Nemertes 2014 Contact Center TCO Study

17
©Nemertes Research 2014 www.nemertes.com 8882412685 DN3477 1 Contact Center Total Cost of Ownership By Lisa Durant Research Analyst, Nemertes Research Executive Summary Understanding the Total Cost of Ownership (TCO) of contact center alternatives is critical for sound decisionmaking. Nemertes recommends considering capital, implementation, and ongoing operational costs in TCO analysis. Nemertes conducted a research study including 114 organizations (most of which used multiple contactcenter vendors) and gathered 190 data points covering cost data for a number of contactcenter vendors. Five vendors received enough responses for us to analyze them independently: Aspect, Avaya, Cisco, Interactive Intelligence, and NEC. We then used this data to develop a firstyear TCO analysis comparing each vendor to overall aggregate costs. Nemertes analysis shows that capital, implementation, and operation costs vary between vendors, especially for rollouts of fewer than 100 agent licenses. Costs vary based on deployment size, type (onpremises vs. cloud), and product, meaning no single vendor is the most costeffective (defined as offering the lowest TCO) for every situation. For example, Interactive Intelligence provides the lowest firstyear cost (a combination of capital, implementation, and operational metrics) for contact center rollouts of all sizes. But, when just evaluating subsequentyear operational costs for all sizes of rollouts, Avaya comes out on top. However, when looking only at rollouts under 100 licenses, Interactive Intelligence provides both the lowest first and subsequentyear costs. Small contact centers (under 100 agent licenses) are receiving a lot of attention as vendors expand cloud and onpremises offerings aimed primarily at small and midsize customers. Small contact centers generally use similar applications as larger contact centers, including automatic call distributors (ACDs), interactive voice response (IVR), call recording, and workforce optimization (WFO) applications. Although organizations still deploy the majority of small contact centers onpremises, a growing number are using and evaluating private cloud, public cloud, or hybrid (onpremises and cloud combination) environments.

Transcript of Nemertes 2014 Contact Center TCO Study

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

1

Contact  Center  Total  Cost  of  Ownership    

By  Lisa  Durant  Research  Analyst,  Nemertes  Research  

Executive Summary

Understanding  the  Total  Cost  of  Ownership  (TCO)  of  contact  center  alternatives  is  critical  for  sound  decision-­‐making.  Nemertes  recommends  considering  capital,  implementation,  and  ongoing  operational  costs  in  TCO  analysis.    Nemertes  conducted  a  research  study  including  114  organizations  (most  of  which  used  multiple  contact-­‐center  vendors)  and  gathered  190  data  points  covering  cost  data  for  a  number  of  contact-­‐center  vendors.  Five  vendors  received  enough  responses  for  us  to  analyze  them  independently:  Aspect,  Avaya,  Cisco,  Interactive  Intelligence,  and  NEC.  We  then  used  this  data  to  develop  a  first-­‐year  TCO  analysis  comparing  each  vendor  to  overall  aggregate  costs.  Nemertes  analysis  shows  that  capital,  implementation,  and  operation  costs  vary  between  vendors,  especially  for  rollouts  of  fewer  than  100  agent  licenses.      Costs  vary  based  on  deployment  size,  type  (on-­‐premises  vs.  cloud),  and  product,  meaning  no  single  vendor  is  the  most  cost-­‐effective  (defined  as  offering  the  lowest  TCO)  for  every  situation.  For  example,  Interactive  Intelligence  provides  the  lowest  first-­‐year  cost  (a  combination  of  capital,  implementation,  and  operational  metrics)  for  contact  center  rollouts  of  all  sizes.  But,  when  just  evaluating  subsequent-­‐year  operational  costs  for  all  sizes  of  rollouts,  Avaya  comes  out  on  top.  However,  when  looking  only  at  rollouts  under  100  licenses,  Interactive  Intelligence  provides  both  the  lowest  first-­‐  and  subsequent-­‐year  costs.      Small  contact  centers  (under  100  agent  licenses)  are  receiving  a  lot  of  attention  as  vendors  expand  cloud  and  on-­‐premises  offerings  aimed  primarily  at  small  and  midsize  customers.  Small  contact  centers  generally  use  similar  applications  as  larger  contact  centers,  including  automatic  call  distributors  (ACDs),  interactive  voice  response  (IVR),  call  recording,  and  workforce  optimization  (WFO)  applications.  Although  organizations  still  deploy  the  majority  of  small  contact  centers  on-­‐premises,  a  growing  number  are  using  and  evaluating  private  cloud,  public  cloud,  or  hybrid  (on-­‐premises  and  cloud  combination)  environments.    

