• No results found

Implementing a Customer Lifetime Value Approach to Sales and Marketing Strategies

N/A
N/A
Protected

Academic year: 2021

Share "Implementing a Customer Lifetime Value Approach to Sales and Marketing Strategies"

Copied!
13
0
0

Loading.... (view fulltext now)

Full text

(1)

1

growth team m e m b e r s h i p™

guidebook summary

Firm: Farmers Group, Inc.

Industry: Multi-line, multi-company insurer and financial services provider Headquarters: Los Angeles, California, USA

Geographic Footprint: United States

Ownership: Wholly owned subsidiary of Zurich Financial Services Revenue (2011): ~$10 billion USD

Implementing a Customer Lifetime Value Approach to Sales

and Marketing Strategies

Problem:

Farmers’ traditional business metric, policies in force, lacked customer focus and resulted in sales and marketing strategies focused on the quantity, rather than the quality, of customer relationships.

Solution:

The Market Analytics team creates a customer lifetime value (CLTV) model to predict the dollar revenue a customer will generate over his/her lifetime with the company. This allows Farmers to:

• Identify the CLTV of each individual customer • Measure the effectiveness of their marketing spend • Attract and retain higher CLTV customers

• Embed CLTV-based strategies within the sales function

Business Results:

• 20% improvement in overall marketing ROI

• Incremental increase in Farmers average customer profitability through an improvement in their market share of higher-value customers

• 100% improvement in direct mail ROI

Resources Required:

• Building and testing the CLTV model took six to eight months • The CLTV initiative involved a ten-member team composed of

analysts and managers within Market Analytics, as outlined below: - Analysts: SAS programmers to perform statistical modeling - Analysts: Staff familiar with the firm’s internal data infrastructure - Managers: MBAs with a strong technical and business background • Building, testing, and rolling out the CLTV model cost approximately

$1 million

• Senior Leadership Executive—a champion for CLTV initiatives within leadership ranks, e.g., Chief Marketing Officer, Chief Operating Officer

Applicability of Best Practice to Executive Functions:

Function Applicability

Corporate Strategy Marketing

Market Research

(2)

2

growth team m e m b e r s h i p™

best practice guidebook

key takeaway:

Create a model to calculate customer lifetime

value (CLTV)—the monetary revenue each customer generates over their

lifetime with the firm—and establish the foundation for a data‑driven marketing strategy

Farmers’ model calculates CLTV using three variables: total relationship revenue,

predicted customer tenure, and predicted future purchase revenue

Farmers’ Customer Lifetime Value Model

critical data inputs for the cltv model

1. Four to five years of historical customer data to analyze the initial and subsequent purchase relationship with the company; 2. transaction‑based data that is aggregated at the customer (household) level;

3. customer information such as age, marital status, credit history, and number and type of policies held

Note: Of the six to eight months it took to build and test the model, 80% of the time was spent cleaning and understanding the data and 20% was spent modeling and determining the strategic implications of its output.

1. Total Relationship

Revenue Customer Tenure2. Predicted 3. Predicted Future Purchase Revenue

Model Variables

This is the total revenue an individual customer generates from the total number of relationships they have with the company. This can be calculated daily, weekly, monthly or annually. Farmers looks at the total number of policies and the monthly premium revenues associated with each.

This is defined as the entire time period an individual customer has a relationship with the company. Farmers uses survival analysis because it allows them to predict how long a customer will remain a policyholder and differentiate between customer groups.

This is the expected future purchases customers will make based on their current relationship with the company, historical data, and demographic profiles. Farmers uses logistic regression to calculate the probability that a customer will make additional purchases.

CLTV Example

Jane Doe is 44 years old, married, living in Nevada and needs full-coverage automobile insurance for her new domestic sedan and liability-only insurance for her 10-year old family car. The monthly premium for the full-coverage policy is $50. The monthly premium for liability policy is $25.

