SlideShare uma empresa Scribd logo
1 de 5
Baixar para ler offline
Page 1
 

Mathematics, Statistics, and Sales Chat
A Web Retailer Case Study

 
 
Page 2
 

Introduction
With the coming of age of web as a
mainstream sales and marketing channel,
companies have invested substantial
resources in enhancing their web presence.
This includes large investments in web
dvertising. In addition, companies are
looking for ways to improve sales conversion
and customer experience for web shoppers.
Sales chat is a medium that can provide a lift
in both these areas. With the growing
popularity of chat as a communication
medium, particularly among the new
generation of consumers, potential for
revenue generation from this channel is
enormous.
The obvious analogy is to consider the sales
chat agents as a “virtual sales force” for a”
virtual store”. However, a key difference
exists. In a real store there are relatively few
visitors and a significant fraction of the
visitors come with intent to buy i.e. they are
“hot” prospects. On the other hand, major
web stores such as Amazon, eBay and
Overstock have millions of visitors every
week and an overwhelming majority of these
visitors do not intend to buy. They also have
the ability to switch from one store to another
at the click of a mouse button. Considering
the visitor volumes and the low average
likelihood to buy, it is not profitable to
randomly engage in chat with every visitor. It
is imperative to identify a subset of this
visitor
population that has a substantially greater
likelihood to buy. The following case study
presents applications of
statistical/mathematical models in identifying
“hot” prospects and improving conversion,
revenue generation and customer
experience for a major web retailer.

 
Page 3 

The Conversion Funnel
The starting point in understanding and
optimizing the performance of this virtual
sales force is the Conversion Funnel (Figure
1). The funnel helps visualize the size of the
opportunity. Figure 1 represents the funnel
for the web retailer in a particular week.
Layers 1 & 2 are filters, i.e. these are
determined and controlled by the retailers,
while layers 3 to 6 are leakages that are
essentially decisions made by the
customer during the browsing/buying
process and are not in the retailer’s control.
The science is essentially in determining the
appropriate filters to apply in selecting the
right customers and matching them up with
the right agents to minimize the leakages in
the funnel.

Statistical Scoring Model –
Filter (Layer 1)

 

The first filter identifies the “hot leads” i.e. the
people most likely to purchase via chat. In
particular, this identifies customers who have
a significantly higher likelihood of purchasing
from a chat agent than on their own in a selfservice mode. This is an important factor
since self-service is obviously a lower cost
channel than chat and if a customer is very
likely to purchase via self-service then the
business case for inviting them to a chat
engagement is poor. To avoid
cannibalization of a cheaper channel A/B
tests are conducted on a regular basis where
a fraction of the “hot leads” are not invited to
chat and their self-service conversion rates
are compared to conversion rates of the
remaining “hot leads” who are invited to chat.
Typically, conversion rate for chat
engagements among this “hot lead”
population is substantially higher (5x-10x)
than that for self-service engagements.
Identification of hot leads is accomplished
using a statistical scoring model. The scoring
can be done in real time while the
visitor/prospective customer is browsing on
the website. The scoring is based on a
Page 4 

number of attributes including time of the
day, day of the week, geographical location
of the customer, product category, exhibited
behavior on the web site etc. Figure 2
schematically illustrates the scoring model.
Essentially certain patterns of behavior
exhibit a much greater propensity to buy
than others.
The scoring model essentially estimates a
probability of purchase (P(sale)). Statistical
and Data Mining techniques such as Naïve
Bayes, Logistic Regression or Neural
Networks are used to develop these scores.
A threshold score can be set above which
customers are invited to a chat. Based on
variations in traffic and availability of agents
the threshold score can be modified. As
more data is generated, the system learns
and the scoring model becomes better at
identifying the hottest prospects.

