BASIC DIALER MANGEMENT

Outbound contact center managers face an array of challenges. Managing everything from employees and call solutions, strategies and processes. Todays managers have become multidisciplinary, always on the lookout for new formulas and best industry practices and golden rules of various outbound contact centers, from the telemarketing and collection perspective. The most common question that an outbound call center managers faces is How do I optimize the soliciting activities The general challenges being segmentation strategies, dialer productivity optimization, resource management and tools. These terms mentioned above are more synonymous with utomatic dialers! these dialers increase contact center efficiency because they eliminate the possibility of reaching bad numbers, save time and make contact center agents more productive. "y automatically dialing and screening for busy numbers, no answers and answering machines, dialers ensure that agent time is not wasted in non value added routine tasks, Only when the dialer reaches a live contact will it transfer the call to the ne#t available agent, with a delay that should not take more than a few seconds according to the pacing desired by the Operations managers. There are si# modes for the outbound dialing$ %redictive, progressive, %review, database, speed dial and manual. &ach has the screening capabilities mentioned above as well as common management tools and processes for reporting and list creation, but that is where the similarity ends, for this reason, the dialing mode to apply to any given

campaigns must be carefully selected, since each available dialing mode has advantages and disadvantages of its own as well as cost and efficiency considerations. Predictive dialing mode This is the most commonly used dialing mode$ %redictive dialing this mode is widely used because it is, without doubt, the most productive for agents. %redictive dialers use historical statistics and sophisticated algorithms to calculate agent availability. Thus based on past trends, the dialer will know e#actly how many dialing attempts and how much time required to reach a live contact as well as e#actly how many dialing attempts and how much time is required to reach a live contact, as well as e#actly when an agent will become available to take the ne#t outbound call. Therefore, the dialer can start dialing to reach the ne#t available and connect this live contact as soon as or shortly after an agent becomes available. 'ince the predictive algorithm analyses a number of factors, including the number of available telephone line, agents and the probability of a call not being completed (e.g. busy signal, no answer, etc) it becomes impossible to establish a so called * magic+ rule to determine the number of agents required for the planning algorithms pf the predictive mode to select the perfect rhythm. ,owever, a dynamic rule can be constructed according to three closely linked criteria! in order of importance, these are$ -all length .ength of after call/wrap up commonly known as wrap time The number of gents for one campaign and its sub lists, 0hen properly used, predictive dialing can have a dramatic effect in the contact center. 1ormally, the busy rate for the

agents between manual and predictive modes increases by 234 to 534

The golden r le !or "redictive dialing i#

ni!ormit$

0hether there are 6 or 76 agents, the important thing is that all agents have an average talk time and after call wrap up time that varies from one agent to another, so that there is constancy to the events for the predictive algorithm. Obviously, the shorter the calls are, the more this consistency of one call to another can be respected. Thus, a number of best contact centers are using the predictive dialing mode for some of their outbound campaigns with as few as 5 or 8 agents and with an average abandon rate of 24 to 64, because calls are short and very homogeneous. 9t is thus optimal to create groups of agents with the same e#perience and skills, since the predictive algorithm will be at its best when all calls from a specific agent group have a very similar talk time, after call work time, and not ready time. 1ew agents should therefore belong to a separate group until their productivity becomes more aligned with other groups! they can then be paired with other more efficient agent groups. 0ith governments in various countries imposing restrictions on the number of abandoned calls as well as attempts made intelligent use of a predictive dialer is not a choice :its a necessity. 9n predictive dialer terms, an abandoned call is often one where the dialer releases the line after a person answers the call because there is no agent available to take the call. Obviously, the abandon rate will be directly linked to the productivity criteria. Thus, some contact centers use the predictive mode even with as few as five agents and maintain normal abandon rate, while other contact centers with three times as many agents, do not even achieve a good rhythm in predictive mode, because their calls are too

lengthy and lack uniformity in talk time, after call work, and not ready time.

The cam"aign "ace i# the %e$

Many dialers allow contact center administrators to set an acceptable abandon call rate. ;overnment requirements and <M (<irect Marketing ssociation) recommendations are usually (less than 64). The dialers will automatically ad=ust the call initiation rates the abandoned threshold approaches or recedes. 9t is thus important to remember that the faster the pace, the more abandoned calls will occur, but too slow a pace can affect profit potential. 9t is also recommended to optimize management of dialing rates in predictive mode by adapting the percentage of agent busy rate. This parameter enables a level of uniformity between telephone performance to be achieved, and can have a very positive effect on abandon rates and wait time between calls. On the other hand ad=usting this parameter too frequently can destabilize the dialing algorithm by increasing the abandon rate or the wait time between the calls.