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

2

The Issue

Nemertes  conducted  an  independent  study  (no  vendors  sponsored  it)  to  evaluate  contact-­‐center  costs.  The  research  documented  capital,  implementation,  and  operational  costs  associated  with  implementing  and  maintaining  contact-­‐center  applications.  We  also  gathered  data  on  factors  such  as  rollout  size,  vendor  selection,  and  type  of  deployment  (cloud  vs.  on-­‐premises)  to  evaluate  how  those  factors  affect  cost.  Nemertes  studied  114  enterprises  and  gathered  190  data  points  providing  detailed  contact-­‐center  cost  data  for  several  vendors.  Five  vendors  received  enough  responses  for  us  to  analyze  them  independently:  Aspect,  Avaya,  Cisco,  Interactive  Intelligence,  and  NEC.  Those  vendors  that  did  not  receive  enough  response  were  rolled  into  an  “Other”  category.  Overall,  we  found  that  costs  varied  most  significantly  for  deployments  under  100  licenses.      Nemertes  recommends  companies  of  all  sizes  evaluate  a  number  of  factors  when  making  contact-­‐center  vendor  and  product  decisions.  Initial  capital  costs  are  important,  but  the  lowest  up-­‐front  price  doesn’t  always  mean  the  highest  value.  And  because  organizations  don’t  know  what  the  ongoing  operational  costs  will  be  until  they  actually  operate  the  system,  it’s  difficult  to  evaluate  such  costs—yet  extremely  important.  The  cost  data  here  provides  that  critical  piece  of  the  puzzle  by  gathering  that  real-­‐world  data  from  organizations  already  using  these  vendors.  We  recommend  IT  leaders  consider  all  costs—with  an  emphasis  on  ongoing  operational  costs—the  proposed  solution’s  technical  capabilities,  existing  relationships  with  the  vendor,  the  solution’s  customer-­‐service  capabilities,  the  solution’s  ability  to  integrate  with  existing  business  applications,  and  referrals  from  existing  customers.  

The Data

Nemertes  analysts  conducted  in-­‐depth  phone  interviews  with  17  IT  professionals  to  establish  a  reasonable  range  for  each  category  of  costs.  We  then  sent  email  invitations  to  a  pre-­‐qualified  list  of  several  hundred  IT  professionals.  The  email  contained  a  link  to  an  electronic  survey  containing  a  subset  of  the  cost  data  questions  asked  during  the  phone  interviews.  We  analyzed  the  survey  data,  eliminating  respondents  whose  cost  entries  fell  outside  of  a  “reasonable”  range  established  by  the  interviews.  Through  this  process,  we  received  97  valid  responses.  Most  companies  used  multiple  contact  center  vendors,  so  we  received  190  data  points  for  contact  center  vendors.    Nemertes  used  the  following  definitions  of  cost  data  for  the  phone  interviews,  online  surveys,  and  subsequent  analysis  and  calculated  costs  per  license  using  the  formulas  noted.    

± Capital  cost  –  Includes  all  hardware,  software,  and  licenses  required  for  contact  center.    

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

3

o The  formula  we  use  is  as  follows:  !"#$%"&  !"#$#

!"#$%&  !"  !"#$%&$&  (!"#ℎ  !"#$%  !"#  !"#$%&'!(%)  

± Implementation  cost  –  Includes  staff  time  and  third-­‐party  consultants  or  integrators.  o The  formula  we  use  is  as  follows:  

 

!"#$$  !"#$  ×  !"!"#"  ℎ!"#$%  !"#$ +  !ℎ!"#  !"#$%  !"#$#!"#$%&  !"  !"#$%&$&  

 ± Operational  cost  –  Includes  staff  time,  equipment  maintenance  costs,  third-­‐

party  managed  services,  training,  and  certification  costs.  We  gathered  four  types  of  operational  data:  o Internal  staff  –  Includes  the  total  loaded  cost  of  internal  staff  (measured  as  

full-­‐time  equivalents)  divided  by  the  number  of  agent  and  supervisor  licenses.  

o Annual  equipment  maintenance  –  Includes  the  amount  that  the  organization  pays  to  the  vendor  or  VAR  for  annual  equipment  maintenance.  