 

Agent Optimization – Filter
(Layer 2)
Once the customers are scored and the “hot
leads” identified, the next step is to invite
these “hot leads” to a chat. The number of
“hot leads” invited is based on tactical and
strategic considerations. On the tactical
front, it depends on several factors such as
the number of agents available, acceptable
abandonment rate without significantly
affecting customer experience, average
handle time and concurrency (how many
chats can an agent handle at a time). This is
a routine scheduling problem.
The more interesting strategic problem is to
determine the right number of agents to
maximize profits. The scoring model only
Page 5 

prioritizes the visitors to the site. It does not
automatically provide a threshold score for
the “hot leads”. We provide the threshold
score. This in turn determines the number of
people invited to chat which establishes our
agent staffing levels. But what is the right
threshold score? How is it determined?
The customers are being prioritized based
on how “hot” they are. Based on the average
order value for a given product type and the
probability of a given “hot lead” to buy, the
expected revenue from the transaction can

be calculated. To increase the number of
chats and hence the overall revenues, we
lower the threshold score inviting less
qualified leads, at the same time increasing
the number of agents.
These less qualified customers on an
average generate lower revenues per
customer i.e. the marginal revenue of these
customers is lower. This implies, as we keep
adding agents to interact with less and less
qualified leads, the marginal profit generated
by the additional agents keeps declining. We
keep lowering the threshold, increasing the
number of “hot leads” and adding agents till
we stop making a marginal profit. To
estimate the number of agents
corresponding to this, we use an
optimization algorithm

Finally, scoring techniques are also used to
match the right agent to the customer. The
essential concept is displayed in Figure 3
where we see that the agent Raymond is as

 

such a top performer but is particularly
skilled in selling Electronics products. We
score various product-agent combinations
and manage our chat queues and routing
based on not just the overall performance of
the agent but also on historical performance
in various product categories. The goal here
is not just to look at product-agent
combinations but to develop a
comprehensive scoring model that scores
the agent for a set of customer/product
attributes and determines the best agent to
talk to a given “hot lead.”

Conclusion
Internet chat is a growing channel for sales
over the web and retailers are adding this
capability to their websites. However, like in
self-service web retailing, success in driving
up sales chat revenues and profitability will
go to players who use advanced data-driven
approaches to drive customer intelligence
and chat engagement decisions.

References
ITSMA and PAC, How Customers Choose
Study, 2009
ITSMA and PAC, How Customers Choose
Study, 2009
ITSMA and PAC, How Customers Choose
Study, 2009
W_MATH_1012

Mais conteúdo relacionado

Último

Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...
Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...
Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...panagenda
 
ERP Contender Series: Acumatica vs. Sage Intacct
ERP Contender Series: Acumatica vs. Sage IntacctERP Contender Series: Acumatica vs. Sage Intacct
ERP Contender Series: Acumatica vs. Sage IntacctBrainSell Technologies
 
Hyatt driving innovation and exceptional customer experiences with FIDO passw...
Hyatt driving innovation and exceptional customer experiences with FIDO passw...Hyatt driving innovation and exceptional customer experiences with FIDO passw...
Hyatt driving innovation and exceptional customer experiences with FIDO passw...FIDO Alliance
 
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...FIDO Alliance
 
2024 May Patch Tuesday
2024 May Patch Tuesday2024 May Patch Tuesday
2024 May Patch TuesdayIvanti
 
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptxHarnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptxFIDO Alliance
 
Breaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdfBreaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdfUK Journal
 
ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...
ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...
ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...FIDO Alliance
 
Structuring Teams and Portfolios for Success
Structuring Teams and Portfolios for SuccessStructuring Teams and Portfolios for Success
Structuring Teams and Portfolios for SuccessUXDXConf
 
The Metaverse: Are We There Yet?
The  Metaverse:    Are   We  There  Yet?The  Metaverse:    Are   We  There  Yet?
The Metaverse: Are We There Yet?Mark Billinghurst
 
Introduction to FIDO Authentication and Passkeys.pptx
Introduction to FIDO Authentication and Passkeys.pptxIntroduction to FIDO Authentication and Passkeys.pptx
Introduction to FIDO Authentication and Passkeys.pptxFIDO Alliance
 
1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT
1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT
1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPTiSEO AI
 
Oauth 2.0 Introduction and Flows with MuleSoft
Oauth 2.0 Introduction and Flows with MuleSoftOauth 2.0 Introduction and Flows with MuleSoft
Oauth 2.0 Introduction and Flows with MuleSoftshyamraj55
 
WebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM PerformanceWebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM PerformanceSamy Fodil
 
Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...
Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...
Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...ScyllaDB
 
Intro to Passkeys and the State of Passwordless.pptx
Intro to Passkeys and the State of Passwordless.pptxIntro to Passkeys and the State of Passwordless.pptx
Intro to Passkeys and the State of Passwordless.pptxFIDO Alliance
 
Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...
Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...
Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...FIDO Alliance
 
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdfThe Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdfFIDO Alliance
 
Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024Hiroshi SHIBATA
 
FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...
FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...
FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...FIDO Alliance
 

Último (20)

Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...
Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...
Easier, Faster, and More Powerful – Alles Neu macht der Mai -Wir durchleuchte...
 