The full-coverage automobile

insurance policy is estimated to last 48 months. The liability-only automobile insurance policy is estimated to last 30 months.

Based on the customer’s data, Farmers’ future cross-sell model predicts a 60% probability that Jane will purchase a homeowner’s insurance policy. The average home policy’s life span is 120 months and premiums are $100 per month.

Future cross-sell models are built on five years of data from a sample of 100,000 customers and validated on the balance of the customer base over the same five years. Market Analytics develops product-specific future cross-sell models to calculate the probability of adding a home policy, an automobile policy, etc.

New Sedan Policy: $50 × 48 months

(3)

3

growth team m e m b e r s h i p™

best practice guidebook

key takeaway:

Analyze existing customers through CLTV and

leverage results to secure buy‑in for a CLTV‑based segmentation study

Market Analytics segments existing customers into three

lifetime value segments and evaluates their profitability…

CLTV-based Segmentation Analyses*

…and utilizes the results to demonstrate CLTV’s impact

and secure sponsorship for additional segmentation work

Stakeholder Engagement Meeting

Customer Lifetime Revenue (CLTV) $6,000 $3,000 $1,050 Cost of Goods Sold (over lifetime) $3,500 $1,500 $1,000 Acquisition/Marketing Costs $125 $125 $125 Lifetime Margin/Profit $3,875 $375 ($75)

% of Existing Customer Base 15% 50% 35%

High-CLTV Customer Director, Market Analytics VP Marketing Middle-CLTV

Customer Low-CLTV Customer

Profitability calculations for each segment requires allocating the elements of cost of goods sold (cost of coverage, i.e., losses, claims support, etc.) to each customer.

Customers with the highest CLTV bring in 25 times more revenue than the lowest value customer. High CLTV Middle CLTV CLTVLow 0 0.5 12.5 Customer Lifetime Value ($000) CLTV Segments Director, Customer Insights VP Sales

The analysis highlights a major opportunity to grow the high-CLTV customer base and the need to change marketing strategies that are acquiring unprofitable low-CLTV

customers. Further segmentation work is necessary to be able to target high-CLTV prospects and maximize profitability.

* Numbers are illustrative only.

stakeholder engagement strategy

1. Highlight sense of urgency through CLTV’s financial implications:

• Emphasize the major differences between revenues from different customer types—the best customers bring in 25 times more revenue than lower value customers.

• Stress that current sales and marketing expenditures do not differentiate between high‑ and low‑value customers—acquiring customers whose acquisition costs exceed their revenue.

2. Establish trust in accuracy of the CLTV model:

• Demonstrate model’s strength by applying it to historical customer data—2000 customers from three years prior—and showing the high correlation between its predictions and actual customer loyalty. 3. Issue call to action on new segmentation:

• Establish that current segmentation scheme has resulted in over one third of current customers generating losses.

(4)

4

growth team m e m b e r s h i p™

best practice guidebook

key takeaway:

Develop actionable CLTV segment profiles

to enable a CLTV‑focused marketing strategy

Market Research conducts a survey to identify the

predictors of CLTV segment membership and

develop in-depth CLTV segment profiles

Market Analytics engages Marketing

to demonstrate the potential of a

CLTV-based marketing strategy

Marketing Leadership Engagement Segment Classification, Behaviors,

and Attitudes Survey High-CLTV Segment Profile

Illustrative

Customer Survey

What do you think about insurance in general?

Scale

Agree strongl

y

Agree somewhat Neither agree nor disagree Disagree somewhat Disagree strongl

y

1. I would have insurance even if it wasn’t required 2. I buy the minimum amount

of insurance that’s required 3. I keep my deductibles high

to lower my premiums 4. The price I pay for

insurance is worth it for the peace of mind it gives me 5. It’s important to have

adequate insurance coverage at all times 6. I don’t pay much attention

to my insurance policies and probably won’t until I have a claim

7. Insurance is too

complicated to understand without an sales

representative

• Process took three months and involved close to 2,000 interviews—1,400 prospects/600 existing customers.

