Marketing Information Systems: part 3 Course code: PV250 Dalia Kriksciuniene, PhD Faculty of Informatics, Lasaris lab., ERCIM research program Autumn, 2013 2 Customer Relationship Management •Customer relationship management (CRM) is a broadly recognized, widely-implemented strategy for managing and nurturing a company’s interactions with clients and sales prospects • The overall goals are: -to find, attract, and win new clients, -nurture and retain those the company already has, -retain former clients back, -and to reduce the costs of marketing and client service (Pepper, Rodgers, 2004) The spectrogram principle of the customer analysis •The success of the enterprise highly depends on the “prism” as analytical model which can convert “white light” of information to the swath of colours with different brightness: identify compounds of customer portrait by characteristics, their importance and effects to the financial results of the enterprise. 3 Components of CRM Systems •The software producers understand the structure of CRM differently •You can find CRM, which mean different goals: sales module, communication module, performance of sales personnel, distribution channel analysis, loyalty “point” systems, etc. (what type is Sugar CRM?, MS CRM ?, SAP CRM?) G:\Powerpoint\Pe_Uk\PE159-Curtis\Final files\Gif\ch07\C07NF08.gif 4 Customer Relationship Management (CRM) Systems – general understanding •Provide information on existing customers, their loyalty and churn •Identify and target new markets •Enhance customer’s satisfaction •Manage relationships with partner organizations •Marketing: cross-selling, upselling, bundling •Customer service •Partner relationship management •Internal marketing (making enterprise attractive for its workers for keeping their knowledge) 5 What’s hot from Gartner 2012 6 What’s hot from Gartner 2012 7 CRM- is a philosophy of management enterprise resources (4+1 main types). 8 9 Customer capital management goals CRM information needs CRM goal Information need Capability of accounting systems to supply info Profitable customers New and old customers Profit per customer No Profit calculation per unit Keep profitable customers as long as possible Communication history Sales info is available Limited info about reaction to promotions Win profitable customers back from competitors Customers of the competitors Who were won back No Convert not-profitable customers to the profitable Expenses per customer Sources for turnover No Provide incentives to get additional products Know individual needs No Enhance positive reference Opinion, referrals, impact No Enhance customer service programs Effectiveness of programs No 10 IDIC model for CRM D. Peppers ir M. Rogers (2004) 11 Application of IDIC model •Identify customers- explore individual characteristics. Needs variables for identification: tel.no. address, email, psychographic characteristics, preferences, habits •Differentiate customers- searching for different characteristics which enable segmenting. Definition of similar segments helps to focus attention to best (most profitable) groups, and create scenarios evoking specific behaviors •Interact with customers- search for tools and technologies for creating perception of the enterprise to its customer in attractive way, get feedback, avoid information distortion due to attitudes (e.g. caused by resistance to spam) •Customize treatment- maximize profit due to meeting individual needs • • 12 Two tasks for managing CRM 13 What is indicator? •Indicator is a common language among managers •Instead of evaluations “good”, “bad”, the numeric evaluations, rankings, graphical visualizations, etc. could be more effective •Indicator is a lever which we have to envisage, and use proper impulse of sufficient power to make impact on it. •Indicator reveals influences which affect enterprise. It is important to notice these influences, to know how they are created, what efforts are needed to make them serve to the enterprise needs. 14 Integrated approach- CRM perspective 15 Problems of getting right data for analysis •Accounting information is limited, there is need for contact points, where customer information can be recorded (loyalty cards, personalized access points, transaction terminals, call centres, web pages or social networks) •The best descriptive is qualitative data, but it is collected in inconsistent way (surveys), or stated by subjective judgments, or classified by subjectively extracting characteristics of communication •Therefore our challenge is to apply the historical purchase data, utilize information from access points and capture qualitative data consistently 16 How to create indicators ? •Traditional commonly understood marketing indicators? •What is missing? What direction should be followed in order to enhance power of indicators? •How to understand gap? •Common rules for creating indicators : absolute (turnover, profit), relational (EBITDA), percental (impact of marketing for “bottom line” in accounting), complex interpretation (RFM), formulas (LTV), ranking (loyalty) •Analytical report types : summarization, queries, trends, anomalies, extremities. •Textual, numerical, color, graphical 17 18 Information for evaluation •CRM evaluation based on accounting information •Defining loyalty and its relationships to sales •Using non-financial information •Balanced scorecards •Internet technologies based indicators •Social network analytics Gap of the indicators 19 Source: Zumstein, D. How to fill the gaps to final indicator •Making qualitative indicators. Negative side- hard to transform to measurable •Creating lead indicators which are going to influence factual results in (lag indicators). Negative side- some relationships between them are missing or misleading •Proxy indicators try to created intermediate links leading to final values Proxy—Financial—Statistical •Creating indicators similar to financial philosophy : Return on Customer Investment (ROCI); Return on Relationship (ROR); – similar to ROI (return on investment in finance) •Longitudinal metrics – involve dynamics •Refining indicators by learning relationships philosophy • 20 CRM variable types •Simple transactional variables – purchase value, frequency •Derived variables- CLTV- customer lifetime value •Survey-based: satisfaction, knowledge, preference •Event-based: churn, complaint •Expert-evaluation-based: loyalty •Compound variables – RFM •Proxy variables- compound-weighted-ranking based •Models: Pareto, Whale curve, custom designed models 21 Promising variable types 22 CRM indicator and metric samples •Customer profitability metrics •Cross-sell change •Process and operation cost change •Credit usage level •Change of number of customers and their structure: •attrition, •churn rate, •Naming groups by character: “vintages”, “cohort”, “VIP” •satisfaction changes according to survey data 23 CRM indicator and metric samples •Value of customer •Evaluation in monetary terms by assumption that customer is the asset of enterprise •NPV-net present value •Potential value (IRR) •Current and potential value according to survey data •ROI – return on investment to customer •Cycles among purchases: •Cycle duration (shorten, lengthen, regularity) •Buyer trajectory – characteristics accumulated during purchase history 24 CRM indicator and metric samples •Evaluation of purchase structure: •Large purchase buyers •Petit purchase buyers •Frequent purchase return makers •Grouping, segmenting metrics: •Decile analysis (divide by 10% segments) •Pareto principle •Whale curve •Share of customer (e.g. VISA uses share of wallet) •Share of personal consumption, expenditures •Customer satisfaction • • • 25 CRM indicator and metric samples •Life cycle value •Most valued customer segment- MVC •Relationship value •Relationship duration •Migration •Loyalty metrics •Specific behavior: “bought in past and will buy in future” •Attitude, brand preference •Tenure functions •Ranking according loyalty strength 26 27 Loyalty categories –their variety •Loyalty pyramid expresses levels of loyalties •No loyalty– first level of loyalty when it is simply absent The user freely searches for product by changing suppliers, not bonding to them. If he bought during promotion period, the sales of this loyalty group return back to previous level •False loyalty: customer does not feel any difference among products of suppliers, but he has no need to change them –behavior by inertia •Hidden loyalty- customer has preference to some product or supplier but not always keeps buying it •Real loyalty- the customer has clear preference and uses it even when there exists sufficient choice Loyalty categories –their variety 28 Compound variables – RFM •Variable R (Recency) show the number of days since the last visit till the date set for analysis •Variable F (Frequency) indicator is equal to the number of visits of the customer. •The M (Monetary value) is equal to the total value of purchases during all the history of communication. •CRM task lays in defining RFM combination matrix for decisions. E.g. how we treat recent customer who comes often, pays much ? How do we treat if he comes rarely? Do we change opinion if he comes only during holiday time? If we waited for his holidays and he missed – did he chose competitor? 