|
We are a small company employing 20 people operating in the electronics industry. We regularly receive magazines and direct mail from organisations that clearly have no interest in maintaining an accurate database.
Just today we received six copies of a magazine (remember, we only employ 20 people from the MD to the cleaner). Three copies were addressed to employees who have long since left the company (One in Dec 2000 and two in Sept 2002). One copy was addressed to the company secretary (whilst the name was correct the title was wrong. Plus, this person would have no interest in the contents of the magazine). This leaves just two copies correctly addressed; well almost, as there was a spelling mistake in one of them.
The reason why this situation exists was explained by another recent experience.
Wishing to undertake a direct mail campaign aimed at a section of the industry
we had not previously addressed, we approached a leading electronics magazine
and purchased part of its subscription database. This cost approximately £5k
for a 12-month contract, with regular updates.
When we received the database, we were appalled by the quality of the data. Apart from spelling mistakes, the data had no fixed format. Company names and addresses were abbreviated in a completely random way. For example, “Business Park” would be abbreviated as “Bus Park”, " B’ness Pk”,” Business Pk”, etc. There were literally 100s of examples of this type of sloppy data entry. Another range of entries demonstrating an incredible variety were the many different counties this country seems to have. Now whilst I will admit that over the past few years the government has done its best to confuse the UK population by rearranging county boundaries and inventing new county names, plus using postal districts in many metropolitan areas, there is a definitive list. This list was completely ignored.
This situation gives rise to two major concerns: 1) how on earth can duplicate records be spotted and: 2) given the poor standard of data entry, how can you be confident that any particular record is correct?
We naturally complained about the quality of information received and, after a very long time eventually had a meeting with the Sales Manager from the company which provided the database.
When asked to give an overview of its position in the market, the Sales Manager explained that the company she worked for was one of the largest in the UK, if not the world, and that it provided databases for all types of organisations large and small. In short, considered itself to be the market leader.
We explained our concerns and were absolutely amazed at the reason given for the state of the data. It appears that when information is received, it is entered onto the main database “exactly” as it was received. This means that, not only are you at the mercy of the person providing the data, but also the person reading and entering it; which if the data is received in hand-written form. This adds further to the possibility of errors.
What amazed us was that no ‘data entry rules’ were used. For example, on our own database we always enter “Ltd” for “Limited”, or “LTD”. We have a fixed list of counties and link this to postcodes to ensure we always enter the correct county or post district.
Large companies and organisations are other major data entry hazards. Take BT as a classic example; we use “BT Plc” as the standard entry for a range of names that include, “British Telecommunications PLC”, “British Telecom”, “BT Ltd”, etc.
We also ask all our sales force to keep us informed about name changes. Consider the changes that have taken place within the companies that made up the old Racal group. Some are now names Thales, whilst some were Thales for a short time, only to revert to Racal a few months later. By ensuring we have the latest information we can change whole blocks of data in one single action so that all records have the correct new name.
The Sales Manager fully understood our point and commended us on our rigorous approach to data management. She then told us that with only 7000 names on our database, it was an easy task, whereas it was impossible with a database with in excess of 30,000 names.
I didn’t believe there was a difference, and that if a company was a “market leader” it should demonstrate it and not make excuses.
The Sales Manager said that the standard of data was “as good as it gets” and other organisations do; employ another company to clean the data before use.
One other interesting fact also came out of our discussions. No record is amended unless the company receives instructions to do so. This is typically following confirmation that a magazine subscription is to be renewed. If the person “goes away” then it would appear the record is just left to gather dust. There is a period after which the record would be deleted, but going back to the example given at the beginning of this letter, that would seem to be a very long time.
The above means that if you buy a database and use it to mail a catalogue or brochure, you could be wasting a significant proportion of your budget on printing and postage sending information to people who just don’t live there any more.
About the author
Howard Venning is Managing Director of Aspen Electronics Ltd. For more information please contact Howard.Venning@aspen-electronics.com |
