applying text mining & analytics to consumer satisfaction surveys - case study

3
2040 14th Suite 200 Boulder, CO 80302 720-259-3060 http://www.collectiveintellect.com Our Client’s Goals Collective Intellect (CI) worked with a leading Fortune 500 hospitality company to analyze their customer satisfaction surveys. Our client understood a portion of their customers were unhappy with their hotel rooms; they just didn’t know why. Existing client research methodologies and analytic systems had failed to uncover the reasons customers were dissatisfied, so CI was tasked with surfacing more precise causation for customer complaints. Results Applying sophisticated language modeling, CI was able to isolate negative guest sentiment to specific areas in the hotel room and property. Collective Intellect utilizes Latent Semantic Analysis (LSA) and Natural Language Processing (NLP) to deliver highly accurate analysis and associated observations from unstructured data. Whether it’s social, email, chat transcripts or surveys, and if data is text then CI possesses the capability to run analysis. Project Highlights Goal – Analyze customer surveys to better understand consumer dissatisfaction with hotel room features. Results – Our semantic and text analytics technology was able surface precise insights into why customers did not enjoy staying in their hotel room.

Upload: collective-intellect

Post on 10-Mar-2015

64 views

Category:

Documents


0 download

DESCRIPTION

Collective Intellect (CI) worked with a leading Fortune 500 hospitality company to analyze their customer satisfaction surveys. Our client understood a portion of their customers were unhappy with their hotel rooms; they just didn’t know why.

TRANSCRIPT

Page 1: Applying Text Mining & Analytics to Consumer Satisfaction Surveys - Case Study

 2040 14th Suite 200 Boulder, CO 80302 720-259-3060 http://www.collectiveintellect.com

 

Our  Client’s  Goals  Collective  Intellect  (CI)  worked  with  a  leading  Fortune  500  hospitality  company  to  analyze  their  customer  satisfaction  surveys.  Our  client  understood  a  portion  of  their  customers  were  unhappy  with  their  hotel  rooms;  they  just  didn’t  know  why.    Existing  client  research  methodologies  and  analytic  systems  had  failed  to  uncover  the  reasons  customers  were  dissatisfied,  so  CI  was  tasked  with  surfacing  more  precise  causation  for  customer  complaints.    Results  Applying  sophisticated  language  modeling,  CI  was  able  to  isolate  negative  guest  sentiment  to  

specific  areas  in  the  hotel  room  and  property.    Collective  Intellect  utilizes  Latent  Semantic  Analysis  (LSA)  and  Natural  Language  Processing  (NLP)  to  deliver  highly  accurate  analysis  and  associated  observations  from  unstructured  data.  Whether  it’s  social,  email,  chat  transcripts  or  surveys,  and  if  data  is  text  then  CI  possesses  the  capability  to  run  analysis.      

Project  Highlights  

Goal  –  Analyze  customer  surveys  to  better  understand  consumer  dissatisfaction  with  hotel  room  features.    

Results  –  Our  semantic  and  text  analytics  technology  was  able  surface  precise  insights  into  why  customers  did  not  enjoy  staying  in  their  hotel  room.  

Page 2: Applying Text Mining & Analytics to Consumer Satisfaction Surveys - Case Study

 2040 14th Suite 200 Boulder, CO 80302 720-259-3060 http://www.collectiveintellect.com

 How  CI  Approached  the  Project  

Working  with  our  client,  CI  first  determined  the  status  of  their  existing  analytics  effort.    Together,  we  concluded  that  CI’s  unique  approach  to  how  data  is  handled,  categorized  and  measured  for  relevancy  could  uncover  additional  insights  their  existing  systems  and  methodologies  were  unable  to  

surface.    

Using  CI’  robust  categorization  engine,  we  were  able  to  isolate  specific  attributes  from  consumer  responses  related  to  dissatisfaction  with  their  hotel  rooms  and  other  features  of  the  property  such  as  the  spa  and  onsite  restaurants.  Thereafter,  we  were  then  able  to  identify  and  dissect  specific  root  causes  for  dissatisfaction  such  as  cost,  convenience,  and  comfort.    

Suggested  Next  Steps  CI  recommended  that  its  hospitality  client  automate  their  analysis  of  customer  

surveys  in  an  “always  on”  dashboard  for  real-­‐time  tracking.  This  approach  would  ensure  that  consumer  reaction  to  changes  could  be  monitored  and  tracked.  We  also  recommended  that  the  resulting  information  be  integrated  into  an  existing  CRM  system,  including  mapping  surveys  results  to  any  social  media  profiles.  

The  CI  Semantic  Methodology  CI  utilizes  semantic  technology  based  on  Latent  Semantic  Analysis  (LSA)  to  expose  latent  contextual-­‐meaning  within  a  large  body  of  text.    Sophisticated  language  processing  is  the  critical  foundation  for  successful  analysis.  Accurate  insights  from  text  analytics  that  are  repeatable,  scalable  and  consistent  can  

Page 3: Applying Text Mining & Analytics to Consumer Satisfaction Surveys - Case Study

 2040 14th Suite 200 Boulder, CO 80302 720-259-3060 http://www.collectiveintellect.com

only  be  achieved  using  technology  that  is  able  to  categorize  enormous  volumes  of  structured  and  unstructured  data  via  dimension  such  as  consideration  and  preference,  intent,  demographics  and  sentiment,  just  to  name  a  few.  

CI  Services  &  Tools  Used  on  Project  CI:Insight  -­  multi-­‐dimensional  analysis,  blended  qualitative  and  quantitative  analysis,  demographics  and  psychographics  analysis,  in  an  integrated  consumer  conversation  analytical  hub  intended  for  Power  Analyst  functions  and  accessed  via  SaaS  model.