|
Customer Data Quality Issues are partly due to internal gaps and partly due to extrinsic reasons mostly out of an organization's control.
Most of the inputs to an organization are coming from 'Customer' entity (Application forms, invoices, loyalty card membership forms, registration at the web…) and most of its end outputs are going to 'customer' entities (Commission payments, bill payments, mailing and promotion campaign communications etc.). 'Customer' out here is any 'master entity' with name, address, contact details, identification code and more. This includes individuals, purchasers, suppliers, companies, groups and households. Let us see what could happen, if you don’t have clean data:
- Wastage in promotion campaigns: Duplicate customer records.
- Missing out volume deals: Same supplier appearing thrice (say) will impair your negotiation leverage, as the total business is not evident.
- Business risk: Same customer carrying multiple credit cards, resulting in a bigger credit risk not in line with his profile.
- Return mail OR untraceable customers: The addresses and telephone numbers, which don’t exist.
- Analysis failures: Same location appearing in ten different forms (New York, NY, New York State, New-York,…) can make the system treat them as separate locations, spoiling all location based analysis.
- Customer complaints: Same customer getting multiple junk mails. Customers already having the bank account are receiving calls for opening a bank account.
Customer Data Quality is different from the transaction data quality
Transaction data quality is a different subject from the 'Master' customer data quality. Transaction data quality includes the quality of financial and non-financial transactions. Here are the key differences:
Data Quality Controls are more extensive for transaction data vis-a-vis customer data
You can have many process flow controls to have data quality of transactions. You can lot of business rules to make sure that data stays consistent. For example, matching credit and debit balances, reducing finished good inventory levels by as much amount as your ship out the goods etc. The control of the quality of transactional data can rely majorly on the way you design your applications and the kind of controls you place in the same. Customer Data Quality however, depends a lot on the external data inputs that you receive from the system. As you will see that there are many data quality challenges, which are out of your control.
Transactional data quality is more visible vis-a-vis customer data
A wrong bank-statement OR a mismatched inventory levels OR a wrong compensation statement to your agents will be quickly pointed out. Any transaction which has a financial impact are typically caught by the end-users or internal system checks. On the other hand, the customer data quality typically does not have the same level of impact for carrying out the business of an organization. A good bit of customer data may not be having transactional relevance. for example sales leads data. The critical areas related to impact is customer address, which results in the returned mail. For example, you may be having same physical customer appearing in three different customer records. This may increase risk levels (if same customer is holding multiple credit cards under the guide of being different entities), but will not impact the day to day running of business.
Transaction Data Quality gets more focus vis-a-vis data of the customer
This is extension of the previous point. As transactional data impacts day to day running of business and has more direct link to the financial exposure, it gets more focus and budgets to fix. This leads to systems being more responsive. The cost of fixing transactional data in post-facto can be sometimes huge (just imagine, if you have sent the cheques of wrong-amounts to your 10000 sales agents). Therefore, typically organizations go out of the way to fix transactional data.
The above differences does not mean that you don't have to be concerned about the transactional data quality. These differences guide us to adopt different strategies and approaches to address data quality issues. You will find most of the controls mentioned in data quality assurance and Data quality program, to be applicable to both transactional and master (i.e., customer) data. However, this chapter focuses on additional data quality elements which are specific to customer (or in wider sense- any master data like vendor, employee..) data quality. Most of the techniques mentioned in this chapter will not be applicable to the transactional data. |