Progre##ive Dialing&
nother widely used dialing mode is the progressive mode. 9t works in a similar manner to the %redictive mode, e#pect that dialing only begins once an agent becomes available, which of course will almost completely eliminate abandon calls. This is due to customers hanging up before the dialer can transfer the call to an agent once a live answer is detected. &ven though a call is not automatically ready when an agent becomes available, this dialing mode still has good advantages and will greatly optimize agent productivity, because the dialer quickly handles all the call screening. 9n fact, studies show that the busy rate for the agents should increase by >64 to 234 using this mode.

Previe' dialing
Obviously certain sales strategies cannot be implemented in either %redictive or %rogressive mode, particularly if the agent must study the customer details before making the call. 9n these cases, %review dialing can be the best choice. This dialing mode shifts the dialing control from the dialer to the agents. "y doing so, it allows agents to carefully study a contact before any dialing occurs! the agent gives the dialing command to the dialer, only then will the dialer screen for a live contact. 9f a live contact cannot be reached, the agent can still add notes to the client file, decide whether to reschedule this call and move on. %review dialing mode is therefore used for "usiness/t o: business sales, where contacts are limited, but the value of each is great, and where opportunity, preparation, and strategy are key elements. This dialing mode still offers good benefits, especially for outbound campaign management and monitoring and since the acquisition and costs deployment costs of this mode tend to significantly lower than for other modes, the return on investment remains very interesting. 9n conclusion, when used as part of a well : planned campaign and with respect for regulations concerning abandoned call rates and Do Not Call requirements, predictive dialers can produce impressive results. "ecause a higher number of calls can be handled, there is significant increase in agent productivity and sales. &fficient dialing and screening for machine responses reduce operating costs. gent productivity is also a desired outcome. ?eal : time campaign monitoring and historical reporting are also key benefits of any dialing solution, because you can only manage what you can measure! ()NDERING (*+ +), RES,LTS ARE-NT AMA.ING/
<ialer users capture statistics on lists and campaigns, on agent telephone

data, and on the success or failure of their promotions. However, despite the

information gathered, some information remains difficult to access because it is scattered among different reports. Even wore, the perception, evaluation and conclusions reached may vary from one group to another, because there is no unified view of the statistics collected. Knowledge is power! This is why a good contact center solution must include adequate measurement units, both real-time and historical, and why good contact center management must refer to best practices when using metrics and creating reporting templates. In managing outbound call activities, it is important to know what to measure, why you are measure it, and how to measure it. What, Why and How to measure The first step is to determine e#actly what should be measured to produce meaningful statistics that can be measured to give a true picture of contact center activity. There are four areas that should be measured$ The performance of the dialer and its calling lists, agent productivity, the profitability of the activity, and the criteria that determine success or failure. Dialing results The main metrics to measure and some ob=ectives to target$ Non0tran#!erred call#$ These are very good indicators of the contacts quality o a given calling list, as well as the call timing used for a specific campaign. 9tems should include$
Signals Invalid signal/No signal/Fax Busy No Answer oice !ail "ox A"andoned call Contact center average 2% 8% 10% #1% #%

 An#'ered call#& These are the targeted outcome of a list, and where the work of the agent begins. 9tem measured should include$ 0rong numbers ?ight party connect 9t is very hard to give an ob=ective for right or wrong party connect, 'ince this will vary significantly depending on the type of the contact strategy, e#isting customers versus prospects, as well as on the age gender, and role of the contact within the household. 'cheduled call back to complete the offer -ompleted offers Messages left, 9f part of the calling strategy 0hether to leave messages depends entirely on the contacts, the campaign, the list, and the nature of the promotion. 'ince outbound dialing is often a numbers game, leaving messages is often an obstacle to agent productivity and sales activities. 9deally, messages should not be left for prospects, since they do not know or have a relation ship with respective contact center. 9n some "7" situations, however, messages will be left, for e#isting customers, contact centers tend to leave messages in highly targeted campaigns or promotions, like renewals. These cases, a more significant number of messages tend to returned. 9f messages are left, two important actions must be taken. @irst, on the dialing strategy side, the contact center must identify when to leave a message, after how many attempts, and how long should they wait before dialing this contact again, 'econd, on the measuring side, they need to capture statistics regarding the number of messages left, the ratio of returned calls, and delay before callback, in order to improve their approach the ne#t time around.

