Make every interaction count

Published on November 30, 2010 in Uncategorized by Donna Prestwood. 0 Comments Tags: , , , , , .

Following up on my prior post regarding Customer Intimacy, a few articles passed my desk this week about organizations that were looking at customer interactions in new ways.

My colleague Cynthia Williams shared a story about how one of her clients was rethinking marketing ROI and placing a greater emphasis on business analysis.

Likewise, executives from US Bancorp were also looking to rethink how they approached marketing when they made the switch from response modeling to uplift modeling.  This article from American Banker describes how they “stopped wasting marketing dollars on people who were naturally inclined to buy us.”  With this one change, they improved year-over-year results by 174%.

Going beyond pure marketing initiatives, some companies are even rethinking the roles and responsibilities of different functional groups.  This research report, for example, details how large business-to-consumer organizations are shifting the focus of their call centers to become revenue generators rather than simply answering customer questions or executing transaction requests.

What’s clearer than ever is that every phone call, every email and every Web experience is an opportunity to impress your customers… or disappoint them.  Today’s top organizations are finding innovative ways to automate insight-driven dialogues with their customers.   With the right technologies, you too can ensure that every customer interaction is focused on growing long-term customer relationship, value and profitability.

If that sounds interesting, you can start by learning more about the Portrait Customer Interaction Suite. 

Part of Pitney Bowes Business Insight, Portrait offers industry-leading solutions for customer analytics and both inbound and outbound customer communications—marketing solutions that rank high on three Gartner Magic Quadrants and two Forrester Waves.

With the right tools, you can make every interaction count. 

Customer intimacy

Published on November 2, 2010 in Data Quality, Data Stewardship and data integration by Donna Prestwood. 0 Comments Tags: , , , .

A roadmap to really knowing your customers

The idea of “customer intimacy” is not a new one. In fact, it headlined a Harvard Business Review article back in 1993. Yet this term is now seeing resurgence across the blogosphere.

So, what exactly is “customer intimacy?” – Why is it important now?  – And how does one achieve true intimacy with one’s customers?

► What is “customer intimacy”?

Customer intimacy is the state of knowing your customers really well in ways that can help you serve them better. It establishes a “virtuous circle”: you know them better, you serve them better, they like you better, they use you more, you learn more about them, you know them better still, you serve them better…and on it goes.

► Why is it so important now? 

These are tough economic times.  It’s cheaper to keep customers and deepen relationships than it is to win new ones – and the best way to deepen relationships is to establish and maintain customer intimacy, to really know your customers and meet/anticipate their needs. What’s more, when you can better define and target your best prospects for new customers and customer segments, your marketing dollars can be better spent.

These are fast-paced times. Think about how much you can learn in a few short minutes through a web search.  Customers know it’s just that easy to get information in the public domain – and they now fully expect, being the important people they are to your business, that a) it’s that easy for you to know everything about them, and b) you really should.  They expect you will have a single, unified view of them as your customer regardless of the numbers and variety of departments and/or channels of distribution they contact you through.

► How do you achieve it? 

Step #1 Better Data:  Customer intimacy is a discipline.  It requires actionable information; and, to get there, data integration, data quality and location intelligence all have to be part of the mix.

Data Integration: You need to pull together all the different, disparate sources of data you have on each customer from all across your organization.  There are so many sources to draw from: for example, you’ll want to include your own internal records from places like:

  • Sales and purchasing (catalog, online, telephone)
  • Service (repairs and replacements; complaints and compliments)
  • Finance (credit/payment history, account activity, current contact information – mailing/delivery address/phone/email)

Data Quality:  As a part of this integration effort, you will need to clean, correct, normalize, match and remove all the duplication in your data too. This can be done regularly and routinely on a spot basis, or managed as an ongoing process. Either way, it is important to recognize a) that data deteriorates over time – people move, preferences change, life events take place, purchases are made…b) As you integrate and make your data fit for use, you’ll also need to set standards and processes for data collection going forward – and governance practices to help ensure that your new data syncs up as it enters your newly cleaned database. 

Data enrichment:  As you’re integrating and cleaning up your data, it pays to consider the benefit of enhancing your data with information from other sources.  For example, infusing your data with demographic and/or lifestyle information and/or information garnered from social media can provide you with more ways to look at, group, and better understand and serve your customers.

Location Intelligence:  If your data is like most, 70% or more of it will include an address and/or other location component (zip code, area code, etc.)  Geo-coding this data can help operations in many ways from helping to pinpoint the best sites for new-store locations, to increasing the efficiency of your delivery routes, to ensuring that you’re applying the right taxes based on jurisdiction. 