- Correlate survey data of the existing customers and prospects to develop a model and algorithm to predict CLTV.

• Uses behavioral, attitudinal, and asset-based questions to provide multiple factors for segment/ cluster analysis. Analyze survey results by factors that best distribute customers into CLTV segments.

Family Focalist

• Who am I - Married Female - Mid 40’s to Early 60’s - Have children that drive - Higher Education - Upper income bracket

- Loyal, confident and detail oriented - Purchase insurance for peace of

mind

• Insurance Wants - Personal contact

- Excellent customer service - Competitive rates with minimal

increase

- Established Company

CLTV—$5,500 (High-CLTV segment)*

CLTV-based segmentation provides a revenue value to each cluster. This makes resource allocation and targeting decisions more objective and straightforward.

CLTV Business Case Objective: Create buy-in for marketing

strategies that target CLTV-based segments by emphasizing the measurable benefits for Marketing:

• Actionability: CLTV enables better targeting

of customer segments based on rigorous financial analysis.

• Profitability: CLTV focuses marketing

decisions on the profitability of every customer, providing financial justification to the profit and loss owners for resource allocation, and generating an ROI for all marketing expenditures.

• Consistency: CLTV provides Marketing with

a consistent, data-driven way of identifying, evaluating and prioritizing customer acquisition and retention.

Chief Marketing

Officer Director, Market Analytics

(5)

5

growth team m e m b e r s h i p™

best practice guidebook

key takeaway:

Institute a CLTV‑based marketing strategy

focused on the long‑term profitability of each customer

Market Analytics provides CLTV-based segmentation data

to focus marketing efforts on the most profitable customer segments…

Pre-CLTV Segmentation*

…and demonstrates results to entrench the strategy

Optimizing the Marketing Mix

CLTV Case-in-Point: Direct Mail*

cltv implications for marketing

• Optimize marketing mix model based on CLTV and improve ROI

- Identify the most effective marketing channels for high‑CLTV segments then reallocate funds to those channels to optimize marketing spend

• Create CLTV‑focused advertising

- Perform ad‑ and copy‑testing with high‑CLTV customers to improve advertising effectiveness • Monitor and report on key metrics that influence customer profitability

- Acquisition rate and cost—a customer’s total acquisition cost must not exceed their CLTV - Retention rate and cost—it is more cost efficient to retain customers than to acquire new ones - Margin revenue growth—grow customer revenue by increasing share of wallet

Before CLTV Focus Mailed 500,000 customers

• 0.05% Return • 1000 Respondents

• 100 Converted Leads at an average of $1,250 lifetime revenue each.

Total Expected Lifetime Revenue Return ≈ $125,000

Optimized for CLTV Mailed 350,000 customers

• 0.04% Return • 750 Respondents

• 75 Converted Leads at an average of $2,750 lifetime revenue each. • 35% of converted new customers

are high-CLTV Customers

Total Expected Lifetime Revenue Return ≈ $206,250

Post-CLTV Segmentation and Profitability Analysis* Segment Happy Lovers FocalistFamily BoomersBaby

Share of Market 15% 15% 20%

CLTV Segment Middle High Low

Customer Lifetime Revenue

(CLTV) $2,500 $5,000 $1,050 Cost of Goods Sold (over

lifetime) $1,500 $3,500 $1,000 Acquisition/ Marketing Cost $150 $175 $35

Lifetime Margin/Profit $850 $1,325 $15

Targeted all customers that demonstrate a high probability to respond.

Reduced cost and increased revenue by balancing the probability to respond with the predicted CLTV of respondents. CLTV enables Marketing

to differentiate between customer segments based on revenue and profitability measures as well as

psychographics, and to target the most viable segment.

* Numbers are illustrative only.

