The Independent reported that “An estimated 5.2 million pieces of direct marketing promotional material were delivered to deceased people during the festive period. As well as being very upsetting for bereaved relatives, this created enough rubbish to fill 245 dustbins.”
Organizations embark on CRM implementations with an objective to improve organization efficiency and effectiveness in connecting to their customers. However, most of CRM implementations fail to achieve the desired ROI. Among all the reasons highlighted, customer data issues stand out. The more effective CRM implementations that have delivered the desired ROI have made a focused effort on customer data integrity.
The more effective CRM implementations that have delivered the desired ROI have made a focused effort on customer data integrity.
Miller Heiman, a leading sales performance consulting firm, during the launch of 2012 Miller Heiman Sales Best Practices Study announced “Lack of data confidence is a significant issue for companies looking to grow beyond where they are today.” According to Gartner, “The ability to create, maintain and leverage a single, trusted, shareable version of customer master data is increasingly seen as an essential requirement in commercial and noncommercial organizations to support business processes and business decision making.”
Let us look at a simple example in a typical B2B scenario. General Machines sells industrial equipment and service contracts to maintain industrial equipment. Some of their customers buy industrial equipment or only service contracts or both. Below is the customer database, which has the following problems:
- For the first record, wrong combination of City and Zip code has been captured. City/Zip code should have been “Springfield Gardens, NY 11413”. In the absence of address validation software, this has gone unnoticed.
- The Organization in the second record is a sister concern of organization in the third record; both come under Amersson Inc and the organization hierarchy has not captured this.
- The organization in the fourth and fifth record are duplicates
Below are the implications of these customer data problems:
- Arrow Nexis Motors has placed an order for equipment but due to a wrong address the shipment got delivered somewhere else
- General Machines launches a promotional campaign to sell Service Contracts with heavy discounts to win more customers. One of these lands with Amersson Inc and they are disappointed to receive an offer at lesser price than the current payments they have been doing. One of the main reasons for failure of marketing campaigns has been inability to narrow down the target audience.
- General Machines decides to have dedicated account managers by identify big ticket customers (Annual Business > 1000M USD) and make dedicated sales push by offering competitive prices. Welstar though being eligible misses out in the analytics report due to duplicate record.
How to build a good customer database?
Based on my experience, below are customer data best practices that have been adopted in successful CRM implementations:
- Build a single customer master
Having a single customer master is one of the important practices that can lead to greater control on customer data. Many organizations are growing through mergers and acquisitions, and each acquired organization continues to operate with its own customer master in their legacy systems. This leads to a greater possibility of duplicates as each customer master has its own unique customer identifiers and customer keys. This can be avoided by investing in a single customer master.
It may not be possible to completely do away with the divisional customer applications as they might have been already integrated with other downstream applications and eliminating them all of a sudden might affect business operations. In such cases, to ensure business continuity and operational efficiency, the master customer records should be consolidated into a central database controlled by a single administrative application and this can become a feeder to the divisional customer applications till the organization feels comfortable enough to sunset the divisional applications.
- Use third-party validation for addresses and demographics
Customer records created with missing zip codes or wrong address information often leads to customer communication breakdowns. Address validation software can be used in conjunction with the in-house ERP or CRM system to capture the correct addresses. There are easy to use off-the-shelf software solutions available that integrate with the popular ERP and CRM systems.
Internal customer data generally deals with the customer’s transactional behavior with the organization. However for complete customer information; you should also capture the demographic information and the social behavior of the customer. Thus, third-party integration with external database like Dun and Bradstreet becomes desirable as reliable data can be sourced and used for customer analytics and obtaining greater insights into the customer.
- Control data through data governance
Often it is seen that too many people and departments have access to customer creation. Unless processes are standardized, people would have their own ways of creating customer records and this would lead to very inconstant customer database. The implications can be even more severe if inaccurate and inconsistent customer data is fed into the downstream applications. For example: same customer location created in different ways like NJ, New Jersey, NJ state etc. can seriously detriment the account assignment rules or the territory management application.