29 “Whale curve“ analytic visualization •Customers are sorted by descending order of their turnover (or profit) values, in order to compute thier cumulative percent values and to plot to Y axis. •In X axis you plot the cumulative percent of the number of customers (e.g. if the enterprise has 10 customers, each of them makes 10% of the enterprise customers, second line will show cumulative of 2 customers which make 20 cumulative percent, etc. •The Whale curve shows what percent of total number of customers in X axis are able to generate their part of the total enterprise turnover (profit) (plotted in % in Y axis). •The final point of curve means total turnover by all customers • 30 31 “Whale curve” of profit, red line denotes loss Using “Whale curve” •Define visually the areas with same growth, split customers to segments accordingly •Ask questions by analyzing behaviors of segments: what we can do in order to convert “second best” customers to the “best” •How we can convert customers who bring “loss” to “profitable •Do we have different rules and personnel for segments? •We can split cumulative curve to “deciles” as well •Pareto “law” is visible in “Whale curve” at 20% in X axis 32 33 CRM for changing customer indicators •Cross sell- offering additional products, which are compatible to those already bought •Up-sell- improvements of the product already bought •Bundling- complex product /service/subscription •“Churn rate” measurement. No precise methods to define. The goal is to elaborate indicators which could make early prediction of churn •Mass customization- exploring customer choices, segmenting them and offering as most popular of them as standardized solutions for best-fit segments (improves costing, reduces waste and stock) •Using strategic games for capturing rules of behavior (e.g. putting advertisements to Second life game) Proxy – creating cause-effect linked indicators Indicator Measure Weight %. Average income Average of present and forecasted income 20 Change of income Annual change 25 Relationship features Duration of contract 15 Tenure of history Technologic involvement System integration 20 Reporting system Tele-Web Email Parrnership value Contact level 10 Refferal Future value 34 Top 5 customers Ranking by „proxy“ Ranking by monetary value Rank difference A 1 1 0 B 2 22 +20 C 3 62 +59 D 4 4 0 E 5 3 -2 Customer portrait •Analytical aspects: •1 The percentage difference of each characteristics of the customer compared to the best value existing in the customer base of the enterprise. •2 customer portrait can be expressed as the area plot of the radar chart. Bigger normalized percentage values of each variable of the customer portrait form larger area plot, which can show, that the particular customer falls among the best customers of the enterprise. •3 possibility of tracking each customer over time by dynamics of each variable and the compound index as well. 35 Source:Kriksciuniene et al 2012 Customer index •Customer portrait index is computed as a mean value of all normalized variables included to the customer portrait •If we assume that each variable has different importance we include including weighting of the variables 36 Source:Kriksciuniene et al 2012 37 Web-page based indicators Application of computerized solutions for CRM •www.sugarCRM.com –registers activities related to customers (contacts, commercial offers, negotiations, sales). Analytic tools. System is cloud based, customized •www.microstrategy.com system for intelligent analysis: aggregation of data, drill-down principle slice-and dice •Campain management- dynamic workflow based solution by microstrategy – provides wizard based, responsibility- based process management analytic support • 38 39 CRM campaign research (Microstrategy) 1.Sales situation is evaluated 2.Loyalty level is evaluated 3.Problem is explored in detail (see the following example of wrongly selected promotion delivery channels (pre-campaign analysis) 4.Campaign is planned, the target group is selected by analytics 5.Post – campaign analysis 40 Example :microstrategy campaign analysis workflow 41 Loyalty Guide Step 2 Customer Profitability This step gives the Director the opportunity to understand how the company’s customers affect overall performance indicators through their buying patterns. The information in this report is again grouped by regular and preferred customers, and the Director finds the figures consistent with the information obtained in previous reports. The Director can see that while profit margins are similar between loyalty cardholders and non-cardholders, average revenue per customer is much higher for cardholders. The loyalty cardholder is a better customer and the director will need to incorporate this into plans. Also, the Director may conclude from the report that non-cardholding MegaStore customers may buy less profitable products or more products that are on sale. The Director clicks on the Step 3 of the Loyalty Analysis Guide. 