o Third-­‐party  services  –  Includes  any  third-­‐party  partners,  systems  integrators,  or  consultants  who  help  with  ongoing  system  operations.  

o Training  –  Includes  training  costs  for  IT  staff  only  (not  end  users).  The  formula  we  use  is  as  follows:    

!"#$%&  !"  !"#$  ×  !"#$%&#  !""#!$  !"#$%$  !"#"$% +

!"#$%&!'(  !"#$%&$"$'& +!"#"$%&  !"#$%&"! + !"!"#"#$!"#$%&%!'$%()

!"#$%&  !"  !"#$%&$&  

 Using  these  definitions  and  formulas,  Nemertes  calculated  total  costs  per  license  for  each  of  the  below  technologies:    

± Automatic  Call  Distributor  (ACD)  –  A  platform  that  routes  incoming  calls  based  on  internally  set  rules  and  instructions  to  specified  groups  of  agents.  

± Interactive  Voice  Response  (IVR)  –  A  technology  that  enables  customers  calling  into  the  contact  center  to  perform  self  service  or  input  information  for  agents  via  keypad  or  spoken  word.  

± Predictive  Dialing  –  A  tool  that  dials  a  pre-­‐defined  list  of  telephone  numbers  and  then  connects  answered  calls  to  agents.  

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

4

± Skills-­‐Based  Routing  –  A  routing  strategy  used  to  assign  incoming  contact  requests  to  the  most  suitable  agent  rather  than  just  the  next  available  agent.  

± Call  Recording  –  A  technology  that  records  voice  conversations.  ± Virtual  Hold  –  A  technology  that  allows  customers  to  request  and  receive  

callbacks  rather  than  waiting  in  queue  for  an  agent  to  respond.  

Contact-Center Key Findings

First-Year Costs Overall,  the  median  first-­‐year  costs  for  contact  center  are  $3,139  per  license.  Up-­‐front  contact-­‐center  capital  and  implementation  costs  vary  somewhat  between  vendors,  while  ongoing  operational  costs  vary  significantly;  the  median  operational  cost  is  $1,300  per  license,  per  year.  Avaya,  Aspect,  and  Interactive  Intelligence  all  have  operational  costs  that  fall  below  this  median.    Those  using  Interactive  Intelligence  report  the  lowest  median  first-­‐year  cost,  a  combination  of  capital,  implementation,  and  operational  costs,  for  contact-­‐center  rollouts  of  all  sizes  ($2,021  per  license).  (Please  see  Figure  1.)  Though  this  view  gives  a  snapshot  of  all  sizes  of  rollouts,  the  data  isn’t  a  precise  indicator  of  rollout  size  or  contact-­‐center  technology  in  use  as  the  data  that  follows.    

 Figure  1:  Median  Cost  Per  License  for  All  Size  Rollouts    First-­‐year  costs  especially  varied  for  rollouts  comprising  fewer  than  100  licenses.  Overall,  the  median  first-­‐year  cost  for  rollouts  comprising  fewer  than  100  licenses  is  $8,235  per  license;  Interactive  Intelligence’s  first-­‐year  cost  for  the  smaller  rollouts  is  significantly  lower  than  any  other  vendor,  as  well  as  the  median,  at  $6,357  per  license.  The  lowest  capital,  implementation,  and  operational  costs  are  highlighted  in  blue  in  Figure  2.  (Please  see  Figure  2.)      

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -./0.1 -234 -5/625 -3/042789$'# -./.50 -553 -3/2:; -1/6;17)&<& -0/2:; -.20 -4/:36 -1/;55=(8'> -0/54; -53. -4/.03 -;/:43?@= -0/24: -5;. -3/3.3 -.:/142A#B$% -0/24: -4:: -6/::: -3/64:A)$%&++*C$D(&" -0/6:; -52. -5/444 -1/604

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/0127)&<& -3;2 -33 -654 -./::2?@= -0;0 -55 -;4: -./012=(8'> -22: -.;: -426 -./40.789$'# -./46; -642 -.6. -./;:2A#B$% -./.64 -665 -./..: -6/54;E>#&+ -1:: -.43 -002 -./6;0F!"#$%&'#()$*!"#$++(,$"'$*-(-*".#*%$'$()$*$"./,0*1&23+$*("*#0(1*'&#$,.%4