9f a message is left, the contact center should also a devise a strategy for answering the calls returned calls. 'ince the very nature of the contact centers is to rapidly route a call to the ne#t qualified and available agent, its preferable to route returned calls to the agents of the inbound contact center. The agent who left the message will most likely be online with another customer when the prospect returns the call.

List penetration rates n important indicator to measure when it comes to dialing results is list penetration. The penetration rate of a list must always be measured because it represents the total number of records or contacts completed. good penetration ob=ective is around 864. %enetration of less than 834 indicates a less profitable list, while more than A34 significantly reduces dialing algorithm productivity, since it is more difficult to reach contacts, and most of these are second, third, or more attempts. Agent productivity tatistics

Telephone statistics are the most commonly used indicators of agent productivity, because they can easily be collected manually, even if a dialer is not used. They have also been used in inbound contact centers since the introduction of the -<. gent %roductivity statistics are measured to show the telephone productivity of the agent and to benchmark them to the group or past results. lthough useful, they do not indicate the quality of the response or the relevance of the sales approach. @or this reason, the quality aspects of the monitoring activities are taken into consideration when evaluating global agent performance, as well as sales results. n agent who shows a higher sales ratio often represents higher profitability.

The most commonly measured call statistics are$ -%, (-ontacts per hour) '%, ('ales per hour) -loseBconversion rate or '%- ('ales per -ontact) 0rap time Talk Time .ogin Time 9dle time

9t would be very difficult and pointless to present industry averages, since these statistics vary significantly according to sales strategies, campaigns, lists, and promotions. Call termination codes One source of information used to better understand why sales offers or solicitations fail is, with out doubt, the call termination codes, lmost all contact centers use these codes, but incorrect coding is a ma=or problem in many contact centers. Termination codes can be rich source of information for customer response analysis and the subsequent ad=ustment of offers, agent skills, and e#pertise. ,owever, in order get the correct information, it is necessary for the agents to take the time to code each of their calls properly. 9f calls are not coded appropriately, the e#ercise is useless. 0hen coding calls, less is more! the list of call termination codes should therefore never contain more than 73 choices, including sale code and removal from calling list. %referably, >3 would be more efficient. lso, the choices must be sufficiently clear and varied so tat each choice can be revealing enough to use as a basis for the improvement of sales arguments, agent skills, calling lists and promotions, This method also avoids over/use of *other+. 9deally if an

agent selects the *other+ choice, heBshe should be required to enter a brief e#planation to improve the code choice menu or the understanding of the codes. The e#planation could be simply two or three words that clarify what *other+ means to the particular call. The target ob=ective should be not more than 74 to 24 of calls coded as *other+. To improve the success rate of code use, it is important to offer a scrolling menu that displays all possible choices, and to require that a code be selected before he agent can complete the file and close it.

+), *A1E ALL T*E STATISTICS23 N)( (*AT/ 0e have discussed what, why and how to measure/but what do you do with all this informationC 0hether you look at the raw data or create reports based on the statistics you have collected, the results will clearly indicate what you are doing right and where your strategies or processes need improvement. "ecause some of the data collected depends on input from other departments within your organization. The results can often have an effect beyond the immediate realm of the contact center. @or e.g. dialer results depend mainly on the quality of calling lists used, for this reason list accuracy with valid contacts and their telephone numbers, with the fewest possible wrong numbers is e#tremely important. The number of busy signals and voice mailbo#es received also has an important effect on agent performance. -alling lists are usually the responsibilities of clients or the marketing department of the contact center. ;ood detail reports from the contact center will help to improve accuracy of these lists. The number of busy signals and voice mailbo#es reached by the system (dialer) becomes the direct responsibility of the contact center campaign management team.

t the list creation stage knowing who to call and when to call are key elements in the campaign timing and contact center hours of operations. On the daily campaign management side setting the dialer call back strategy and call rescheduling when reaching the third/party connect become key elements for cost efficiency and good agent productivity. t the end of the campaign, the analysis done should be re/used when creating sub/lists or campaigns for future purposes. <uring the calling period we should consider list penetration, number of attempts per lead, abandon rate, new leads attempts on a regular basis to maintain uniformity of performance. There is a direct correlation between the number of times the system dials before reaching a contact and t he agent *wait time+ between calls. @or voice mailbo#es, it is worth using the best calling hours according to the geographic zone. 0ith planned strategies with limited number of callback tries at different times. @or busy signals we recommend a tighter call back strategy, for e.g. following a busy signal the strategy would be to call back >3 minutes and then >6 minutes later and then the following day. limited number of tries should also be planned. To optimize calling lists and campaign productivity the recommended strategy for dealing with voice mailbo#es and busy signals is D/6 attempts in a week (sub=ect to client requirements.) @inally if the data are not in line with the established ob=ectives, different contributors will have to analyze the situation diagnose the problem and then find the solution.