Step #2:  Putting Your Better Data to Work:  Just having great data doesn’t get you to the point of true customer intimacy.  Today’s advanced analytical options provide a range of solutions that can draw from your data the types of insightful, actionable information that will help you to jump start the “virtuous circle” of mutual benefits described above.

There are three key types of analytics that drive customer intimacy:

  • Standard analytics are typically used to help segment your data into useful subgroups.
  • Predictive analytics go a step further – as the name implies, they can help you predict what a customer or customer segment is likely to do, so that you can plan accordingly.
  • Network analytics is the most sophisticated of the three – providing a very complex analysis that seeks out patterns in the relationships between people; between things; and between people and things. 

Of these three, network analytics helps to inform the highest levels of customer intimacy.  It determines degrees of separation, and it provides multi-dimensional insights that add further texture to market segmentation exercises. For example, network analytics can help determine leaders and followers within specific market segments. It is also used very effectively for prevention of fraud and other criminal activity.  

► Want to become more intimate with your customers? 

Future blog posts will delve deeper in the benefits of applying analytics to know and serve your customers better.

You may want to check out this white paper: Managing Data Assets.  

And you can contact us at Pitney Bowes Business Insights today to learn about improving your data and taking advantage of the myriad insights it can provide.

What does it take to model risk?

Published on October 6, 2010 in Data Quality, Geocoding and data integration by Donna Prestwood. 0 Comments Tags: , , , , .

One of my colleagues, Burchard Hillmann-Koster, recently published a white paper that details what insurance companies in Europe must do to comply with the new Solvency II directive.

By setting minimal capital requirements, the EU wants to provide consumers with an adequate level of protection. These capital requirements could total hundreds of millions of dollars or more, so the ability to accurately measure assets and calculate risks can have a significant impact. Solvency II allows insurers to develop and certify their own internal model to calculate the solvency capital requirements—but the effectiveness of these internal models cannot be guaranteed without easy access to high quality, historical and predictive data.

In some ways, the principles behind Solvency II are similar to the Basel II regulations in the banking industry.  In fact, any industry where portfolio and risk management come into play must deal with the same issues and challenges on a day-to-day basis.  So even if you are not in the insurance industry, you may find value in the discoveries and best practices identified in this paper, Five Steps Toward Solvency II and Beyond.

Here are a few samplings from this informative read: 

One: start with high-quality data

Many organizations are not satisfied with their data quality, citing incorrect information, missing or misfield data, duplicated records and inconsistent standards that lead to significant costs, delays and an incomplete understanding of the truth. Actuaries and compliance groups responsible for doing the necessary calculations will need accurate data, which can be delivered though data audits, data cleansing and validation, data matching and consolidation, and data integration.

Two: geocode with confidence

When it comes to assessing risk, location is everything. Geocoding turns addresses into geographical coordinates that can be measured, compared, accumulated and analyzed using location-based analysis. Tools should offer the ability to cleanse, parse, standardize and validate addresses before determining location, which adds confidence to the process. Be wary of “false positives” that can increase the risk for poor decisions.

Three: make the necessary connections through predictive analytics

Combine geocoding with the ability to spatially enrich the data, perform analysis, calculations and predictive analytics. If you can verify that an insured is not in a high-risk area such as a flood zone or hurricane path, your aggregate risk will be lower – as will your solvency capital requirements. 

Four: find ways to integrate multiple functions

Maintaining one platform reduces cost of ownership and can speed up system integration. A single interface also simplifies training and education, and makes it easier for your company to gain the skills and capabilities needed to achieve a competitive advantage.

Five:  add value beyond Solvency II

While there is a place for point-level solutions, organizations may be better served by building and enhancing their overall capabilities in data integration, data quality, geocoding and spatial analysis. These core capabilities can help you reduce solvency capital requirements—but they also add value across your entire operation.

Obviously, the white paper goes into much more depth on these topics.  If risk management is one of your concerns, take a moment to learn more and download Five Steps Toward Solvency II and Beyond.

A Services Approach for Data Quality

Published on September 29, 2010 in Data Governance, Data Quality, Data Stewardship and data integration by Donna Prestwood. 0 Comments Tags: , , , .

Is it true that departments only look after their own needs?

For example, in most organizations the data entry required for order processing is based on the data needs for order processing.  The fact that this data could be accessed later for marketing segmentations, compliance audits or portfolio risk management may or may not come into play while a phone rep is keying in information about a recent order.