42 At the referring domain level things look a little bit different. Now the Director can determine which domains didn’t work for the month of August. Two domains get the Director’s attention: euroseek.net with an ROI of -100% (which means that there was no profit at all), and altavista.com with an ROI of -95.22%. It is obvious that while some domains like gotto and lycos are working very well (with high ROI’s), the rest are negative, which is driving the average to be negative. This shows that the promotion itself is not the problem, but perhaps the site selection could be improved. Having run 3 separate ads within this promotion, the Marketing Director is interested in finding out if any ads were more successful than others. The Director has a particular interest in altavista.com. Therefore, the Director selects that row of the report and drills within the same dimension to the Ad Banner level. 43 In order to get a clearer picture of the products most commonly combined with Pop music, the Analyst sorts by attachment rate in descending order. Note the strong analytic capabilities demonstrated by the calculation of attachment rate. 44 Wanting a more precise look at the figures, the Analyst changes the report to grid format. What the manager is looking at now is a list of customers that are between 21 and 30 years old and make between $30-$45K per year. The analyst decides that the promotion should be sent to this group, and in order to do so, a mailing list is necessary with all the pertinent information. With that in mind, the analyst clicks on the templates tab... 45 Specialized CRM systems and integrated solutions •Siebel •SAP •Oracle •Baan •Peoplesoft •Microstrategy •Microsoft CRM •Microsoft Attain •Microsoft Axapta •SAS •Remedy •Goldmine • • 46 Axapta CRM main menu • • • • • • 47 AXAPTA CRM modules •(Sales Force Automation): •Business relationships information (communication with sales partners, customers •Sales orders and quotes information •Preparing customer data according to their relationships to business areas, price quoted, processes or contacts •Managing correspondence of business relationships •Contact management- task management, their status 48 Business relationship definition 49 Business relationships •It is the centralized place where you can find any information registered in Axapta about any business relationships filtering by any instances describing them: contact data, persons, etc. •Business relationship window has three parts: •Filter window •Main area •List area •Menu button areas 50 AXAPTA CRM module functions •Sales Management –information for effective management •Graphic evaluation of marketing personnel members •The ordering probability is computed, time forecast of probable orders is designed •Real results are compared to forecasted 51 Sales Quote window 52 CRM quote preparation and transfer to orders •Initial contact •Confirmation of interest •General evaluation •Sending offer •Preparation for negotiation in written or meeting •Setting status: (“in progress”) allows editing. “Final” status includes archiving and non-editing modes •As soon as relationship status meets order, the functional button allows to design order for sales 53 AXAPTA CRM modules •Marketing automatisation, telemarketing •Synchronization to Microsoft® Outlook® about contacts and meetings •Each marketing worker has the individual workbook where his activities are registered •Not complex business prospects evaluation tool •Mailing list creation, Microsoft® Word bookmarks 54 Managing correspondence 55 SAP integrated system: CRM module •Module is composed of various functional blocks. 56 Analytic scenarios •Analytic scenarios- multi purpose analysis •Customer analysis – value analysis per customer •Product analysis – observation of product, promotion optimization •Communication channels – analysis of regular and e-channels •Marketing analysis- allows to select new markets. Cross-sell scenario design •Sales analysis – extensive reports “win or lose” analysis for competitive evaluation •Customer oriented business management by differentiating approaches to customers 57 Structure of analytics scenario 58 “Best practice” application in SAP •The analytic scenarios idea is to evaluate them at all enterprises which implemented SAP solutions. Successful scenarios are standardized and implemented. Benchmark of scenario effectiveness is provided CRM tasks related to social media analytics- new source for deriving value indicators •Development from Web1.0 to Web4.0 59 Social networks: nine most popular (2010) 60 The Web 2.0 characteristics: Social Media, and Industry Disruptors •The ability to tap into the collective intelligence of users •Data is made available in new or never-intended ways •Relies on user-generated and user-controlled content and data •Lightweight programming techniques and tools let nearly anyone act as a Web site developer •The virtual elimination of software-upgrade cycles makes everything a perpetual beta or work-in-progress and allows rapid prototyping 61 62 63 Online Social Networking: Basics and Examples •New Business Models •social network analysis (SNA software) • The mapping and measuring of relationships and information flows among people, groups, organizations, computers, and other