Segment 40-Something Female 30-Something Male 20-SomethingMarried Share of

Market 20% 10% 10%

(6)

6

growth team m e m b e r s h i p™

best practice guidebook

key takeaway:

Analyze existing sales via CLTV to identify

and embed customer acquisition best practices

Market Analytics analyzes the CLTV of sales

per representative to identify best practices in

attracting and retaining high-CLTV customers

Sales Force Assessment via CLTV

Market Analytics utilizes adoption imperatives and

training to embed the CLTV initiative in the sales force

Sales and CLTV Adoption

Top 5% of

Sales Reps Bottom 5% of Sales Reps 0 2 4 6 Average CLTV ($000) Percentage of Commissions for a Sales Rep

Sales CLTV Best Practices Survey

This anonymous 15‑minute survey will uncover practices that lead to sales success.

1. Currently, which method is most effective for bringing in new business? Please rank the top 4 in order of effectiveness, with “1” being the most effective.

____ Direct Mail ____ Telemarketing ____ Internet leads ____ Personal contact

____ Local advertising (newspapers, billboards, local cable, etc.)

____ Phone book ad

____ Community marketing (M.I.L.K., FPP, March of Dimes, etc.)

____ Referrals from business professionals ____ Personal referrals

Other, please specify _____________________

Conduct survey of sales reps to identify the best practices of sales reps with high average CLTV. Farmers turned the best practices into a “Habits” toolkit for sales reps.

Sales Manager Director, Market Analytics Training Manager Training Sales Manager Sales Representative Imperatives for Adoption

of CLTV by Sales • Be relevant—provide compelling

evidence that a CLTV approach improves sales and commissions

• Target key influencers—secure initial

buy-in from sales managers who will help embed CLTV at the front line

• Build CLTV credibility—establish

credibility of the CLTV model and the segments, especially with regard to previous segmentation

• Be prescriptive—use training to

underscore what works and what doesn’t in building a high-CLTV portfolio

• Refine infrastructure—modify processes,

systems, tools and rewards structures that drive desired behaviors

Value of a Customer Percentage of Customers Top 20% 50% Middle 20% Bottom 20% 3% One in five of your customers generates more than 50% of

your commissions.

Sales Representative

Habits of Highly Effective Sales Representatives

• Cross-sells and up-sells constantly • Prioritizes hiring and retaining quality staff • Operates with a marketing budget • Makes more frequent customer contact

High CLTV Low

(7)

7

growth team m e m b e r s h i p™

best practice guidebook

key takeaway:

Deploy CLTV scorecards to infuse a CLTV focus in sales behavior

Market Analytics creates customized CLTV-based scorecards to provide performance guidance to sales reps and sales managers

Sales Representative Quarterly CLTV Scorecard *

sales management tool

The transparency and usefulness of the CLTV‑based scorecard provides sales reps with an effective tool to manage customer acquisition strategy. Market Analytics generates an aggregated version of the scorecard for sales managers. The use of CLTV metrics enables managers to evaluate how each sales rep is performing in terms of individual customer quality as defined by their lifetime revenues. These data reveal performance issues, including the need for coaching on high‑CLTV customer acquisition, and the adherence to a customized policy renewal strategy for different CLTV segments.

Joe Smith

New Automotive and Home Insurance Customers

Customer Lifetime Value Analysis for Households Acquired in Q4 2008

CLTV Ranks by Premium revenue: High is $5,000 or more; Middle is $1,500 to $5,000; Low is $1,499 or less

Your Q4 2008 New Customers

CLTV Rank Percentage of New Customers 60% 30% 0%

High Middle Low

Your Top CLTV Customers

Household CLTV Rank

Jane Doe High

John Doe High

Al Doe High

Your Bottom CLTV Customers

Household CLTV Rank

Jane Smith Low

John Smith Low

Al Smith Low

What Separates Your high-CLTV from low-CLTV Households?