David Loshin, President of Knowledge Integrity, discusses this phenomenon in an upcoming webinar scheduled for October 13: Considering a Service Approach for Data Quality. 

Even though data may be suitable for its original purpose, each downstream user is forced to define their own expectations and rules to ensure that this data meets their business needs.  Data quality functions are replicated throughout organizations, which leads to inefficiency and waste.

As an alternative, data quality techniques can be encapsulated as services – and standardizing the tool selection, implementation, reference data integration and data quality rules can reduce overall operating costs while improving the consistency and quality of your data. 

If more predictable data at lower costs sounds like something that can benefit your organization, take a few moments out of your day on Wednesday, October 13 @ 11AM ET.  Register today for this free data quality webinar and learn how you can standardize data quality techniques and rules today—and increase the consistency of your data through a services approach.

Financial data quality

Published on September 15, 2010 in Data Governance, Data Quality and Spectrum Technology Platform by Donna Prestwood. 1 Comment Tags: .

On the blog at CIO.com, we are reminded that the overall quality of organizational information requires an approach to data management that includes a “virtuous cycle of continuous analysis, observation and improvement”.

It comes as no surprise that data quality and data governance rank as the top initiatives among business intelligence professionals.    However, when it comes to financial data quality there are still significant hurdles to overcome.  Some estimate that more than 65% of fixed asset data within US business audit trails is typically misclassified, unrecorded or floating in limbo in corporate financial records—a compliance issue that has become thorny and expensive in light of Sarbanes-Oxley.

Clearly, financial data is different from other data assets such as customer addresses.  Unlike character-based records, tools like “spell check” and data parsing offer little assistance; currency conversion can complicate matters; and speed is critical given the time-sensitive nature of financial reporting.

Data access is hampered by the fact that records are stored on disparate systems, different platforms (or even desktop Excel files.)  Inconsistencies in definitions and labels are likely when ERP, order entry and CRM systems are compared, and the same company may refer to the same product using product codes, part numbers, brand names or abbreviations. 

Ultimately, one’s ability to streamline and automate the quality of financial data begins with an understanding that financial data is just another data domain—once you can deal with the unique attributes of numeric data.  Fortunately, these capabilities exist today.  Here are five things to look for: 

1. Control Totals Validation

Today’s leading-edge solutions can “error-check” totals.  Data quality platforms that were once limited to character-based functions can now handle the complex numeric calculations needed to check and validate these computations. 

2.  Business Rules

Can you tell when 123 should have been 213?  Your data quality tools should incorporate business rules into their logic that compare actual totals to expected ranges and flag items that may be incorrect.  When a service that normally sells for $5,000 reflects $50,000 in revenue, that may be a sign of data entry error, miscoding or a bundled offer.

3.  Normalizing Product Definitions

A product may be listed by a product code in one system and a brand name in another. At times it may be sold as a stand-alone offer, other times as part of a bundle.  Tools should normalize these labels and provide for consistent product definitions that make it easy to aggregate financial data across platforms, divisions and geographies.

4.  Tax Calculations

Many industries are required to assess and remit sales & use, property and payroll taxes—and rates are often based on local tax jurisdictions that cannot be determined by simple ZIP Code look ups.  By assigning jurisdictions at the rooftop level, finance departments can automate tax compliance.

5.  Overall Governance

As data quality is a process, organizations should also set in place guidelines as to how financial data quality will be managed. What issues or gaps exist today?  Who can or should develop rules?  Who can administer changes to rules?  Etc.

At the end of the day, your data must be “fit for use” – and that is especially true when it comes to your financial data.  We invite you to learn more about enterprise data quality today.

GeoConfidence: How’s your location intelligence?

Published on September 10, 2010 in Data Quality and Geocoding by Donna Prestwood. 0 Comments Tags: , .

More and more companies are basing essential business decisions on geocoding – and these decisions run the gamut from where to place a new store site, to how much tax to charge, to designating flood-insurance zones, and more. The fact is, geocoding is becoming a go-to resource for those with high hopes of increasing revenue, improving compliance, reducing expenses, even driving up customer loyalty and satisfaction. 

With geocoding though, as with all data-based analyses, the old adage “garbage in; garbage out” can still apply.  Basing geocoding – and subsequent predictions and decisions – on incomplete or inaccurate addresses, is like building a house on a fault line. It may be fine for a while. But when it isn’t fine the results can be disastrous.

Here’s a quick take on ways old and new to feel surer about your geo-based decisions:

Do it the old-fashioned way:  Geocoding is automated – and asking an expert is not. But for some decisions, confidence comes from confirming automated results with someone in the know.  Insights an expert brings to the evaluation process can go beyond the data.  However, relying on experts instead of automation is much more time consuming and is often only applicable on a limited case-by-case basis.