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -135 -446 -3:4 -6/:6.789$'# -./5:1 -61: -234 -6/5407)&<& -6/.2. -.53 -404 -6/14.?@= -;.2 -.21 -6/625 -0/03;=(8'> -6/:;3 -002 -./0:3 -0/201A#B$% -./4.: -501 -./4:5 -0/546E>#&+ -./413 -640 -./0:: -0/.0;

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*L*.::*H('$"8$8

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*M*.::*H('$"8$8

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*7++*N(O$8

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

5

 Figure  2:  Median  Cost  Per  License  for  Rollouts  <  100  Licenses    As  stated,  the  data  becomes  more  meaningful  and  precise  as  we  evaluate  it  by  rollout  size  or  application.  Over  time,  low  operational  costs  will  negate  higher,  one-­‐time  upfront  capital  and  implementation.  We  also  find  that  organizations  sometimes  spend  more  with  staff  and  third-­‐party  partners  on  the  implementation  to  get  everything  functioning  near-­‐perfectly  right  from  the  start,  which  ultimately  improves  ongoing  operational  costs.    The  overall  first-­‐year  cost  for  rollouts  of  more  than  100  licenses  was  fairly  consistent  between  vendors,  with  a  $1,293  average  first-­‐year  cost  per  license.  However,  operational  cost  for  these  rollouts  still  varied  significantly.  Avaya  has  the  lowest  first-­‐year  cost  ($1,007  per  license)  for  rollouts  of  more  than  100  licenses.  The  lowest  capital,  implementation,  and  operational  costs  for  rollouts  over  100  licenses  are  highlighted  in  blue  in  Figure  3.  (Please  see  Figure  3.)    

 Figure  3:  Median  Cost  Per  License  for  Rollouts  >  100  Licenses    Contact-Center Channels Contact  centers  of  all  sizes  are  now  engaging  more  than  just  voice;  they  are  enabling  channels  like  email  and  real-­‐time  chat.  Figure  4  shows  the  channels  most  commonly  monitored  in  contact  centers  of  all  sizes.    

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -./0.1 -234 -5/625 -3/042789$'# -./.50 -553 -3/2:; -1/6;17)&<& -0/2:; -.20 -4/:36 -1/;55=(8'> -0/54; -53. -4/.03 -;/:43?@= -0/24: -5;. -3/3.3 -.:/142A#B$% -0/24: -4:: -6/::: -3/64:A)$%&++*C$D(&" -0/6:; -52. -5/444 -1/604

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/0127)&<& -3;2 -33 -654 -./::2?@= -0;0 -55 -;4: -./012=(8'> -22: -.;: -426 -./40.789$'# -./46; -642 -.6. -./;:2A#B$% -./.64 -665 -./..: -6/54;E>#&+ -1:: -.43 -002 -./6;0F!"#$%&'#()$*!"#$++(,$"'$*-(-*".#*%$'$()$*$"./,0*1&23+$*("*#0(1*'&#$,.%4

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -135 -446 -3:4 -6/:6.789$'# -./5:1 -61: -234 -6/5407)&<& -6/.2. -.53 -404 -6/14.?@= -;.2 -.21 -6/625 -0/03;=(8'> -6/:;3 -002 -./0:3 -0/201A#B$% -./4.: -501 -./4:5 -0/546E>#&+ -./413 -640 -./0:: -0/.0;

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*L*.::*H('$"8$8

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*M*.::*H('$"8$8

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*7++*N(O$8

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -./0.1 -234 -5/625 -3/042789$'# -./.50 -553 -3/2:; -1/6;17)&<& -0/2:; -.20 -4/:36 -1/;55=(8'> -0/54; -53. -4/.03 -;/:43?@= -0/24: -5;. -3/3.3 -.:/142A#B$% -0/24: -4:: -6/::: -3/64:A)$%&++*C$D(&" -0/6:; -52. -5/444 -1/604

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/0127)&<& -3;2 -33 -654 -./::2?@= -0;0 -55 -;4: -./012=(8'> -22: -.;: -426 -./40.789$'# -./46; -642 -.6. -./;:2A#B$% -./.64 -665 -./..: -6/54;E>#&+ -1:: -.43 -002 -./6;0F!"#$%&'#()$*!"#$++(,$"'$*-(-*".#*%$'$()$*$"./,0*1&23+$*("*#0(1*'&#$,.%4

