Profits from Closed Customers


Closed Customers aren’t really closed; they stay on in your database.

Database marketing can help you win revenue even after a customer relationship is discontinued. This is illustrated by the following example – A prominent global financial services giant, with nearly 100 years of history, faced a unique problem while operating in India. While it had been one of the earliest entrants in the credit card industry in India, it had rapidly been losing market share to newer and nimbler more aggressive local competitors.

Indian customers have one of the lowest levels of debt worldwide due to cultural aversion to debt and lack of competition in the pre 1990 era. The credit cards receivables business in India is also a loss making operation as of 2006, because of rampant competition and discounting on annual fees and charges. Lending in India is complicated because the credit bureau CIBIL was only in nascent stages, and declared income and actual income of people varied due to the tax laws and ‘black money’ economy.

However the average receivables per card had been steadily increasing in India and it had potential to make huge profits once Indian customers became comfortable with rotating balance and paying finance charges. The credit card division also had a culture of conservative lending only to prime customers, with a good track record. On the other hand, the company’s personal loan business was making great strides in both revenue and profitability growth due to aggressive selling to both prime and sub prime customers. As a result of this the company had built up a database of 3 million customers, out of which nearly 2 million had paid off their loans.

To improve the profitability of the credit card division, and offer its customers a more value added portfolio of financial services, the company embarked on a data mining project of cross selling to its closed personal customers. After extensive tests and research based on selective tele-calling to its customer database, the company found out the following analytical findings-

1) Customers who had paid back their loans on time were the customers who were good credit customers. These customers had also increased their income since the time they had closed their personal loan.

2) People who had closed their personal loans were targeted for re churning by the person loans business. However after 6 months of closing their loans, if the customer did not take another personal loan, they were unlikely to ever take a personal loan. Thus if these customers were called again for personal loan, it would be unprofitable since the incremental expenses were not justified by incremental revenues.

3) People who bounced cheques but paid off their entire loan were bad credit risks, especially for a revolving line of credit as in for credit cards.

4) People who were called by the credit card division had better brand recall if they had an earlier relationship with personal loans division. Since they paid off their loans on time, their experiences with the company as a whole were very positive. This goodwill of the company’s brand helped to trigger higher response ratios (almost 20 % of such people took the credit card compared to only 5 % for the general population)

5) De to regulatory reasons both the credit card division and the personal loan division had to maintain an arm’s length distance. In order to do so, the credit card division decided on a transfer price of 600 rupees plus 1% of average receivables to the personal loan division. This helped track the profitability of the exercise better.

As a result of the exercise the company managed to sell an extra fifty thousand credit cards. The program was such a success that it was adopted world wide. The personal loan division earned tens of millions of rupees from its closed customer database, and the credit card division managed to increase its market share slightly.

Thus mining its own database of customers helped the company achieve the following-

a) Increase profitability
b) Improve brand recall and enhance the existing relationship
c) Cut down on marketing costs by targeting more responsive customers
d) Improve the life time value of revenues earned from each customer

This article  builds up an argument for using internal data at a customer level for decreasing marketing costs and enhancing brand recall.

Corporate Success Tips

ops1.jpg 

OFFICES

Never walk down the hall without a document in your hands. People with documents in their hands look like hardworking employees heading for important meetings. People with nothing in their hands look like they’re heading for the cafeteria.

People with the newspaper in their hands look like they’re heading for the bathroom. Above all, make sure you carry loads of stuff home with you at night, thus generating the false impression that you work longer hours than you do.

COMPUTERS

Use computers to look busy. Any time you use a computer, it looks like work to the casual observer. You can send and receive personal e-mail, calculate your finances and generally have a blast without doing anything remotely related to work.

These aren’t exactly the societal benefits that everybody from the computer revolution expected but they’re not bad either. When you get caught by your boss–and you will get caught–your best defense is to claim you’re teaching yourself to use the new software, thus saving valuable training dollars.

You’re not a loafer, you’re a self-starter. Offer to show your boss what you learned. That will make your boss scurry away like a frightened salamander.

DESK

Messy desk. Top management can get away with a clean desk. For the rest of us, it looks like you’re not working hard enough. Build huge piles of documents around your workspace.