Thus, it is important to establish control by assigning customer creation responsibility to a specific group. Specific roles like Data Steward or Data Governor can be created that would have administrator privileges to create customer or verify the customer created before it can be used in the IT database for further sales and marketing activities. Libraries can be applied for standardization of key customer fields, example Organization names. This would mean stricter control in customer creation and cleaner data as duplicates and inconsistency could be eliminated at the source itself and most importantly a standardized way of creating customer records.
- Cleanse data with match-merge, de-dupe and purge strategies
Even with all the data controls established, there would be possibilities of duplicates creeping into the system. The same customer appearing twice will blight negotiation pull since the total business volume will not be apparent on the customer dashboard. Thus, it would be essential to establish match-merge process. The match-merge process can be in the form of a program that would be scheduled at a predefined timeframe (e.g. weekly or monthly) that would identify the duplicates. Business rules based on fuzzy logic and/or acronym identification can be incorporated that would identify duplicates and do the match-merge. For example: Child records/entities associated to the duplicates (Victim) will get associated to the existing record (Survivor) after the merge process and victim will be purged. Maintaining established customer identifiers—for example Dun and Bradstreet numbers—help in identifying duplicates and maintenance is easier.
It is also necessary for the organization to establish purge strategies for maintaining active information. Organizations having customer records inactive for last 10 or 15 years stand a little chance of hearing from them again. However, some might argue that even if the customers have been inactive for a decade, would still need to be retained as they represent the voice of the market. In such cases, one of the possible ways to resolve this would be inactivating the customer status or archiving the customer record on less costlier infrastructure. Having a defined archiving strategy would be a step in the right direction.
- Establish customer data monitoring metrics
Customer data gets updated and refreshed regularly hence it is prudent to establish data monitoring business rules. Following are tools and techniques that can be utilized to monitor data:
- Use of workflows and alerts to maintain data quality
- Use of organization specific data security policies
- Maintaining Dashboards to spot data errors
- Use of Exception reports run monthly to find erroneous or incomplete records
- Data Stewards or Governors can do spot check and monitor data on a regular basis
- Establishing measures to cleanse data at the source itself
- Restrict access to customer views
Business rules should be established that would determine who has access to what data. For example: a salesperson has only read-only access of customer financial data, but someone in the finance department can have administrative access to customer credit information. Also, some of the vital customer information like credit card numbers and bank account numbers would need to be encrypted to avoid data theft or misuse. It is often advised to have customer data protection written policies that the customer can access. This creates a greater comfort level for the customer to reveal critical information.
- Balance quality vs. quantity
Sometimes organizations get carried away with their CRM implementations and try to incorporate all the possible data. The question to be asked is: “Is this data required at all”? Often times, a lot of invalidated data is stored but never used. It is sometime appalling to see CRM system adoption increasingly very low even after having spent a lot of money in the implementations. Sales teams complain that in the myriad information available in the CRM system, they don’t know which screens they need to go to seek information they are looking for. Organizations need to a focused exercise on what to put in their CRM systems and ensure that it is not the quantity but quality and usability that will yield better returns.
Way Forward
The industry is buzzing about Master Data Management (MDM) solutions to ensure data consistency. Gartner says MDM is “critical to achieving effective information governance,” and estimates worldwide MDM software revenue will reach $1.9 billion in 2012, a 21 percent increase from 2011.
MDM solutions are being increasing considered as one of the solutions to equip organizations to maintain reliable and controlled customer data, and ensure a consistent customer experience. The organizations can consider the best practices mentioned above while designing their customer management systems.
Concise information and well documented. Just a point to ponder, MDM has been spoken alot these days, however enterprise is finding it difficult to put that too practice. They end up buying the product but transgressing to an MDM meta is indeed a herculean task.Only hope that there are better governing principles elucidated in days to come… Anyways it was indeed a pleasure reading this blog.
The information touches upon a myriad of best practices that many enterprises can mine gold from for their customer B2B and B2C processes and data. Is informative and at the same time puts some relevant questions for consultants to answer and guide their customers.
Subodh, I am interested in validations of: addresses, mobile phones and e-mails before entering the data into CRM systems.
Where could I find deeper information, best practices or business cases about validation practices from an IT perspective and validation services as well?
Thks,