ERP Contender Series: Acumatica vs. Sage Intacct
ERP Contender Series: Acumatica vs. Sage IntacctERP Contender Series: Acumatica vs. Sage Intacct
ERP Contender Series: Acumatica vs. Sage Intacct
 
Hyatt driving innovation and exceptional customer experiences with FIDO passw...
Hyatt driving innovation and exceptional customer experiences with FIDO passw...Hyatt driving innovation and exceptional customer experiences with FIDO passw...
Hyatt driving innovation and exceptional customer experiences with FIDO passw...
 
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
 
2024 May Patch Tuesday
2024 May Patch Tuesday2024 May Patch Tuesday
2024 May Patch Tuesday
 
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptxHarnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
 
Breaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdfBreaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdf
 
ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...
ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...
ASRock Industrial FDO Solutions in Action for Industrial Edge AI _ Kenny at A...
 
Structuring Teams and Portfolios for Success
Structuring Teams and Portfolios for SuccessStructuring Teams and Portfolios for Success
Structuring Teams and Portfolios for Success
 
The Metaverse: Are We There Yet?
The  Metaverse:    Are   We  There  Yet?The  Metaverse:    Are   We  There  Yet?
The Metaverse: Are We There Yet?
 
Introduction to FIDO Authentication and Passkeys.pptx
Introduction to FIDO Authentication and Passkeys.pptxIntroduction to FIDO Authentication and Passkeys.pptx
Introduction to FIDO Authentication and Passkeys.pptx
 
1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT
1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT
1111 ChatGPT Prompts PDF Free Download - Prompts for ChatGPT
 
Oauth 2.0 Introduction and Flows with MuleSoft
Oauth 2.0 Introduction and Flows with MuleSoftOauth 2.0 Introduction and Flows with MuleSoft
Oauth 2.0 Introduction and Flows with MuleSoft
 
WebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM PerformanceWebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM Performance
 
Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...
Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...
Event-Driven Architecture Masterclass: Engineering a Robust, High-performance...
 
Intro to Passkeys and the State of Passwordless.pptx
Intro to Passkeys and the State of Passwordless.pptxIntro to Passkeys and the State of Passwordless.pptx
Intro to Passkeys and the State of Passwordless.pptx
 
Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...
Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...
Choosing the Right FDO Deployment Model for Your Application _ Geoffrey at In...
 
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdfThe Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
 
Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024
 
FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...
FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...
FDO for Camera, Sensor and Networking Device – Commercial Solutions from VinC...
 

Destaque

2024 State of Marketing Report – by Hubspot
2024 State of Marketing Report – by Hubspot2024 State of Marketing Report – by Hubspot
2024 State of Marketing Report – by HubspotMarius Sescu
 
Everything You Need To Know About ChatGPT
Everything You Need To Know About ChatGPTEverything You Need To Know About ChatGPT
Everything You Need To Know About ChatGPTExpeed Software
 
Product Design Trends in 2024 | Teenage Engineerings
Product Design Trends in 2024 | Teenage EngineeringsProduct Design Trends in 2024 | Teenage Engineerings
Product Design Trends in 2024 | Teenage EngineeringsPixeldarts
 
How Race, Age and Gender Shape Attitudes Towards Mental Health
How Race, Age and Gender Shape Attitudes Towards Mental HealthHow Race, Age and Gender Shape Attitudes Towards Mental Health
How Race, Age and Gender Shape Attitudes Towards Mental HealthThinkNow
 