To the observer, last year’s work looks the same as today’s work; it’s volume that counts. Pile them high and wide. If you know somebody is coming to your cubicle, bury the document you’ll need halfway down in an existing stack and rummage for it when he/she arrives

11 Analytics Softwares

1) R

R is open source, is similar to S and you can find ample support groups online. The only down side is that R does not have a good GUI (yet ..). It does have a rudimentary data GUI

http://www.r-project.org/

2) WEKA

http://www.cs.waikato.ac.nz/ml/weka/index_downloading.html

3) Minitab

  • http://www.minitab.org
  • 4) E Views

    http://www.eviews.com

    5) For Bayesian Inference

    http://www.mrc-bsu.cam.ac.uk/bugs

    6) Matlab

    http://www.mathworks.com

    7) Jim LeSage’s econometrics toolbox

    http://www.spatial-econometrics.com

    8) Office 2007 and Excel 2007 (www.microsoft.com) 🙂

    Office 2007 and Excel 2007 now supports up to 1,048,576 rows. If you were going to make the upgrade anyhow, it might save you a few bucks over an expensive statistics package. The statistics plugin for Office 2007 is a lot better than previous versions as well. Runs almost as many analytics as SPSS.

    9) JMP

    http://www.sas.com/apps/demosdownloads/jmptrial_PROD_7.0_sysdep.jsp?packageID=000415&jmpflag=Y

    10) www.sas.com

    11) www.spss.com

    Sites for country analysis Data

    Please note these websites will have their own citation /quotation/copyright  policy. This is in addition to the seperate list created on qualitative data (see archives) –

    1) Eurostat’s web site (at ec.europa.eu/eurostat) for data on European Union Member States

    2)International Monetary Fund’s website.

    http://www.imf.org

    3) EarthTrends

  • http://www.earthtrends.wri.org/searchable_db/index.php?theme=5
  • 4) Nation Master

    www.nationmaster.com

    5) France

    http://www.insee.fr/fr/home/home_page.asp

    6) CIA World Fact Book

    https://www.cia.gov/library/publications/the-world-factbook/index.html

    7) World Bank (Covered earlier in achives of decisionstats.com)

    http://www.worldbank.org/

    8) UN Stats

    http://unstats.un.org/unsd/default.htm

    9) BBC

    http://news.bbc.co.uk/1/hi/country_profiles/default.stm

    10) Economist.com

    http://www.economist.com/countries/

    11) OECD – Organisation for Economic Cooperation and Development

    http://www.oecd.org/statistics/

    Please do let us know of feedback

    5 Graphing Softwares for the Web

    Here is a list of softwares apart from Adobe Flash and Microsoft Silverlight

    1) PHP offers the ability to dynamically create Shockwave Flash files.
    In addition, you can try the tool named PHP/SWF Charts.
    Link:
    http://www.maani.us/charts/index.php

    2) Another library for PHP is Open Flash Chart
    An article illustrating how to use it is here
    http://www.linux.com/feature/121823

    3)  you want a nice lightweight charting facility for the web with a simple API, you could try the Google Charts API, which is free to use and incorporate into your own applications
    http://code.google.com/apis/chart/

    4)Crystal Xcelsius

    Industry leading interactive data visualization.

    Create interactive Excel dashboards, business presentations and visual calculators from ordinary spreadsheets – then integrate them into PowerPoint, Word, PDF and the Web.

    5) Some other softwares

  • http://www.flashactionscript.org
  • http://www.flashkit.com
  • http://www.fusioncharts.com/Default.asp
  • http://www.spreadsheetconverter.com/excel-web.htm
  • Politics, Polls, Globalization

    1) US Presidents for past three decades

    Bush-Clinton-Clinton-Bush-Bush- Clinton (standing for election)

    Indian PM’s for first three decades

    Nehru-Gandhi-Gandhi-gap years -Gandhi (now standing for election)

    2) Polls

    Prediction -Result

    Clinton will win IOWA. Hence Obama wins IOWA by 9 points

    Obama lead in New Hampshire by 10 points . Hence Clinton wins New Hampshire.

    Polls predict close race in Gujarat. Hence Narendra Modi wins 2/3 rds majority.

    Globalization (in Movies)

    Richard Attenbourgh makes Gandhi

    Shekhar Kapur makes Elizabeth

    Do Monkeys Pay for Sex?

    According to the paper, “Payment for Sex in a Macaque Mating Market,” published in the December issue of Animal Behavior, males in a group of about 50 long-tailed macaques in Kalimantan Tengah, Indonesia, traded grooming services for sex with females; researchers, who studied the monkeys for some 20 months, found that males offered their payment up-front, as a kind of pre-sex ritual. It worked. After the females were groomed by male partners, female sexual activity more than doubled, from an average of 1.5 times an hour to 3.5 times. The study also showed that the number of minutes that males spent grooming hinged on the number of females available at the time: The better a male’s odds of getting lucky, the less nit-picking time the females received.