Bill Holt
3 Tips to Avoid Product Data Management Hell

During an online discussion I received the following question from someone trying to bring sanity to their B2C catalogue:

I’m struggling to find any case studies (non system/commercial) on product data best practices.  Any ideas?

I don’t know why there is a lack of useful material on this subject.  It is not an uncommon question for anyone looking to put large amounts of data online for an ecommerce website and feeling daunted by the task.

My best guess is that, because the margins in delivering a cataloging only solution are so low, that most folks are forced to either cobble together their own solution and don’t publicize it, or they resort to one of the larger Product Information Manager solutions.  These PIM tools are feature rich and they bind up your processes and your budget for years.

If you don’t want to go a PIM route, or something similar, you can do it yourself.

The good news is that you can control costs by customizing a solution to the requirements of your business rather than following a bloated process that you don’t need or one that entangles you with a particular vendor for years.  Rely less on “domain expertise” and more on solid process analysis, project management and stakeholder communication. 

My “best-practices” tips for B2C, e-commerce product data aggregation and maintenance are the following:

1. Know the business requirements well and meet the MINIMUM requirements.  

To an outsider, managing product data looks simple. Everyone thinks they are a product data expert, everyone thinks they are a web-functionality expert, everyone can look at a web page and criticize, everyone knows you can throw data into a database or an index and get it back out again. However, the tools and processes supporting data quality (setting up and maintaining) for web based processing are expensive.

A great investment would be to map data flow from reception by the supplier through marketing’s requirements for the customer experience on the website on to finance and the data warehouse if you use one.

For each data field, ask yourself these questions:

  • How critical is the field for the current business?
  • How critical is the field for future business requirements?
  • How easy is it to get the data populated?  Who will populate it?

2. Decentralize product data handling as much as possible.  

Push as much data acquisition and exception handling to suppliers as possible.  More than likely this means MS Excel spreadsheets.  Once you’ve mapped the data above you can build a templates that include critical data fields and list standard field values if you have them.  Eventually a dedicated data acquisition and exception handling tool may be appropriate to make or buy.  If you get to this stage be cautious. Fully understand costs and process impacts.

If you must have folks build catalogues for you make sure it is based from somewhere where labor costs are low.

This step is critical.  Many companies have tried to set up their own catalogue factories which are too expensive to maintain and end up capitulating to an expensive vendor product or service. 

3. Know and consistently communicate the costs (both monetary and time costs) of process and tool change to executives and stake-holders.

Business is constantly evolving.  Your product data processes will change and hopefully improve.  If marketing wants to add a new field, know what it will cost to handle it along the entire data flow.  This will allow the decision makers to fully understand the impacts of change.

Let me know your product data management stories and challenges:

  1. 4billholt posted this