AI Trends in Creative Operations 2024 by Artwork Flow.pdf
AI Trends in Creative Operations 2024 by Artwork Flow.pdfAI Trends in Creative Operations 2024 by Artwork Flow.pdf
AI Trends in Creative Operations 2024 by Artwork Flow.pdfmarketingartwork
 
PEPSICO Presentation to CAGNY Conference Feb 2024
PEPSICO Presentation to CAGNY Conference Feb 2024PEPSICO Presentation to CAGNY Conference Feb 2024
PEPSICO Presentation to CAGNY Conference Feb 2024Neil Kimberley
 
Content Methodology: A Best Practices Report (Webinar)
Content Methodology: A Best Practices Report (Webinar)Content Methodology: A Best Practices Report (Webinar)
Content Methodology: A Best Practices Report (Webinar)contently
 
How to Prepare For a Successful Job Search for 2024
How to Prepare For a Successful Job Search for 2024How to Prepare For a Successful Job Search for 2024
How to Prepare For a Successful Job Search for 2024Albert Qian
 
Social Media Marketing Trends 2024 // The Global Indie Insights
Social Media Marketing Trends 2024 // The Global Indie InsightsSocial Media Marketing Trends 2024 // The Global Indie Insights
Social Media Marketing Trends 2024 // The Global Indie InsightsKurio // The Social Media Age(ncy)
 
Trends In Paid Search: Navigating The Digital Landscape In 2024
Trends In Paid Search: Navigating The Digital Landscape In 2024Trends In Paid Search: Navigating The Digital Landscape In 2024
Trends In Paid Search: Navigating The Digital Landscape In 2024Search Engine Journal
 
5 Public speaking tips from TED - Visualized summary
5 Public speaking tips from TED - Visualized summary5 Public speaking tips from TED - Visualized summary
5 Public speaking tips from TED - Visualized summarySpeakerHub
 
ChatGPT and the Future of Work - Clark Boyd
ChatGPT and the Future of Work - Clark Boyd ChatGPT and the Future of Work - Clark Boyd
ChatGPT and the Future of Work - Clark Boyd Clark Boyd
 
Getting into the tech field. what next
Getting into the tech field. what next Getting into the tech field. what next
Getting into the tech field. what next Tessa Mero
 
Google's Just Not That Into You: Understanding Core Updates & Search Intent
Google's Just Not That Into You: Understanding Core Updates & Search IntentGoogle's Just Not That Into You: Understanding Core Updates & Search Intent
Google's Just Not That Into You: Understanding Core Updates & Search IntentLily Ray
 
Time Management & Productivity - Best Practices
Time Management & Productivity -  Best PracticesTime Management & Productivity -  Best Practices
Time Management & Productivity - Best PracticesVit Horky
 
The six step guide to practical project management
The six step guide to practical project managementThe six step guide to practical project management
The six step guide to practical project managementMindGenius
 
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...RachelPearson36
 

Destaque (20)

2024 State of Marketing Report – by Hubspot
2024 State of Marketing Report – by Hubspot2024 State of Marketing Report – by Hubspot
2024 State of Marketing Report – by Hubspot
 
Everything You Need To Know About ChatGPT
Everything You Need To Know About ChatGPTEverything You Need To Know About ChatGPT
Everything You Need To Know About ChatGPT
 
Product Design Trends in 2024 | Teenage Engineerings
Product Design Trends in 2024 | Teenage EngineeringsProduct Design Trends in 2024 | Teenage Engineerings
Product Design Trends in 2024 | Teenage Engineerings
 
How Race, Age and Gender Shape Attitudes Towards Mental Health
How Race, Age and Gender Shape Attitudes Towards Mental HealthHow Race, Age and Gender Shape Attitudes Towards Mental Health
How Race, Age and Gender Shape Attitudes Towards Mental Health
 
AI Trends in Creative Operations 2024 by Artwork Flow.pdf
AI Trends in Creative Operations 2024 by Artwork Flow.pdfAI Trends in Creative Operations 2024 by Artwork Flow.pdf
AI Trends in Creative Operations 2024 by Artwork Flow.pdf
 
Skeleton Culture Code
Skeleton Culture CodeSkeleton Culture Code
Skeleton Culture Code
 
