Contingent Demand: What It Is, And Why It Matters

In my last post, I went out of my way to thrash the conventional approachto Workforce Management (WFM). I meant it entirely – the conventional approach to WFM is indeeddoomed to failure. But that doesn’t mean that contact center managersmust be doomed to failure (for one thing, many people have “succeeded” ascontact center managers even while living through continuous customer andagent angst and never-ending budget challenges).

In fact, there is a better way to manage staff scheduling, which involveslooking at the problem from a slightly higher altitude. By “higheraltitude”, I mean to take an old idea – the use of outbound blending totake the edge off the curve – and then to abstract that idea to somethingmuch larger and more meaningful.

This process can be visualized as whathappens when examining an element in a landscape from a balloon whilegaining altitude – what at first might look like an odd detail oraberration turns out, from a higher altitude, to be easily identifiableas, for example, an oxbow loop in a long river. Specifically regardingscheduling etc., what “higher altitude” gains us is a better look at thereal problem.

WFM has traditionally been focused on solving the problem of ensuringthat adequate resources are on hand to handle all inbound demand. Thiswas viewed as a single and challenging problem because customers don’twant to wait, and CFOs don’t want to pay for lots of extra agents. Sinceoutbound calls, for example, are by definition initiated by theenterprise and not by the customer, they have been generally been viewedas presenting a different set of problems than the scheduling oneattacked by WFM (for example, the problems of answering machinedetection, dialing rate control to avoid overdialing, and so forth).

Because the problems of inbound and outbound management have been viewedseparately; because by and large the “solutions” to the problems havecome from different vendors (WFM/WFO vendors and Outbound vendors,respectively); and because the problems were generally by managed bydifferent managers in contact centers (staffing managers and campaignmanagers, for example); the common thread that can help both problems hasbeen missed.

This common thread can be understood by starting with the notion of”demand” in the contact center context, and rethinking what it is and howto manage it. We can think of demand as coming in two basic types:imperative and contingent. Imperative demand is demand which customersplace on the system, and which must be satisfied essentially as soon asit is made (that is, for example, you want to answer inbound calls assoon as they are received). Contingent demand, on the other hand, isdemand that can be managed, and can in some sense be made “contingent” onexternal factors (such as, importantly, the amount of imperative demandthat must be dealt with). Some contingent demand may becustomer-initiated but not time-critical (for instance, emails aretypically best handled as contingent demand). Usually such types ofcontingent demand are not time-critical because the initiating customerdoes not have to “stay on the line” and may not expect an instantresponse (email is this way). Other types of contingent demand may beenterprise-initiated, and may not even involve a customer directly.

Examples of this include work items (screen loan applications, checkingstatus of open trouble tickets, and so forth), chat barge-ins to assistcustomers on a web site (as long as such assistance is perceived asoptional, it can be initiated for some, all, or none of the customerscurrently using a website, as circumstances permit).

One key aspect ofcontingent demand is that it is either optional or can be managed usingthe concept of backlog, which means that, in the sense of routing, theenterprise can turn contingent demand on and off as desired (as long as,over time, all of it eventually gets handled, although even this may beoptional, as some contingent demand types may be truly optional, whileothers such as customer emails are not).

The key to solving the WFM problem is to ensure that a sufficiently largeinventory of contingent demand exists to allow one to schedule a higher numberof agents, for every time slot, than would be needed for any conceivableimperative demand level. Without the available contingent demand tooccupy those agents not needed for imperative demand, such an approachwould be cost-prohibitive, since it would require “overstaffing” by 10%or more – which would normally cause a staffing manager to be promptlyfired. The key is to have useful, economically value-adding activitiesthat can be managed as contingent demand; “make work” won’t do because itjust wastes money.

To make this clear: if enough contingent demand is available, it ispossible to create a “good enough” forecast of imperative demand,schedule a lot of extra agents at all time slots, and fill in the”difference” between actual imperative demand and actual staffing byallocating appropriate proportions of the available staff to handlecontingent demand. Thus in one simple step (OK, maybe not that simple!),two big, important benefits are obtained.

First, customers of such acontact center would NEVER BE QUEUED UP! As soon as they call in, they geta live agent (if needed). Think what it would be like if your customersnever had to impatiently wait to be served, and without breaking thebudget!

Second, agents would be freed from the shackles of Real-TimeAdherence (RTA)andunnatural work schedules. This is because, with sufficient contingentdemand to handle variations, agents don’t need to be scheduled “just so”to meet the curve, and they don’t need to be held to a ridiculously highlevel of adherence to a schedule. It would be good enough for most peopleto show up about when scheduled, to take their breaks when they want, toeat when (and with whom) they want – because the system will adjustautomatically using contingent demand management as the special sauce.

Again, the secret ingredient that makes it possible to delight customers,agents, and CFOs is the availability of “real” contingent demand (“real”means it adds value and is useful to the enterprise, as opposed to”fictional” demand which isn’t demand at all and doesn’t add value, butjust costs money – something no-one contemplates or would allow!).

Thegreat thing is that, with the rapid emergence of many new contact typesthat can be managed as contingent demand (social media posts, textmessages, mobile apps that are situationally aware, and so forth), it isbecoming easier to provide the requisite contingent demand, and in factintegrating imperative demand and contingent demand provides a means ofbreaking the enslaving tendencies of WFM AND to effectively handle theemergence of many new contact types without batting an eye.

So, if we can step back from the problem, gain a little altitude, andlook at things with a bigger picture perspective, it is possible to dowhat WFM vendors have only dreamed of – to delight customers always (atleast in terms of responding quickly), without breaking the bank. Andbest of all, such an approach makes fewer demands on WFM systems, so youdon’t need all the bells and whistles to do great things.

Give it a think and consider the possibilities!

Brian Galvin
Brian Galvin

Brian joined NewVoiceMedia in July 2011 as Chief Science Officer. He has 30 years experience in technology-focused enterprises including AMD, SAIC and Genesys. He heads NVM Labs, the research arm of NewVoiceMedia.

Deskphone with Vonage logo

Talk to an expert.

UK free phone number: 0330 808 9348