Lifetime Value Drivers HouseholdsTop CLTV Bottom CLTV Households

Pricing Credit Tier 85% 15%

Average Household Policy Count 4.0 1.0

Average Annual Revenue $8,000 $400

Percent Cross Sold (Auto & Property) 88% 12%

This identifies by name a sample of the highest and lowest CLTV households within the portfolio. This puts a face to CLTV in a way that reinforces the need for the sales rep to distinguish high-CLTV customers and the tactics used to acquire them.

This differentiates high-CLTV households from low-CLTV ones in ways that underscore the significant drivers for targeting, acquiring and retaining high-CLTV customers.

* Numbers are illustrative only.

The rep’s scorecard classifies the lifetime value of the new customers acquired in the quarter and how that compares with the CLTV of all the customers acquired by the reps in the area. These ongoing classifications are leading indicators of the quality of the rep’s customer portfolio and their performance relative to peers

MIDDLE/LOW Your Overall New Customer CLTV Rating for Q4 2008:

(8)

8

growth team m e m b e r s h i p™

best practice guidebook

Business Results

The CLTV model is transforming the way we do business, as it enables us to optimize everything from marketing spend to the customer experience. CLTV and increasing our share of the high-CLTV market are becoming the cornerstones of our growth strategy.

—Director, Insight & Innovations, Farmers Insurance Group Marketing Mix ROI

ROI Before and After Optimization for CLTV Marketing Mix Allocation Model

Before and After Optimization for CLTV

Q3 2008 Q4 2008 Q3 2008 Q4 2008

Marketing refined its marketing

mix to concentrate on high-CLTV customers

leading to a double-digit increase in ROI

Sales’ focus on enhancing

the CLTV of their customer portfolios has

increased customer profitability

Market Share by CLTV Segment (Indexed) Total Total Radio Radio Internet Internet TV TV Print Print

Direct Mail Direct Mail

Marketing worked with its advertising agencies to determine the media preferences of high- and middle-CLTV customers. This combined with a CLTV-focused marketing mix model enabled Marketing to allocate resources to optimize marketing effectiveness.

Radio Internet TV Print Direct Mail Radio Internet TV Print Direct Mail Middle CLTV High CLTV Q1 2008 Q1 2009 Q1 2008 Q1 2009 100% 104% 100% 106%

CLTV enables Marketing to justify expenditures to the profit and loss owners and makes it easier to calculate ROI. As a result, overall marketing ROI increased 20%.

Having a sales force focused on maximizing the CLTV of customers versus increasing the number of policyholders has enabled Farmers to:

• Improve their market share of middle- and high-CLTV customers

(9)

9

growth team m e m b e r s h i p™

best practice guidebook

Key Lessons Learned

Profiled Company Perspective

• Building the CLTV model is an iterative process; do not expect to get it right immediately. Because of the need

for advanced modeling capabilities, vendors are often required. Farmers engaged an advanced analytics provider

with SAS programming capabilities to assist in the creation of the model.

• The skill profile of the Market Analytics team at Farmers was critical to the development and deployment of

the CLTV model. For instance, the management team comprises MBAs with quantitative backgrounds, which

provides a facility with data and an understanding of their strategic applications. This ability to synthesize data

and develop strategic recommendations is vital to selling in CLTV to executive management.

• Since the CLTV model creates a financial valuation for every customer, it focuses directly on revenue

generation and can be applied to the customer-facing functions—sales, marketing, product development, and

customer service. Accordingly, CLTV provides a sound financial case to engage stakeholders on the benefits of

CLTV-based strategies.

• To demonstrate the tactical and strategic applicability of CLTV it is important to create a success story quickly

with a prominent stakeholder. In Farmers case, the Chief Marketing Officer was the first to be engaged because

CLTV enables Marketing to institute strategies based on quantifiable customer value not just market size.

In addition, CLTV overcomes concerns over the effectiveness of marketing expenditures by providing clear

pre- and post-ROI calculations for marketing initiatives. To gain momentum, it is also critical that you share

success stories broadly.