Start with better data:  There are an ever-expanding number of tools out there today to help ensure that address data is both accurate and complete, and starting with better data will yield a much more reliable result.  Today’s tools can even append your data with additional information that can drive more informed decision-making. 

Apply GeoConfidence:  A relative newcomer to the geocoding process, GeoConfidence is built on a process called Geographic Determination. This process overlays basic geocoding data with additional data such as Zip+2 or Zip+4 determinations – and it creates a buffer, known as a “confidence surface” around data points.  The importance of this is simple: even when address-quality is high, there are certain degrees of variance in longitude/latitude estimations.  Geographic Determination helps to pinpoint the extent to which variation may come into play by taking each geocoding data point and creating a circle around (or near it) that helps determine how reliable that geocoding point really is.

The further beauty of Geographic Determination is that it is the product of automation.  It’s fast, and it can run through and compile results based a whole collection of data points. Like the so-called “confidence interval” that states a statistic is “accurate to plus or minus X%,” Geographic Determination provides added insight to the quality and range of the geodata upon which decisions are to be made.  And, like good statistical analysis, it also narrows the margin for error, homing in on the smallest possible point wherein an address is likely to be located.

Better data, personal insights, and now, Geographic Determination all contribute to better address-based decision making. To learn about these can have a positive impact on your business, please contact us at Pitney Bowes Business Insight (PBBI). We will help you walk through all the options, and provide you with a true boost to your GeoConfidence.

To learn more about GeoConfidence, try this whitepaper from Pitney Bowes Business Insight – or call us today at 1-800-327-8627.

Upcoming data webinar series

Published on August 11, 2010 in data integration by Donna Prestwood. 0 Comments Tags: , , , , , .

We are excited that David Loshin, President of Knowledge Integrity, will be collaborating with us this fall on a series of webinars and learning opportunities.

 The first event, slated for Wednesday, September 8, covers Data Integration Alternatives—Managing Value and Quality.

With data volume growing at an alarming pace, it has never been more critical for organizations to manage data and maximize the value of this information.  Over the years, most companies have gotten good at creating transactional and operational business applications, but these disparate systems have led to virtual “islands of data.”

On the other hand, there is a growing list of enterprise applications (such as business intelligence and data warehousing, customer relationship management, enterprise resource planning, and even more complex approaches to business analytics) that require access to data sets from a variety of sources.

 In this environment, data centralization has become critical.  In this upcoming webinar, we will discuss the options and alternatives available to you today that can help you become a master in data sharing. Topics to be covered include:

  • Exploring Data Integration Alternatives – from traditional ETL (Extract Transform Load) to Data Virtualization
  • Understanding Data Integration Challenges – including completeness, consistency and reasonableness
  • Adding Value to Data Integration – using governed data quality services

This informative webinar event is scheduled for Wednesday, September 8 @ 11AM ET (15:00 UTC/GMT) and is offered by Pitney Bowes Business Insight at no cost to you. Please take a moment to register today.

Best Practices in Geocoding

Published on August 11, 2010 in Data Quality and Geocoding by Donna Prestwood. 2 Comments Tags: , , , , , , .

As more than 70% of all business records include a location component, it is not surprising that location accuracy has become such an important part of data quality. Today, organizations are using location data to administer market analysis, risk assessment, effective targeting, network investments, site selection and portfolio management.

Before you can analyze, extrapolate or profit from location data, you first need to associate each record with an accurate latitude and longitude coordinate. That’s why so many organizations employ geocoding. Geocodes translate common reference points, such as customer addresses, into latitude and longitude coordinates that makes it easier to analyze data.  If your geocode is wrong, however, your analytics are wrong, your insights are wrong—and your decisions are wrong—so it pays to be accurate.  Today’s best practices include: 

1. Validate source addresses

Geocoding tools should offer the ability to cleanse data, standardize addresses and validate that source addresses are correct before applying geocodes.

2. Validate geocode results

Accuracy has another element, positional accuracy, which measures how close the geocode is to the reference point. Geocoding an address to the center of a city, for example, will be less positionally accurate than one centered on a precise parcel or rooftop.  Today’s leading solutions provide a ‘geo-confidence index’ that estimates the probability that the latitude and longitude assigned correspond to the place intended. 

3.   Utilize precise, up-to-date reference data

How often you update your reference data is important, as reference points such as roads, addresses and developments are always being added and modified. Many companies do quite well with quarterly data refreshes.