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -135 -446 -3:4 -6/:6.789$'# -./5:1 -61: -234 -6/5407)&<& -6/.2. -.53 -404 -6/14.?@= -;.2 -.21 -6/625 -0/03;=(8'> -6/:;3 -002 -./0:3 -0/201A#B$% -./4.: -501 -./4:5 -0/546E>#&+ -./413 -640 -./0:: -0/.0;

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*L*.::*H('$"8$8

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*M*.::*H('$"8$8

C$D(&"*=>8#8*G$%*H('$"8$I*J>++>K#8*7++*N(O$8

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

6

 Figure  4:  Channels  Monitored  in  the  Contact  Center    All  vendors  supply  a  solution  that  supports  the  most  common  channels  including  e-­‐mail  and  real-­‐time  chat;  however,  prices  differ  based  on  how  vendors  bundle  and  sell  channels  for  specific  solutions.  IT  decision  makers  should  consider  these  pricing  differences  when  deciding  on  a  specific  solution  and  vendor.    Contact Center Applications ACDs  (66%),  IVRs  (55%),  and  call  recording  (53%)  are  the  most  common  applications  used  by  contact  centers  of  all  sizes.  Workforce  optimization  (47%)  and  skills-­‐based  routing  (45%)  also  are  used  fairly  frequently.  (Please  see  Figure  5.)    

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

7

 Figure  5:  Contact  Center  Applications    The  average  first-­‐year  cost  for  an  ACD,  the  most  commonly  used  contact  center  application,  is  $966  per  license;  the  average  ongoing  operational  cost  for  an  ACD  is  $288  per  license.  Interactive  Intelligence  provides  the  lowest  first-­‐year  and  ongoing  operational  costs  for  ACD  deployments  across  all  size  rollouts.  Figure  6  shows  how  ACD  capital,  implementation,  and  operational  costs  differentiate  between  vendors;  the  lowest  costs  for  each  category  are  highlighted  in  blue.  (Please  see  Figure  6.)      

 Figure  6:  ACD  Median  Cost  Per  License  (All  Size  Rollouts)  

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

8

 The  median  first-­‐year  cost  for  an  IVR  is  $884  per  license,  with  the  ongoing  operational  cost  at  $271  per  license.  Avaya  had  both  the  lowest  average  first-­‐year  and  ongoing  operational  costs  for  IVR.  The  lowest  capital,  implementation,  and  operational  costs  for  IVR  are  highlighted  in  blue  in  Figure  7.  (Please  see  Figure  7.)    

 Figure  7:  IVR  Median  Costs  Per  License  (All  Size  Rollouts)    The  median  first-­‐year  cost  for  call  recording  is  $766  per  license,  and  the  median  ongoing  operational  cost  is  $288  per  license.  Interactive  Intelligence  has  the  lowest  first-­‐year  ($401  per  license)  and  ongoing  operational  costs  ($77  per  license)  for  call  recording.  The  lowest  capital,  implementation,  and  operational  costs  for  call  recording  are  highlighted  in  blue  in  Figure  8.  (Please  see  Figure  8.)    

 Figure  8:  Call  Recording  Median  Costs  Per  License  (All  Size  Rollouts)  

 The  same  vendor  does  not  necessarily  provide  these  contact  center  applications.  Only  22%  of  contact  centers  use  only  one  vendor.  (Please  see  Figure  9).          

!"#$%& '()*+(, -.),"."#+(+*%# /)"&(+*%#(, 0/012!"#$%&'#()$*!"#$++(,$"'$ -.// -.0. -11 -23.4(5'6 -237 -89 -/37 -113:)&;& -8/2 -28 -<11 -987=>4 -171 -/1 -.?1<8 -<?8<0@#A$% -818 -.<< -..1 -9.2@)$%&++*B$C(&" -2<0 -83 -<99 -177

4&++*D$'6%C(",*E$C(&"*465#5*F$%*G('$"5$H*:++*I(J$*D6++6K#5

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

9

   The  number  of  vendors  used  within  a  contact  center  also  differentiated  based  on  deployment  size.  Figure  10  shows  the  percentage  of  companies  using  single  or  multiple  vendors  broken  down  by  deployment  size.      