SEGMENTING LISTS0 (*+ BIGGER IS N)T AL(A+S BETTER

9t is a well/known fact that the calling list can either make or break an outbound campaign. -ontact centers often use a

single bulky calling list. This list is usually divided or segmented into several campaigns. 'tudies show that the size of the calling list within a given campaign can vary from >333 to >63,333 leads. ccording to the circumstances contact centers create campaigns and sub/lists. @or e.g. -reate campaigns time zone wise, area code wise etc. well/segmented list can produce e#cellent results improve relationships with customers and provide both, motivation for contact center agents. -alling lists segmentation is done at two levels$ -lient or the marketing team and at the contact center itself, -lient or the marketing tea is responsible for creating the master list of e#isting customer or prospects, while clientBmarketing teams segmentation can optimize the campaigns success, contact center list segmentation allows the proper allocation the proper allocation of agent groups and sub/lists, which enables some elements to be divided and measured precisely and thus feed statistics back to the clientBmarketing team. The feedback gives clientBmarketing team the opportunity to analyze and better segment future campaign lists. The first segmentation done by the clientBmarketing team is based primarily on demographic factors and past purchasing behavior. ;iven the money invested in telemarketing and the very limited number of times customers can be targeted, no *strikes+ are allowed. -hoosing the best time and matching the best offer to the most likely customers is an art and an important contributing factor in acceleration the profitability of a contact center. The second segmentation done at the contact center level divides each promotion into distinct campaigns and sub/lists based on various criteria like language, geography, and promotions. These campaigns can then be assigned to agent groups with the appropriate or specialized skills to handle them.

LIST AND CAMPAIGN LI4EC+CLE0(*+ IT IS ALL AB),T TIMING3

Once the segmentation is over, knowing how to use the list is essential. Timing (how long the lead list can be used) the length of the campaign is crucial. To optimize the predictive aspect of calling strategies, the average length of the campaign should never be less than 7/2 hours. 9t is thus vital to avoid over segmenting your campaign. On the other hand the lifespan of a list should not e#tend beyond 2/D weeks because customers called towards the end of the campaign would have been attempted before and were reschedules. Eery long lists can limit the ability of contact center to be proactive because it is difficult for them to measure the relevance and quality of contacts on the list before it is completed. @urther more towards the end of the call list, the quality of remaining contacts and productivity deteriorate. Thus if the list is very large the period near the end of it can e#tend over several hours, which can affect agent focus, sap their motivation and significantly reduce productivity because of longer wait time between calls. The shorter the list, the more these unproductive periods can be avoided. 9n the collection environment, calling lists must be even shorter to take into account the daily payments made by the customers without the intervention of outbound collection calls. The lists must either be the appropriate size to cover one day of outbound collection calls or if a bigger list is chosen, it must be refreshed everyday. G)LDEN5 N)T MAGICAL R,LES3 -ontact centers differ in size, purpose, business type, and technological tools. Making it impossible and undesirable to prescribe a rigid set of rules or behaviors that guarantee successful campaigns. There are however

certain guidelines which result from commonsense and in/ depth e#perience, which can be recommended as golden rules, these are$ >. The size of a list must be calculated according to the number of agents, call length and calling hours and not =ust according to the number of contacts. 7. 9t is important to have a central calling list large enough to allow segmentation into sublists, but not large enough to stretch over several weeks of calls. 2. 0hen it comes to solicitation *less is more+. @or e#isting customers it is preferable not to solicit by telephone more than D times per year, while for all direct marketing communications an organization should not be solicitating more than A times per year. There are no magical rules that can guarantee success but by keeping the best practices and golden rules in mind when creating and segmenting lists and when creating and assigning agent groups for these lists will ensure that managersBsupervisors can track the progress of their campaigns and will go a long way towards ensuring that resources are motivated and satisfied. braham .incoln once said, *I! I had #i6 ho r# to cho" do'n a tree I-d #"end the !ir#t !o r #har"ening the a6e creating and segmenting your list is were you need to invest your time so you can reap all the benefits of the outbound campaign.