information- or knowledge-processing entities. •The nodes in the network are the people and groups, whereas the links show relationships or flows between the nodes. SNAs provide both visual and mathematical analyses of relationships 64 Business and Enterprise Social Networks •social marketplace • The term is derived from the combination of social networking and marketplace. An online community that harnesses the power of one’s social networks for the introduction, buying, and selling of products, services, and resources, including one’s own creations. Also may refer to a structure that resembles a social network but is focused on individual members 65 Commercial Aspects of Web 2.0 and Social Networking Applications •Consumers can provide feedback on the design of proposed or existing products etc. •Word-of-mouth (viral marketing) is free advertising •Increased Web site traffic brings more ad dollars •Increased sales can come from techniques based on personal preferences such as collaborative filtering 66 CRM in virtual community • 67 Web 2.0 data types •Rating •Tagging •Forum content •Blog •E-newsletter •Video materials •Competitions •Search engine analysis •Shopping in social networks •Feedback from customers: conversational marketing • • • 68 Advertising using social networks, blogs •Viral (Word-of-Mouth) Marketing done by bloggers •Classified Ads, Job Listings, and Recruitment •Special Advertising Campaigns •Mobile Advertising 69 The Future: Web 3.0 And Web 4.0 •Web 3.0: A term used to describe the future of the www. It consists of the creation of high-quality content and services produced by gifted individuals using Web 2.0 technology as an enabling platform •Semantic Web: An evolving extension of the Web in which Web content can be expressed not only in natural language, but also in a form that can be understood, interpreted, and used by intelligent computer software agents, permitting them to find, share, and integrate information more easily •Web 4.0: It is still an unknown entity. However, it is envisioned as being based on islands of intelligence and as being ubiquitous • • • 70 Blog record preparation for analysis 71 Internet pages analysis: data preparation 72 Hot to use ontologies •By interlinking information from various sources, it is possible to define if “the person knows book author” 73 Links among individuals and their types 74 Internet query analytics •Grouping by topics •Defining group sizes •Detailed information of the query success •The suitable formats and algorithms for queries can be designed 75 Conversion analysis •User decision making process, affected by social networks: •Likes •Impressions •Friends impressions •Clicked •Share •Comments •Total fans •Young Ae Kim; Srivastava, J. (2007) Impact of Social Influence in E-Commerce Decision Making • 76 Conversion analysis •The conversion rate (Z-axis) is affected by the likes (X-axis) and Clicks (Y-axis). •The correlation among the indicators for this case is 0,98. However each business case tend to be unique and should be explored by the enterprise in long term for its customer base 77 How it spreads when in need: Katrina PeopleFinder •Hurricane 2005 • •1.1 M people were on search • •Blogger initiative for search • •PeopleFinder Information Format PFIF system was implemented during 24 hrs 78 Peoplefinder query sample • 79 Project scope and data management problems •7,000 records on Sunday. 50,000 records on Monday evening •4000 volunteers •Total 640,000 records •ShelterFinder – other project where all shelters for people were registered •Katrina PeopleFinder project data was passed to Google and used together with American Red Cross and Microsoft for finding people •Project is now closed for preserving sensitive data •The processes can be transferred from non-profit to the commercial area for analysis of referral information 80 Assignment 2 •Tools &software: •MS Excel pivot module, •Statistica advanced models, •Viscovery SoMine •2nd team assignment and lab work training •Task : •The data file for analysis CRM_data_for_analysis.xls •The task description is in file Assignment2_CRM_Analysis.pdf 81 81 Assignment 2 – Task description •The data file for analysis CRM_data_for_analysis.xls •The task description is in file Assignment2_CRM_Analysis.pdf 82 Literature •Berry, M.,J.A., Linoff, G.S. (2011), "Data Mining Techniques: For Marketing, Sales, and Customer Relationship Management", (3rd ed.), Indianapolis: Wiley Publishing, Inc. •(Electronic Version): StatSoft, Inc. (2012). Electronic Statistics Textbook. Tulsa, OK: StatSoft. WEB: http://www.statsoft.com/textbook/ •(Printed Version): Hill, T. & Lewicki, P. (2007). STATISTICS: Methods and Applications. StatSoft, Tulsa, OK. •Sugar CRM Implementation http://www.optimuscrm.com/index.php?lang=en •Statsoft: the creators of Statistica http://www.statsoft.com •Viscovery Somine http://www.viscovery.net/ 83