Figure  10:  Number  of  Vendors  and  Deployment  Size    The  choice  to  use  one  or  multiple  vendors  to  provide  contact  center  applications  affects  total  cost  of  ownership.  For  example,  the  median  first-­‐year  cost  of  ownership  for  an  Avaya  ACD  alone  is  $1,043  per  license,  and  the  median  first-­‐year  cost  of  

Number of Vendors in Contact Center

Figure  9:  Number  of  Vendors  in  Contact  Center  

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

10

ownership  for  an  Avaya  IVR  is  $773  per  license.  However,  Nemertes  found  that  the  median  first-­‐year  cost  of  ownership  for  an  Avaya  ACD  and  IVR  in  combination  is  $1,677  per  license  (a  difference  of  $139  per  license).  This  is  similar  for  other  vendors  as  well.    Deployment Model Most  vendors  currently  offer  contact  center  solutions  both  on-­‐premises  and  in  the  cloud;  many  also  offer  hybrid  models  combining  both  on-­‐premises  and  cloud  solutions.    

 On  average,  53.8%  of  participants  deploy  contact  center  solutions  on-­‐premises;  41.4%  deploy  contact  center  solutions  through  public  cloud,  private  cloud,  or  hybrid  (on-­‐prem/cloud  combination)  models.  (Please  see  Figure  11.)      Of  those  deploying  SaaS/public  cloud,  72%  are  small  contact  centers  (<100  licenses).    The  majority  of  most  vendors’  customers  are  using  on-­‐premises  solutions  though  most  vendors  offer  both  cloud  and  on-­‐premises  solutions.  However,  80%  of  Interactive  Intelligence  customers  and  64%  of  NEC  customers  are  using  public  cloud,  private  cloud,  or  hybrid  solutions.  Figure  12  shows  deployment  type  by  vendor.  

 

Deployment Method

Figure  11:  Deployment  Method  

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

11

 Figure  12:  Deployment  Type  by  Vendor  

 Interactive  Intelligence  and  NEC’s  higher  number  of  cloud  customers  reflects  the  current  direction  of  the  contact  center  industry.  Cloud  and  hybrid  contact  centers  are  gaining  more  traction,  especially  with  smaller  contact  centers.      Deployment Length Another  aspect  to  consider  when  evaluating  contact  center  solutions  and  deployment  type  is  the  time  it  takes  to  deploy  the  solution  and  how  that  fits  in  with  company  needs  and  timelines.  Lengthy  deployments  can  also  affect  costs  should  a  company  need  to  pay  specific  individuals  to  manage  implementation.  For  example,  ACD  deployment  time  differentiates  depending  on  deployment  type.  Nemertes  found  that  55.6%  of  those  deploying  SaaS/public  cloud  ACDs  stated  that  deployment  took  12–24  months;  39%  stated  that  it  took  less  than  12  months  to  deploy  their  SaaS  ACD  solution.  Figure  13  shows  length  of  deployment  and  deployment  type.  (See  Figure  13.)    

!"#$%& '#()&"*+,"-..-/0)123+40

43%1$ 562&+$7&+8.9"0:3%1$

;,)"49 !"#!$ %#%$ &#'$ %#%$;8.6. !%#($ )#&$ *#*$ ''#*$:+,4% )*#"$ '%#'$ '&#%$ '"#!$<#9"&.49+8"0<#9"33+="#4" '%#%$ )%#%$ "%#%$ '%#%$>?: +!#%$ !#,$ ")#&$ "&#($'9@"& ,,#%$ ,#%$ '(#%$ '(#%$;8"&.=" )+#*$ &#!$ '!#"$ ',#)$

*Where total does not add up to 100%, the balance are “unsure”

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

12

 Figure  13:  Length  of  Deployment  and  Deployment  Type

Conclusion and Recommendations

Calculating  a  solid  business  case  for  contact  center  implementations  isn’t  only  table  stakes  –  it’s  also  job  security.  Without  a  compelling,  functional  business  case  based  on  solid  data,  IT  staffs  risk  project  cancelations  during  times  of  budget  cuts.  It  is  important  for  companies  of  all  sizes  to  consider  all  cost  areas  when  making  contact-­‐center  deployment  decisions:  capital,  implementation,  and  ongoing  operational.  Nemertes’  research  indicates  that  capital  and  implementation  costs  are  highly  negotiable,  and  enterprise  IT  leaders  should  aggressively  deal  with  vendors  to  reduce  these  costs.  Operational  costs  are  more  difficult  to  identify  but  have  greater  impact  on  total  cost  of  ownership.  IT  leaders  should  invest  the  necessary  time  and  care  to  make  sure  that  they  understand  these  costs.  We  also  recommend  the  following:  

± Consider  all  deployment  models,  including  public  cloud  (SaaS)  and  hybrid.  Cloud  contact-­‐center  solutions  may  provide  cost  savings,  can  also  stabilize  costs,  allow  for  fluctuations  in  staff  size,  and  make  expenses  more  predictable.  