PEPSICO Presentation to CAGNY Conference Feb 2024
PEPSICO Presentation to CAGNY Conference Feb 2024PEPSICO Presentation to CAGNY Conference Feb 2024
PEPSICO Presentation to CAGNY Conference Feb 2024
 
Content Methodology: A Best Practices Report (Webinar)
Content Methodology: A Best Practices Report (Webinar)Content Methodology: A Best Practices Report (Webinar)
Content Methodology: A Best Practices Report (Webinar)
 
How to Prepare For a Successful Job Search for 2024
How to Prepare For a Successful Job Search for 2024How to Prepare For a Successful Job Search for 2024
How to Prepare For a Successful Job Search for 2024
 
Social Media Marketing Trends 2024 // The Global Indie Insights
Social Media Marketing Trends 2024 // The Global Indie InsightsSocial Media Marketing Trends 2024 // The Global Indie Insights
Social Media Marketing Trends 2024 // The Global Indie Insights
 
Trends In Paid Search: Navigating The Digital Landscape In 2024
Trends In Paid Search: Navigating The Digital Landscape In 2024Trends In Paid Search: Navigating The Digital Landscape In 2024
Trends In Paid Search: Navigating The Digital Landscape In 2024
 
5 Public speaking tips from TED - Visualized summary
5 Public speaking tips from TED - Visualized summary5 Public speaking tips from TED - Visualized summary
5 Public speaking tips from TED - Visualized summary
 
ChatGPT and the Future of Work - Clark Boyd
ChatGPT and the Future of Work - Clark Boyd ChatGPT and the Future of Work - Clark Boyd
ChatGPT and the Future of Work - Clark Boyd
 
Getting into the tech field. what next
Getting into the tech field. what next Getting into the tech field. what next
Getting into the tech field. what next
 
Google's Just Not That Into You: Understanding Core Updates & Search Intent
Google's Just Not That Into You: Understanding Core Updates & Search IntentGoogle's Just Not That Into You: Understanding Core Updates & Search Intent
Google's Just Not That Into You: Understanding Core Updates & Search Intent
 
How to have difficult conversations
How to have difficult conversations How to have difficult conversations
How to have difficult conversations
 
Introduction to Data Science
Introduction to Data ScienceIntroduction to Data Science
Introduction to Data Science
 
Time Management & Productivity - Best Practices
Time Management & Productivity -  Best PracticesTime Management & Productivity -  Best Practices
Time Management & Productivity - Best Practices
 
The six step guide to practical project management
The six step guide to practical project managementThe six step guide to practical project management
The six step guide to practical project management
 