4.   Geocode to multiple levels of accuracy

There will be times when it is not possible to deliver a geocode centered on a specific address or parcel.  The tools you use should recognize this and apply consistent rules, automatically cascading to the next most-specific point of reference, from address point, to street level, to postal code, city, state, etc.

5.  Combine geocoding and spatial analysis

Ultimately, the goal of any solution is to provide answers, not latitudes and longitudes.  Look for tools that combine geocoding with the ability to perform analysis, calculations and predictive analytics, such as point-in-polygon analysis, closest site analysis and the ability to calculate drive time and distance.

 6.  Integrate into existing workflows

When you can integrate geocode analytics into existing operations and business processes, you can which streamlines workflows, eliminate manual processes and improve decision making. 

 7.   One-stop service

Solutions need to be simple to use and flexible enough to meet different business requirements. A single technology platform that matches up with your overall corporate objectives can help ensure that a consistent standard will be applied in every market.  Likewise, maintaining one platform reduces cost of ownership and can speed up system integration. A single interface also simplifies training and education, and makes it easier for your company to gain the skills and capabilities in Location Intelligence needed to achieve a competitive advantage.

 To help you learn which geocoding solutions are available in your area, Pitney Bowes Business Insight has created a multi-media map.  You can use the interactive map to get detailed information for each country regarding our address correction, geocoding and routing capabilities.

More actionable data

Published on June 8, 2010 in Data Governance, Data Quality, Geocoding and Spectrum Technology Platform by Donna Prestwood. 0 Comments Tags: , , , .

Business users count on high-quality data every moment of every day.  

In practical terms, data needs to be “fit for use”.  That is, you data should add value to your existing operations and help users make better, more accurate decisions in the course of doing business.  For most organizations, ensuring that data is fit for use entails three distinct disciplines. 

  • Traditional data quality: the ability to cleanse, standardize and validate data
  • Location intelligence: the ability to see data in context of location
  • Spatial analysis: the ability to understand the relationship between two or more data points

Today, 70% of all records contain a location element and knowing “where” affects decisions in virtually every aspect of an organization.  Consider:

  • Marketing can gain insights into customer needs and buying habits
  • Insurance underwriters can identify potential risks (such as proximity to a flood zone)
  • Finance teams can assign proper tax jurisdictions
  • Sales managers can develop territories based on true market opportunity
  • Service teams can plan routes to minimize logistics costs
  • Facilities managers can develop optimal network strategies

Location and spatial analysis are so critical, in fact, that organizations are increasingly looking to incorporate geocoding, location intelligence and predictive analytics as part of their core data quality platforms and processes.

This month, Navin Sharma, our Director Global Product Strategy, will share how these market realities play out in the latest release of the Spectrum™ Technology Platform.  This informative webinar will include details on how you can integrate enterprise geocoding, location intelligence and a new geo-confidence model into your core CRM, ERP and legacy systems.  Ideal for data stewards, IT managers and business users, I invite you to register for this Webinar today and get the facts on today’s best practices in data quality, data management and more.

When BI and ETL meet

Published on June 8, 2010 in data integration by Donna Prestwood. 0 Comments Tags: , , , .

Business users and data analysts face two daunting obstacles when it comes to making decisions. First they need to access accurate, timely data.  Then they need to manipulate this information to create meaningful insight.  Increasingly, organizations have found that it’s advantageous when they can tackle both of these challenges at once.

ETL, the short-hand for extract, transform and load, is the backbone of data integration.  This process makes it easy to extract information from multiple sources and transform this data into a consistent, usable format that can then be loaded into a database, data warehouse or data application.  For example, the right tools can read data in its native format from CRM, ERP and legacy systems without needing to write custom programs—and then combine this information to create a 360-degree view of your customers.

BI, or business intelligence, involves analyzing this business data to generate the insights needed to make more effective business decisions.  When you can turn data into meaningful information and useful reports, you empower business users to respond quickly, with confidence.

While ETL has traditionally been an IT function, BI comes to life when managed within the lines of business. Recognizing the need for speed and efficiency, today’s leading-edge technologies combine integration and analytical tools into a single interface that is easy to use and understand.  Designed specifically for business users, now analysts and other decision makers can pull together the necessary data without expensive IT overhead.

Scott Arnett, Product Strategist, shares some details on how Pitney Bowes Business Insight brings together Business Intelligence and ETL in a single platform.  This Webinar will focus on the latest release of Sagent Data Flow, which is built specifically for business users who need to efficiently integrate data from a variety of sources as well as perform analytics.  Take a moment to register today.