely stop dialing. If any of these occur, the system does not generate any calls in that cycle:

• No agents are logged into the campaign.
• No records are available to dial.
• The final call calculation results in zero.
• The campaign is in preview mode.
• There are active time restrictions, such as consumer protection.

In these cases, the callback is not lost but must wait for the next valid cycle.

What Happens When There Are Few Logged-In Agents?

When there are 5 or fewer logged-in agents, the system operates in progressive mode. In this scenario, it generates one call per available advisor. If there are no available advisors, no calls are generated. In this case, there is no complex calculation, just a direct relationship between ready agents and calls.

What Happens When There Are More Than 5 Logged-In Agents?

When there are more than 5 agents, the system uses a predictive model. It calculates how many calls to launch in parallel, considering factors such as:

• Total number of agents.
• Agents available at that moment.
• Average handling time.
• Probability of contact.
• Configured campaign speed.
• Recent service status.

Does the System Dynamically Adjust Dialing?

Yes. Depending on the real-time status, the system can:

• Accelerate dialing if many agents are available.
• Reduce it if few agents are ready.
• Slow it down if there are too many ongoing calls.
• Adjust it if the server is under high load.

Are There Protection Limits That Affect Execution?

Yes. Even if the calculation indicates that calls should be generated, there are limits that can reduce or prevent execution, such as:

• Available server capacity.
• CPU usage.
• Limit of simultaneous licenses.

If any of these limits are reached, the system reduces or stops dialing.

Does Callback Always Have Priority Over Normal Campaign Records?

No. The system does not constantly prioritize callbacks. It only evaluates them at specific times, i.e., at multiples of 5. At other times, it processes normal campaign records. Therefore, a callback competes for a limited execution window.

What Is the General Flow of a Callback?

The flow works as follows:

• The advisor schedules the call,
• The system waits for the next multiple of 6 to activate it,
• The callback is ready to be dialed,
• It waits for the next multiple of 5 to be picked up by the engine,
• If conditions are met, the call is generated.

What Happens If a Callback Is Not Answered?

It is not immediately discarded. The system automatically schedules a new attempt 30 minutes later, and this new attempt is processed under the same conditions as the initial call.

How Many Attempts Does the System Allow?

The system allows up to 3 attempts in total:

• First attempt: Scheduled time,
• Second attempt: 30 minutes later,
• Third attempt: 30 minutes after the second.

If none are successful, the callback stops being processed automatically.

How Long Can a Callback Really Take to Execute?

The total time can vary. A callback can take:

• Up to 6 minutes to be activated,
• Up to 5 minutes to be executed once activated.

In the worst case, it can take approximately 11 minutes to execute from the time it enters the window.

What Is Important for Users to Understand?

There are two key ideas:

• In predictive mode, the system can generate more calls than available agents, anticipating that some agents will become free;
• If conditions are not favorable when the callback enters evaluation, it is not lost, but it is not executed immediately either; it must wait for the next valid cycle.

In wolkvox, call rescheduling does not depend solely on the scheduled time but on an internal activation cycle, dialing windows, operational conditions, engine capacity, and automatic retries. Understanding this behavior allows for scheduling callbacks with better expectations and comprehending why a call may not go out exactly at the expected time, even if it was created correctly.