Reading time : 6 min

Preview, progressive or predictive? Which mode should you choose for your call campaigns?

Whether it is a question of loyalty, data collection, information or commercial promotion, dialing engines are essential tools for companies that carry out inbound and outbound call campaigns.  These engines generally offer three campaign management modes - with contact preview, progressive and predictive - as well as a number of options to increase caller reachability and optimize agent occupancy. 

Before describing these different modes and presenting their respective advantages, it seems essential to insist on a point that remains decisive for any campaign’s success : the quality of the contact file.

Take care of your customer segmentation and your files

It's very simple: the more useful information your file contains about the contacts you will try to reach,  the more satisfactory the effective contact rate will be. This means that you need to work upstream on consolidating the information you have (typically in your CRM, billing system or other) and using this information to segment your file according to the specific objectives of each campaign.

It also means that the campaign management system you use must accept files with as many fields as necessary to properly qualify each contact [demographic data, relationship history, loyalty status...] and provide agents with enough material to personalize each conversation. If the contact form only has a name and a phone number, it will get harder for the agent to convince the contact person,  even with very good arguments...

Here is what you need to know about the different numbering modes to make an informed choice and achieve your objectives. At the end of this article, we will also explain the latest advances that will make it possible to go beyond the limits of each method to reinforce the added value of outbound call strategies.

Preview mode - The most qualitative under certain conditions

When the system detects that an agent is available, it presents him with a file containing all the information contained in the file or in external databases (CRM or other). The agent has time to review the file by previewing the situation and the context of the conversation he or she is about to have. He can therefore prepare his introduction and speech. Once he is ready, he can trigger the call by simply clicking on a button in its interface. Most advanced systems offer options and variants for this numbering mode:

  • give the officer the opportunity to refuse a record and proceed to the next record;
  •  apply a more restrictive mode where the agent has no choice and must make the call;
  •  set a time delay (called "automatic delay") after which the record number is dialed no matter what happens.

While this management mode offers the greatest comfort to agents, it is not optimal in terms of time management and productivity: it is not possible to know if the contact  1/ will pick up and 2/will actually be available. The agent can indeed spend a significant part of his time preparing unnecessary calls that will ultimately not succeed, which can lead to frustration and demotivation...

This is why we recommend using this method of execution for very qualitative call campaigns where contacts could be notified in advance that they are going to be called. It is perfectly suited to campaigns carried out by specialized consultants or experts. The optimal use case is when you start from a list of scheduled appointments or reminders that guarantee a much higher reachability rate than cold calls. The Diabolocom telephony solution makes it very easy to create such lists and, in this way, to process outgoing flows, hitherto diffuse, in the form of structured, more productive campaigns.

Progressive mode - A good compromise between reachability/acceptability

The main difference with preview mode is that there is no preview time. As soon as an agent is available, the engine dials the number and at the same time returns the contact card to the workstation. The agent therefore has only a few seconds to read it before the conversation begins.

The systems offer a very useful option to maximize the agents' useful time: the progressive mode with answering machine detection. In this case, the card is not distributed at the time the number is dialed. The engine makes a pre-call and if it falls on an answering machine, it switches to the next number. On the other hand, if a real interlocutor answers the call, the record is traced back to an agent.

>>> Having less time to prepare the call can be very effectively compensated by next-generation conversation guides/argument guides. Far from the call scripts of yesteryear, these guides are automatically customized from the contact's data. Combined with automatic answering machine detection (AMD, Answering Machine Detection), the progressive mode eliminates "unnecessary" calls, which lead to agent productivity and has the significant advantage of ensuring that callers are put in immediate contact with an agent. The brand or company thus minimizes situations, which are less and less tolerated, where the contact stops but has no one on the other end of the line.

Predictive mode - Delicate to balance and controversial...

Even if everyone talks about predictive call campaigns, it would be more appropriate to talk about over-numbered Call Campaigns. The notion of "prediction" here refers to the system's ability to use activity data (talk time, dropout rate, occupancy rate, etc.) to calculate how many numbers should be dialed at some point. Logically, to achieve this balance between agents and called parties, you should take into account unanswered calls and those resulting in answering machines, which means that there is a need to over-number, i.e. dial many more numbers than the number of agents.

Thanks to the real-time statistical analysis of activity data, it is now possible to go very far in the adjustment between the number of calls and the number of agents. Despite these possibilities,  there are not enough agents available quickly enough when people are called up and drop out. These situations - where the system literally hangs up on the people contacted because there are no available counsellors - are very poorly experienced. Increasingly assimilated to nuisances by individuals, over-numbered campaigns are recognized as such, from a legal point of view, in a growing number of countries. This is the case in the United Kingdom where campaign operators with a drop-out rate (or nuisance rate) above a certain threshold face very heavy fines.

Give more value to call campaigns

Given the cost structure of outbound call platforms, the increasing difficulty of reaching targeted people, particularly in the context of mass market campaigns, and the foreseeable tightening of regulations, prime contractors and campaign operators will necessarily have to change their approach and change their call strategies. In the future, in both B2B and B2C, call campaigns will only be effective and value-creating if two conditions are met:

  • Better qualify the people contacted - As we said at the beginning of this post, any campaign must be based on a detailed analysis of the contacts: when can they be reached? Given their profile, are they really likely to be interested in the offer we are going to offer them? Etc. Rather than focusing on volume, it is necessary to focus on customer knowledge, and to use the ability of call engines to integrate this knowledge to define more targeted call and recall strategies. For example, our solution allows us to specify that people with a certain profile (gender, age, place of residence, etc.) should be called in the middle of the morning rather than around lunch or the other way around. This data-based work results in much better reachability rates and more satisfying and productive work for agents.
  • Preventing people upstream to validate their interest - Consumers feel over-solicited and are increasingly rejecting unsolicited calls outright, without even trying to find out what the call would be about. Diabolocom customers tend to send an SMS or an e-mail to their own customers or prospects, explaining the reasons why they should be contacted. By giving prospects the opportunity to tell what time they are available, the agents can gain time and focus on the people that are really interested in their call.

Going further with deep learning

Campaign operators working on very large volumes will increasingly be able to rely on artificial intelligence, in this case deep learning algorithms, to better analyze past business trends and find new working hypotheses to optimize contact reachability. By crossing multiple criteria on large volumes of data (activity and contacts) and without interpretation bias at the beginning, it is possible thanks to these algorithms to go beyond simple statistical analysis and to uncover patterns allowing to build new reachability strategies, to better dose contact initiation, reminders, etc. The possibility of integrating these scenarios and strategies directly into the call engine to execute them automatically opens up opportunities that could restore the reputation of call campaigns and consign current worst performing and most controversial practices to oblivion. We are working on it!