• CLTV has numerous applications besides those profiled:

- Evaluate M&A targets: In cases where a significant cost of the acquisition is related to the customer base,

CLTV can help determine accurately the current and predicted value of the target company’s customer base

- Drive customer service investments: Knowing a customer’s lifetime revenue allows companies to create

different on-boarding procedures for different CLTV segments and develop tailored services for high-CLTV

segments (Due to regulatory reasons Farmers chooses not to pursue this strategy)

(10)

10

growth team m e m b e r s h i p™

best practice guidebook

Frequently Asked Questions

These FAQ’s summarize the Q&A session with Shiv Gupta, Director of Insights and Innovation, at Farmers Insurance,

during the Ask the Thought Leader Webcast of this guidebook conducted on July 16, 2009.

Q

How did you secure buy-in to develop the CLTV model?

The Innovations and Insight group was given a mandate and budget to change how the company was conducting business—explore new ways and territories. However, there are a lot of examples, such as this Best Practice Guidebook, that can be leveraged to help make a business case for implementing CLTV.

Q

How do you measure customers (Markets) who make big

purchases every 5–8 years?

This does make the purchase frequency more difficult to predict. CLTV becomes more difficult and a company in this segment could fallback on more subjective means of understanding customer value, such as loyalty and client relationships.

Q

How is CLTV applicable in B-to-B companies?

Employing CLTV within a B2B environment can be challenging; it is more difficult to garner customer insights. However, the models do not need to be as complex as relationships are fewer and better understood. Companies may use subjective measurements instead of the complex models, and should be able to identify customer tenure and the projected financial value of a customer.

Q

In what industries is it easiest to implement the

CLTV model?

Ranking them in order from easiest to most difficult:

a. Direct customer facing businesses, with large volumes of customers and frequent interactions (i.e. Insurance, Retail)

b. Consumer package good companies may have to work through data providers and panels to identify who their customers are and which ones are the most profitable

c. B-to-B businesses are challenging for the reasons given on questions three and five

Q

How do you predict how long a customer will stay

with Farmers?

Farmers used a logistical survival model based on years of historical customer data. They used a vendor, Fractal Analytics, for the initial analysis. However, over time, Farmers began developing their own analytics team. Performing the analytics then became more of a partnership between Farmers and Fractal. Farmers recommends using an outsourced analytics team if you do not have the necessary personnel or skill sets in house.

Q

Will the three variables used to calculate CLTV work

for everyone?

(11)

11

growth team m e m b e r s h i p™

best practice guidebook

Frequently Asked Questions (Continued)

Q

How has CLTV helped you defend Marketing’s budget in

these tough times?

Use the CLTV model to cost-justify the work. For example, most companies can look at the customers brought in by an advertising campaign and can predict the revenue of those customers for today and sometimes for a year. With CLTV if an advertising campaign brought in 1,000 new customers you can see what the cost and ROI of each is and forecast what the campaign’s revenue is for the next five years; justifying the initiative.

Q

How will Farmers handle competition when they have a

CLTV model?

It’s all about the execution, not just the idea. CLTV is already out there and many companies already use it in different ways and with different effect. If a company is able to execute the CLTV model well, they are already a competitor to be reckoned with.

questions?

If you have any questions regarding this webcast or the Growth Team Membership™ (GTM), e‑mail us at GTMResearch@frost.com. To learn more about GTM visit us at www.gtm.frost.com or on Twitter

(12)

12

growth team m e m b e r s h i p™

best practice guidebook

Source: Growth Team Membership™ research.

Supporting Tools & Resources

CLTV Implementation Resources

The Growth Team Membership™ research team has compiled this additional suite of resources to assist

members in learning more about developing and implementing CLTV.

Frost & Sullivan Resource—Growth Process Toolkit:

“New Product Development: Accelerating Growth through Unbiased and Rigorous

Early-Stage Product Evaluation.”

http://www.frost.com/prod/servlet/gtm-portal.

pag?tab=two&&&&&&itemId=9818&ctxixpLink=FcmCtx1&ctxixpLabel=FcmCtx2

.