± Consider  the  features  and  channels  that  each  solution  supports,  and  evaluate  how  the  solution  integrates  with  existing  systems.  Proven  integration  will  reduce  implementation  time,  implementation  costs,  and  ongoing  operational  costs  

± Evaluate  how  channels  are  bundled  and  sold  when  seeking  a  solution  for  a  multichannel  environment.  Purchasing  only  specific  channels  most  important  

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

13

to  the  company  rather  than  pre-­‐bundled  channels  that  will  not  be  used  could  reduce  costs.  

± Consider  the  time  it  takes  to  deploy  and  implement  a  solution.  Lengthy  deployment  times  can  affect  costs  –  especially  if  the  company  plans  to  pay  someone  to  handle  implementation.  

± Evaluate  how  combining  common  contact  center  products  and  functions  could  reduce  costs.  Purchasing  some  applications  together  from  one  vendor  can  reduce  costs.  

± Demand  and  contact  referrals  after  creating  a  short  list  of  two  to  three  vendors  to  consider.  This  direct  contact  with  actual  users  will  enable  IT  leaders  to  better  understand  capital,  implementation,  and  ongoing  operational  costs.  

± Evaluate  operational  costs,  which  can  vary  widely  between  vendors  and  deployment  models.  Often,  leveraging  existing  staff  expertise  with  a  specific  vendor  can  reduce  operational  costs.  

   

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

14

Contact Center Cost Study Methodology

For  this  Nemertes  Contact  Center  Cost  Study  Research  Project,  analysts  conducted  in-­‐depth  phone  interviews  with  17  IT  professionals.  Nemertes  also  conducted  several  short  follow-­‐up  calls  or  exchanged  emails  to  clarify  and  augment  data.  During  the  interviews,  each  analyst  asked  a  pre-­‐planned  list  of  questions  to  ensure  that  they  asked  the  questions  consistently.  Many  questions  are  open-­‐ended,  providing  an  opportunity  for  participants  to  provide  their  own  unbiased  insight  and  observations.      After  completing  the  interviews,  we  established  reasonable  ranges  for  each  category  of  costs.  We  then  sent  email  invitations  to  a  prequalified  list  of  IT  professionals.  The  email  contained  a  link  to  an  electronic  survey  with  a  subset  of  the  cost  data  questions  asked  during  the  interviews.  We  analyzed  the  survey  data,  eliminating  respondents  whose  cost  entries  fell  outside  a  “reasonable”  range  established  by  the  interviews.  We  received  97  valid  responses  and  most  companies  used  more  than  one  contact  center  provider.  We  received  190  data  points  for  contact  center  providers.  To  ensure  the  report  is  relevant  to  the  largest  possible  group  of  readers,  we  deliberately  sought  to  reach  the  broadest  possible  range  of  industries  and  company  sizes.    To  determine  what  participants  were  including  in  their  contact  center  initiatives,  we  asked  open-­‐ended  questions  about  contact  center  usage,  plans  and  goals,  and  costs.  The  specific  technologies  included  ACD,  IVR,  call  recording,  workforce  optimization,  skills-­‐based  routing,  virtual  hold,  predictive  dialing,  and  voice  analytics.  We  also  asked  demographic  questions,  including  number  of  employees,  annual  revenue,  job  titles,  and  IT  budgets.  As  a  result,  individual  interviews  and  surveys  varied  considerably  in  the  number  of  questions  answered  and  in  the  number  of  subject  areas  addressed,  as  well  as  in  the  degree  of  insight  provided  for  each  contact  center  topic  covered,  based  on  the  interest  and  expertise  of  the  participants  involved.        For  the  interviews  and  surveys,  Nemertes  drew  participants  from  its  database  of  IT  professionals,  its  non-­‐vendor  client  base,  and  to  a  lesser  extent,  from  publicly  available  lists  of  IT  executives  and  published  case  studies.  Nemertes  gave  all  vendors  equal  opportunity  to  provide  customer  names.    Nemertes  guarantees  confidentiality  and  anonymity  for  participants  and  their  companies.  Any  reports  or  slides  generated  from  this  data  include  quotations  from  real  individuals,  identified  only  by  title  and/or  industry  affiliation.  Please  note  these  quotes  are  verbatim,  with  no  changes  in  content  or  wording,  except  to  correct  grammar.    

