A lot of discussion around master data management (MDM) can be fairly abstract, focusing on technology and different deployment styles, so it is good now and again to remind ourselves of why MDM exists at all – business value.

Retail businesses traditionally got by fine with bricks and mortar stores, but in the late 19th century Montgomery Ward and Sears popularised the mail-order catalogue and so created a new sales channel for retailers.

We had to wait a while for another new channel, the internet, and these days pureplay and traditional retailers embrace it alike, to the point that the internet accounted for 12 per cent of UK sales in 2011. More recently the line between in-store and online shopping has blurred, with more customers are researching purchases online before buying in a store, or browsing in a store and then looking online for the best price for a product.

This shopper behaviour provides both challenges and opportunities for a retail business like Halfords, which has been providing products related to cars, bikes and travel since 1892. The largest of Halfords 460 UK stores stocks 14,000 product lines, but its website lists over 30,000. This means that the website has a significantly extended range, each product requiring factual information, marketing descriptions, images and safety data. A store may stock a single child seat for a car, but there may in fact be several sizes and numerous colour options, so the website needs more information than just what is in the stores.

The issue for a retailer like Halfords is the sheer diversity of the uses of its product data. As well as its web site, it has a core inventory and fulfilment system, a print catalogue system, a labelling application, marketing systems, point of sale systems, packaging applications, kiosks and product quality systems.

Product data often begins with supplier-delivered information, and this is then entered and enhanced, such as adding marketing descriptions to the core product which can be slightly different depending on the channel for which it is intended. Low-resolution photos of a product may be taken and the images used on the web. Images are frequently taken many times because the multiple uses of a photo are not always considered, when it would clearly be more efficient to take all the necessary images at once, and to re‑use existing images where they exist.

All this product data may be shared with third parties such as advertising agencies and public relations firms. It takes an average of 50 minutes to fully on-board a new product in the traditional way, but many weeks to collect all the data required. Any inconsistent product data across all the systems where this data is entered or used could result in all sorts of issues that ultimately cause customer returns and refunds, an expensive end to a disappointed customer transaction.

Clearly it would be better to have a master source of all product data, and allow this to feed all the various systems that need it to ensure consistency. This is what Halfords is now putting in place, with a product information data hub that allows data and images to be entered once and distributed to other systems as needed. The system allows more efficient cloning and mass maintenance of data for products and variants, which has meant that product on-boarding can now be done in just 10 minutes - a massive drop.

Across a growing range of many thousands of products, that adds up. Avoiding the re-keying of data across systems ensures product data consistency, which results in fewer customer issues, reducing returns, and which should lead to greater customer satisfaction.

The six-month project will be implemented in the early summer of 2013, the business case for it being a mixture of the hard savings around data on-boarding processes, as well as softer, but equally important, strategic considerations. A solid platform of consistent product data will make things easier for future systems, such as new labelling or point of sales systems, reducing the costs of implementing these and helping to ensure their success. "Our newly improved product data processes will provide significant strategic advantages across the business and help to deliver our multi-channel ambitions," said Paul Barron, business process consultant at Halfords.

This project is a good example of master data management in practice, showing tangible cost savings as well as strategic business benefits.