It is amazing that thousands of companies have tried to implement global data synchronization yet studies show that the data is still as bad after the project as it was before – perhaps even worse.  One retailer employed a consulting firm to audit the products and data they receive for inaccuracies and they kept multiple consultants busy (perhaps they still are busy) doing this work.  Why would companies continue to pursue stakeholder and customer limiting technologies and practices?

Perhaps retailers and their suppliers feel “in for a penny, in for a pound” is the appropriate approach when it comes to global data synchronization.  My colleague, Melanie Ligons, even puts it as “in for a pound, in for a ton”.  After all, the standards groups will continue to work to make things better, right?  Well, yes.  But how long will it take?  There are things companies can do today to address their real problem – bad data – and not wait for the standards groups.  Another colleague, Steve Keifer, recently wrote (Consumers - Not Retailers - will drive adoption of Data Synchronization) that SAP, Oracle and others have built significant master data management functionality into their solutions.  Of course, it takes a significant investment – both in terms of money and in human resources – to effect the cultural, process and technical change necessary to implement such functionality.  Yet it takes more effort still to modify corporate behavior to want to fix this problem.  The solution may be more costly than the ongoing problems caused by bad data.

It might be, however, that it only looks more costly.  Ongoing analysis at some companies reveal that fixing the bad data problems they face (and almost all companies are facing them) will have substantial positive impact on their businesses.  One electronics company makes between 300 and 400 million parts per day. These are created and shipped to customers to be put into products.  Imagine if you have wrong information – about the components themselves or about the orders for those components.  The mistakes could be costly to both the electronics company and it customers.  Thus, they have embarked on a strategic data management plan that includes technical, process and cultural change.

Still, this company and many others might benefit as much from an externally resident data firewall that protects the company from both bad parts data and inaccurate supply chain data – and does so for both inbound and outbound information.  The electronics company mentioned above has analyzed its internal information and found that, depending on systems and product lines, between 35% and 50% of its data has errors, missing attributes or mismatches.  In addition, they estimate that 50% of their data comes from outside their business.  When this much data originates outside the enterprise, a business is depending on its partners to get the data right, so that the enterprise’s business decisions are based on both complete and accurate information.  Yet I’ve not heard of many companies that perform data audits on their suppliers.  If you audit them for the quality of components and resources, wouldn’t you want to make sure the information about the components – and the information used in developing/producing the components – was accurate, too?

Companies that utilize GXS solutions that provide product and transaction data quality do just that.  They make sure that product information is accurate BEFORE they let it into their businesses.  And they make sure their business transactions adhere to their strict business rules.  This helps make sure that not only the data is accurate, but shipments are more likely to be accurate as well.  And to be on the safe side, companies use these tools for outbound data to make sure their customers and, in the case of retail, their consumers are getting the right information and products.

I usually don’t talk about GXS products in my blog, but felt it was important to do so after participating in the Electronics Industry Data Exchange Association’s (www.eidx.org) spring conference last week.  During the conference it became clear that at least the high tech companies understand the magnitude of the challenge of bad data – and that changing formats from EDI to XML as the retail industry did – is not the answer.  Catalog data format is irrelevant if the data is bad.  They are looking for solutions and companies like Oracle, SAP, GXS and others presented how their solutions work to solve the real business problem of bad data.  In the retail industry, companies like Best Buy in the US and 3663 in the UK have embraced data quality solutions as a principle aspect of business transformation that will ensure their customers have the right information and the right products.  These are the pioneers, though.  Most retailers are relying on global data synchronization to share information and magically fix their data, too.

Yet how can it?  When most users of global data synchronization are inputting data manually, using spreadsheets or web forms that provide little to no data quality validation, retailers are receiving information rife with keystroke entry errors.  And those suppliers that do automate global data synchronization (so that their data is truly synchronized with their customers – a novel idea!) are rarely embarking on data governance and stewardship programs to make sure the data they synchronize is accurate.

If companies think that data sync alone will fix data problems, then neither they nor the consumers will win.  Quite frankly, consumers are demanding more and accurate data, not synchronized data.  In fact, consumers could care less about the technology associated with meeting their needs.  What they do care about is whether the dimensional information about that flat panel TV can be relied on.  They don’t want to have to take it back.  Consumers want to make sure the food they purchase won’t cause an allergic reaction.  If the data they receive is synchronized, but inaccurate, it is worse than not getting information at all.  To quote Barry Bernstein of SonyEriccson with regards to data: “Missing is easy, wrong is difficult.” If consumers are provided with bad data now, then they will be trained to disregard future information – even information that might be right. 

While some might argue that the global data synchronization standards allow for more information to be shared – including information that was difficult to share with older standards – they will also agree that most companies that are doing global data synchronization today are doing so for only a limited number of mandatory attributes – dimensions, weights, GTIN and such.  Their arguments that GDSN was necessary to allow for all the other information are spurious, at best, and hurtful to global retail business at worst.  GDSN took everyone’s eye off the ball.

In order to get the industry behind GDSN, a variety of organizations, like A.T. Kearney, did research that showed the potential value of GDSN to participant companies.  Yet many of the benefits mentioned ASSUMED accurate data as part of the result of global data synchronization.  In fact, accuracy was the primary result that was expected.  Knowing that synchronized data is still bad, then, there is little to support the business benefits found in all those studies.  Even a new product introduction process that takes less time isn’t valuable if the data is inaccurate.

Data Sync is a nice to have.  Data Quality is a need to have.  Once retailers and their suppliers realize that the two are not the same thing, the need will have to be addressed.  The high tech industry already gets it and is starting with data quality initiatives first. Perhaps retail can learn from them.