Articles:

Gupta, Sunil et al. “Modeling Customer Lifetime Value,” Journal of Service Research 9.2 (November 2006):

139–155. Sage Publications.

http://www.anderson.ucla.edu/faculty/dominique.hanssens/content/JSR2006.

pdf

.

Hosford, Christopher. “Measuring for the long haul: Customer lifetime value metrics give marketers

a long view of how to spend.” BtoB Online. May 7, 2007.

http://www.btobonline.com/apps/pbcs.dll/

article?AID=/20070507/FREE/70507015/1034/FREE#seenit

.

Kumar, V. “Tough Times Call for CLV” The Economist: Executive Briefing (March 29, 2009).

http://robinson.

gsu.edu/news/newsmakers/vk_economist09.html

.

Lu, Junxiang “Modeling Customer Lifetime Value Using Survival Analysis—An Application in the

Telecommunications Industry” SUGI 28 (March 30–April 2, 2003): Paper 120–128.

http://www2.sas.com/

proceedings/sugi28/120-28.pdf

.

Rajkumar, Venkatesan, Kumar, V. “A Customer Lifetime Value Framework for Customer Selection and

Resource Allocation Strategy” Journal of Marketing 68.4 (October 2004).

http://www.atypon-link.com/

(13)

13

growth team m e m b e r s h i p™

best practice guidebook

Supporting Tools & Resources (Continued)

White Paper:

Checco, Chris. “Leveraging Customer Lifetime Value to Increase Return on Marketing Investment,”

September 20, 2006.

http://www.customerchemistry.com/DownloadFile.asp?getFile=LeveragingValue.pdf

.

Websites:

“Customer Lifetime Value: How to become a ‘CLTV Aligned’ organization,” Fractal Analytics

http://www.fractalanalytics.com/solutions/customer-lifetime-value/

.

“Special Lifetime Value Download Calculator and Instructions,” Database Marketing Institute,

http://www.dbmarketing.com/special_ltv.htm

.

Books:

Bejou D, Keiningham T, Aksoy L (2006), “Customer Lifetime Value: Reshaping the Way We Manage to

Maximize Profits,” Binghamton, NY: Best Business Books.

Berry M., Linoff G (2004), “Data mining techniques: for marketing, sales, and customer relationship

management,” John Wiley and Sons.

Farris P; et al. (2006), “Marketing Metrics: 50+ Metrics Every Executive Should Master,” Upper Saddle

River, N.J.: Wharton School Publishing.

Gupta S., Lehmann D.R. (2005), “Managing Customers as Investments: the Strategic Value of Customers

in the Long Run,” Wharton School Publishing.

Thernau T, Grambsch P (2001), “Statistics for Biology and Health—Modeling Survival Data: Extending the

Cox Model,” New York: Springer.

References

Related documents

In this section, we develop two ways to categorise character networks by exploiting the distributions of major and minor characters. The first one consists of studying centralisation,

maximizing problem; new models and algorithmic pro- cedures, both exact and heuristic, for solving project scheduling problems in constrained and unconstrained en-

With the release of the previous version (2.1) the horizontal data input function has been already offered for the following windows: Cost centre structure, Fixed investment

For students with emotional disorders or other types of disabilities, cognitive-behavioral interventions that teach students to discuss appropriate behaviors, role-play,

Whilst all sites showed globally very similar communities, cluster analysis for taxonomic patterns of meiofaunal communities based on Sørensen similarities of OTU presence/absence

When applying more informative food groups data based on characteristics relating to animal/plant origins, plant family or processing methods rather than only the main nutrients,

Results show that the period was one of atypically rapid growth of total factor productivity (TFP) relative to levels in OECD countries, and that the rise in firm efficiency

Nampa, Idaho is now part of greater Boise. It was established in 1891 and owes its growth to the Oregon Short Rail. It was an important railroad depot; other industry included