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
 

Mathematics, Statistics, and Sales Chat - A Web Retailer Case Study

  • 1. Page 1   Mathematics, Statistics, and Sales Chat A Web Retailer Case Study    
  • 2. Page 2   Introduction With the coming of age of web as a mainstream sales and marketing channel, companies have invested substantial resources in enhancing their web presence. This includes large investments in web dvertising. In addition, companies are looking for ways to improve sales conversion and customer experience for web shoppers. Sales chat is a medium that can provide a lift in both these areas. With the growing popularity of chat as a communication medium, particularly among the new generation of consumers, potential for revenue generation from this channel is enormous. The obvious analogy is to consider the sales chat agents as a “virtual sales force” for a” virtual store”. However, a key difference exists. In a real store there are relatively few visitors and a significant fraction of the visitors come with intent to buy i.e. they are “hot” prospects. On the other hand, major web stores such as Amazon, eBay and Overstock have millions of visitors every week and an overwhelming majority of these visitors do not intend to buy. They also have the ability to switch from one store to another at the click of a mouse button. Considering the visitor volumes and the low average likelihood to buy, it is not profitable to randomly engage in chat with every visitor. It is imperative to identify a subset of this visitor population that has a substantially greater likelihood to buy. The following case study presents applications of statistical/mathematical models in identifying “hot” prospects and improving conversion, revenue generation and customer experience for a major web retailer.  
  • 3. Page 3  The Conversion Funnel The starting point in understanding and optimizing the performance of this virtual sales force is the Conversion Funnel (Figure 1). The funnel helps visualize the size of the opportunity. Figure 1 represents the funnel for the web retailer in a particular week. Layers 1 & 2 are filters, i.e. these are determined and controlled by the retailers, while layers 3 to 6 are leakages that are essentially decisions made by the customer during the browsing/buying process and are not in the retailer’s control. The science is essentially in determining the appropriate filters to apply in selecting the right customers and matching them up with the right agents to minimize the leakages in the funnel. Statistical Scoring Model – Filter (Layer 1)   The first filter identifies the “hot leads” i.e. the people most likely to purchase via chat. In particular, this identifies customers who have a significantly higher likelihood of purchasing from a chat agent than on their own in a selfservice mode. This is an important factor since self-service is obviously a lower cost channel than chat and if a customer is very likely to purchase via self-service then the business case for inviting them to a chat engagement is poor. To avoid cannibalization of a cheaper channel A/B tests are conducted on a regular basis where a fraction of the “hot leads” are not invited to chat and their self-service conversion rates are compared to conversion rates of the remaining “hot leads” who are invited to chat. Typically, conversion rate for chat engagements among this “hot lead” population is substantially higher (5x-10x) than that for self-service engagements. Identification of hot leads is accomplished using a statistical scoring model. The scoring can be done in real time while the visitor/prospective customer is browsing on the website. The scoring is based on a
  • 4. Page 4  number of attributes including time of the day, day of the week, geographical location of the customer, product category, exhibited behavior on the web site etc. Figure 2 schematically illustrates the scoring model. Essentially certain patterns of behavior exhibit a much greater propensity to buy than others. The scoring model essentially estimates a probability of purchase (P(sale)). Statistical and Data Mining techniques such as Naïve Bayes, Logistic Regression or Neural Networks are used to develop these scores. A threshold score can be set above which customers are invited to a chat. Based on variations in traffic and availability of agents the threshold score can be modified. As more data is generated, the system learns and the scoring model becomes better at identifying the hottest prospects.   Agent Optimization – Filter (Layer 2) Once the customers are scored and the “hot leads” identified, the next step is to invite these “hot leads” to a chat. The number of “hot leads” invited is based on tactical and strategic considerations. On the tactical front, it depends on several factors such as the number of agents available, acceptable abandonment rate without significantly affecting customer experience, average handle time and concurrency (how many chats can an agent handle at a time). This is a routine scheduling problem. The more interesting strategic problem is to determine the right number of agents to maximize profits. The scoring model only
  • 5. Page 5  prioritizes the visitors to the site. It does not automatically provide a threshold score for the “hot leads”. We provide the threshold score. This in turn determines the number of people invited to chat which establishes our agent staffing levels. But what is the right threshold score? How is it determined? The customers are being prioritized based on how “hot” they are. Based on the average order value for a given product type and the probability of a given “hot lead” to buy, the expected revenue from the transaction can be calculated. To increase the number of chats and hence the overall revenues, we lower the threshold score inviting less qualified leads, at the same time increasing the number of agents. These less qualified customers on an average generate lower revenues per customer i.e. the marginal revenue of these customers is lower. This implies, as we keep adding agents to interact with less and less qualified leads, the marginal profit generated by the additional agents keeps declining. We keep lowering the threshold, increasing the number of “hot leads” and adding agents till we stop making a marginal profit. To estimate the number of agents corresponding to this, we use an optimization algorithm Finally, scoring techniques are also used to match the right agent to the customer. The essential concept is displayed in Figure 3 where we see that the agent Raymond is as   such a top performer but is particularly skilled in selling Electronics products. We score various product-agent combinations and manage our chat queues and routing based on not just the overall performance of the agent but also on historical performance in various product categories. The goal here is not just to look at product-agent combinations but to develop a comprehensive scoring model that scores the agent for a set of customer/product attributes and determines the best agent to talk to a given “hot lead.” Conclusion Internet chat is a growing channel for sales over the web and retailers are adding this capability to their websites. However, like in self-service web retailing, success in driving up sales chat revenues and profitability will go to players who use advanced data-driven approaches to drive customer intelligence and chat engagement decisions. References ITSMA and PAC, How Customers Choose Study, 2009 ITSMA and PAC, How Customers Choose Study, 2009 ITSMA and PAC, How Customers Choose Study, 2009 W_MATH_1012