 Timeframe We  conducted  interviews  and  surveys  with  contact  center  cost  study  participants  from  IT  organizations  between  September  and  December  2013.  We  asked  

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

15

participants  to  provide  us  with  insight  into  ongoing  contact  center  initiatives  and  those  planned  for  the  next  two  years.    

Participants

In  selecting  individual  participants,  we  asked  to  speak  with  the  individual  or  individuals  within  IT  most  closely  associated  with  decision-­‐making,  operations  and  overall  knowledge  in  the  contact  center  area.     By Title In  this  study,  Nemertes  gathered  data  from  a  wide  range  of  decision  makers/influencers.  Directors  of  IT,  CIOs,  and  CTOs  represent  the  bulk  of  research  participants  (51.6%).  

 Figure  14:  Participants  by  Title     By Industry As  noted,  we  sought  to  include  the  broadest  possible  range  of  industries  in  our  research.  Manufacturing,  financial  services,  and  professional  services  (broadly  covering  all  professional  services  sub-­‐areas,  including  engineering,  accounting,  law  firms,  etc.)  top  the  list  of  industries  represented,  accounting  for  25.3%,  17.4%,  and  11.6%  of  participants  respectively.  We  also  spoke  to  relatively  high  percentages  in  healthcare  (10.5%),  high  tech  (6.8%),  and  retail  (5.8%).    

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

16

 Figure  15:  Participants  by  Industry  

By Size: Number of Employees We  seek  insight  from  large  and  small  organizations.  For  benchmark  analysis,  we  characterize  companies  as  being  small,  midsize,  and  large  by  several  measures,  including  employee  size.    Our  size  characterization  based  on  employee  size  is  as  follows:  

• Small:    0-­‐250  • Midsize:    251-­‐2500  • Large:    >2500  

 A  total  of  5.8%  of  the  participating  companies  are  small.  There  is  a  fairly  even  split  between  large  (46.8%)  and  midsize  (47.4%)  companies.    

 Although  this  differs  from  the  “traditional”  business  demographics,  in  which  small  businesses  make  up  the  largest  percentage  of  total  businesses,  we  do  talk  to  many  innovative  small  and  large  organizations,  so  a  solid  coverage  of  all  types  of  companies  is  imperative.    

Participants by Industry

Company Size by Number of Employees

Figure  8:  Participating  Company  Size

   

©Nemertes  Research  2014±  www.nemertes.com    ±  888-­‐241-­‐2685  ±DN3477    

17

By IT Culture We  asked  interview  participants  to  describe  the  IT  culture  or  how  the  business  views  IT.  Is  IT  a  strategic  differentiator,  and  how  rapidly  do  they  deploy  new  technologies?  Nearly  one-­‐third  (32.1%)  say  their  IT  culture  is  highly  strategic  (bleeding  edge),  22.1%  say  aggressive,  29.5%  say  moderate,  and  16.3%  conservative.   For Further Information Nemertes  has  thousands  of  charts,  correlation  points  and  data  analysis  for  this  and  numerous  other  topics.  Though  Nemertes  segments  data  using  the  cut  points  noted  in  this  Methodology  statement,  we  can  cut  

and  correlate  data  using  any  numbers  (so,  for  example,  we  can  segment  or  correlate  data  using  only  companies  with  fewer  than  100  employees,  or  only  in  the  financial-­‐services  industry).  

   Nemertes  uses  this  research  data  in  hundreds  of  more  detailed  papers,  presentations,  strategy  sessions  and  Webcasts.  We  also  rely  on  the  findings  to  assist  in  consulting  projects,  as  well  as  conversations  with  our  clients  about  various  technology  and  business  initiatives.  

 If  you  have  further  questions  about  our  methodology,  please  contact  [email protected].  Clients,  please  contact  client-­‐[email protected]  for  any  assistance.  Those  interested  in  engaging  with  Nemertes,  please  contact  [email protected].      

   

About  Nemertes  Research:  Nemertes  Research  is  a  research-­‐advisory  and  strategic-­‐consulting  firm  that  specializes  in  analyzing  and  quantifying  the  business  value  of  emerging  technologies.  You  can  learn  more  about  Nemertes  Research  at  our  Website,  www.nemertes.com,  or  contact  us  directly  at  [email protected].  

 

Participants by Company IT Culture

Figure  9:  Participants  by  Company  IT  Culture