Why is Data Quality an Issue?

We have all heard the stories of letters sent to dead customers either though the company has been told about their death or cheques sent out for £0.00. Companies are continuing to risk damaging their relationship with clients and their hard won professional reputation by paying too little attention to the quality and organisation of their customer data.

Data quality is an uphill struggle for most organisations as volumes of data grow so does the proportion of dirty data. The popularity of software systems such as SCM (Supply chain management), ERP (Enterprise Resource Management) and CRM (Customer Relationship Management) has also highlighted these issues. Add into this mix the M&A activity over the last few decades and the cost cutting activities of the global recession and we have a problem. 

A few years ago I experienced this myself with an incorrectly addressed letter from a well known office supplies business. Months previously I’d opened a new business account and the letter that I received (six months later) was my introduction letter with a few money saving vouchers. The problem with the letter was that it containing an incorrect address (in fact the address didn’t even exist as it contained two street lines in it), my name was spelt wrong, plus it containing vouchers that were three months out of date. Obviously I didn’t deal with that company again. The point of this example is not to embarrass the company; hence they are not named, but to make the point that this is a common problem. The larger the organisation the more of a problem it becomes.

Given that it’s a widely held view that the amount of data in organisations will expand a hundredfold over the next five years, companies must increasingly depend on and develop a coherent and cost-effective data quality strategy.

By data quality we mean more than just validating names and addresses or removing record duplication. It is a complete process for defining and enforcing global business rules for data quality.

The Business Case

The impact of data quality issues on an organisation can be difficult to express in a way that can be understood. It’s obviously important to a business to make decisions based on correct information but on the other hand once the costs involved are indicated companies get cold feet.

To give an example of the business case behind these kinds of activities lets look at a simplistic direct marketing example. In this example we are assuming that we have a customer database of 100,000 customers. Of that number 2% have problems with the address. We have a new product which the marketing department want to mail shot to this customer base. If we then assume a typically response ratio of something like 2% and a cost per mail shot piece of £1 we get a breakdown as follows.

(A) Cost of each mail shot piece	£1
(B) The number of pieces mailed per marketing campaign.	100,000
(C) The cost of each marketing campaign. This is calculated as (A * B)	£100,000
(D) Typical conversion rate	2%
(E) Customers with incorrect addresses	2% = 2,000

Based on these numbers and a product that retails at £150 we get a comparison that shows a difference per marketing campaign of £6,000. The breakdown is as follows:

	With incorrect addresses	With addresses fixed
(F) Number of customers that are actually contacted (B – E)	98,000	100,000
(G) Number of customers that purchased the new product (F * D)	1,960	2,000
(H) Revenue from campaign (G * 150)	£294,000	£300,000
(I) Profit per campaign (H – C)	£194,000	£200,000

£6,000 is not a world shattering amount but if we then multiple this over a number of campaigns, and into the other areas of the organisation that use addresses, we can start to get a sensible figure for the impact of incorrect addresses.  Combine this with the other data quality issues that probably exist in the data such as duplicate customers, incorrect customer profiles and we can get some sizable numbers.

The above text is taken from my up and coming revision of ‘The Enterprise Data Model’ scheduled for publication later this year. Currently the book is in draft so changes are likely to this text in the actual book. More information on the current edition can be found at  http://www.koios-associates.com/TheEnterpriseDataModel.html