Baker Goodchild Management

Fuzzy Matching to Prepare Data for Marketing Campaigns

In News by olga

Data quality in marketing campaigns has never been more important than now.

With a choice of marketing channels available and an ever-discerning customer, organisations using poor data will be quickly “found out” and will suffer from reduced ROI and response rates on campaigns.

With so many disparate data sources and relative organisation laxities on data collection errors do regrettably creep in. There is an ever-increasing trend to “fuzzy matching”, which is a part of data cleansing and enhances data more than traditional basic techniques are able to. In the below article we will discuss fuzzy matching in more detail.

How “fuzzy data” errors arise

We all probably have an idea, but here are some examples of how “fuzzy data” errors arise:

  • Misheard data in call centres
    – Call centre staff work in noisy environments
    – Difficulties with accents
    – Phonetic errors – e.g. “Shore and Shaw”, “Pete and Peat” or similar letters e.g. “Salt” entered as “Walt”, etc.
  • Transcribed in error – Manually entered data from coupons, forms, etc. can be transcribed in error if the initial handwriting is difficult to read or in some cases illegible
  • Online form entry errors – Consumers can innocently or deliberately enter data in error when completing forms online
  • Multi-channel data capture – Data can be captured in a multitude of ways, this can introduce problems. Many systems will allow data collation in different ways, with different validation rules. This can introduce errors which are hard to fix later

Fuzzy matching is a part of data cleansing

Fuzzy matching has been around for many years, but thanks to recent software innovations is now better than ever. Fuzzy data is a more advanced part of data cleansing and enables organisations to cleanse data duplicates where records are not exact matches. Fuzzy matching is primarily used as a form of data deduplication but can also be used in application integration projects where organisations are seeking to provide a single version of the truth.

Software tools for fuzzy matching

A variety of software tools are available for fuzzy matching. For direct mail campaigns it’s best to use specialist software focusing on addressing and direct mail fuzzy matching.

Fuzzy matching algorithms can be created either via SQL in databases and/or through off the shelf tools which allow configuration options. Alternatively databases could be sent to a specialist data cleansing and manipulation agency (such as Baker Goodchild), who will be able to complete this work on your behalf.

Software needs to be configured to identify definite “yes” duplicated records whilst “uncertain” records will need specialist manual intervention and will need to be kept in a quarantine area until they are manually examined.

Contacting Baker Goodchild

So if you’d like to work with a marketing agency that specialise in direct mail but are also experts in data manipulation (including fuzzy matching), contact us here at Baker Goodchild. We can manage as much or as little of the data management for your campaign as you would like.