The Four Legs of Retail Applications

Commentary by Bill Bittner, President, BWH Consulting


One of the nice things about attending conferences like UConnect is that they give you an opportunity to reflect on the technology requirements of a particular business model.
UConnect is conducted by GS1-US, formerly known as the Uniform Code Council. They maintain standards for Identification (Bar Codes and RFID), Communications (EDI), and Business
Processes (Rosetta Net and VICS). Data Synchronization was a big theme at this conference and it inspired this thought.


Retail Applications, especially Consumer Package Goods Applications, are built on a foundation that consists of four legs.



  • Software provides the applications themselves and the control mechanisms necessary to interface the applications with each other, the hardware, and the internet.

  • Hardware or Infrastructure is the second leg, which provides the processing power, storage, and physical connections that interface a particular installation with the outside
    world.

  • Leg three is the transaction data that is generated as the business operates. These are the POS transactions, time card entries, and inventory movement transactions that are
    constantly collected by the various applications.

  • The last leg is the “Master Data”.


Master Data is really what Data Synchronization is all about. Everyone realizes that it does not make sense for hundreds of retailers to measure a can of Del Monte Peas when
they are all the same. Instead, the manufacturer should tell everyone in the supply chain the measurements. The challenge is that this is the classic “necessary vs. sufficient”
situation they taught in elementary logic classes.


The data provided by the manufacturers is necessary, but additional data about the item is also necessary to have sufficient information to sell it. This includes POS descriptions
unique to hardware vendor, tax codes, food stamps, etc. The answer up to now has been for individual retailers to build proprietary portals for capturing the additional data.
Manufacturers are forced to log into various systems to complete the new item data.


Moderator’s Comment: Is there an alternative to having each retailer define their own portal? Is there a way the necessary retail data could be captured
once and shared?


Not all master data is the same. Personnel data is unique to the individuals employed by a certain company. But for retail, a large amount of master data
is the same across organizations. Except for private label merchandise, retailers could share all the same data about consumer package goods. A service that provided this information
would be a big help to the small and medium size retailers because it would save them the effort involved with gathering, verifying, and entering master item data. It might even
make more applications feasible for these organizations. I really think there is an opportunity here.

Bill Bittner – Moderator


BrainTrust

Discussion Questions

Poll

9 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
James Tenser
James Tenser
17 years ago

Maintenance of a digital master product catalog for consumer packaged goods might be a good start. This is not really a brand new idea. The discussion was well underway as the dot-com era commenced. It led to such important considerations as:

What data elements must be included in a complete item record? For a packaged food product up to several hundred fields might be required, to accommodate ingredient and nutrition information, as well as package dimensions and UPC codes. For an item of apparel, color, size and style must be folded in.

What product imagery should be created and to what specifications? For a multi-channel retailer there might be several sizes and image resolutions for each SKU. Then there are esthetic considerations for some products, such as the angle of the lighting or whether an item of apparel can be shown on a model or off.

How to deal with churn in the assortment that may exceed 50% per year. Every thing from seasonality to package changes must be accommodated, on a continuous basis.

There are several firms who have already made a business out of creating and distributing product imagery that may be used online or for ad production. But to my knowledge there is no master product database so far. The challenge: The solution that meets every retailer’s requirements would be sub-optimal for all.

Ken Kubat
Ken Kubat
17 years ago

After sizing up the data management monster from multiple perspectives over the past 20 or so years (including a dozen years at a syndicated data company), I’m more certain than ever that the fundamental challenges here are alignment, integration, AND (to ‘Friar Mathews’ point) … TRUST! Alignment, at a core level, is the art of identifying (and promoting!) common interests among parties, whether those “parties” are internal (e.g. departments within a company) or external (i.e. enterprise trading partners). Without the leverage/synergy of clearly defined common interests, it’s real tough to build a defensible business case for “master” data management … AND it gets tougher. The task of implementing master data management is, fundamentally, an integration activity … and in the profound words of an esteemed Enterprise Architect, “Integration is hard!”

Finally, as Ryan stated earlier, “As to data sharing, the question isn’t can it be done, it’s will it be done?” Whether it’s “master data” or “transactional data,” benefits from multiple party use can ONLY be achieved through collaboration and sharing, and THAT requires trust … and “trust” may be the toughest challenge of all! As a whole, we have NOT been a “trusting industry” (as, say, compared to the Financial Services industry). As an incurable optimist, however, I prefer to look on the bright side … noting not only the significant strides taken in a variety of collaborative activities, but also shining a light on the incredible upside of greater trust and sharing. Just as the end of the Cold War generated a substantial “peace dividend,” I’m convinced that our industry will benefit similarly from a “trust dividend” as companies commit to and participate in the process of collaborative information management.

Mark Burr
Mark Burr
17 years ago

It’s interesting that the poll results and the comments seem quite contradictory…

Sure, sounds good, doesn’t it? We all have common data and we all share it the same way for the benefit of everyone – sounds like data socialism to me. But seriously…

Creating a common data model and standardization throughout one organization’s chain of systems is difficult enough (maybe not impossible). Even following the concept of a single standard doesn’t solve the equation. Within single standards, the definition of data elements and the broad use in implementation renders even a single standard questionable if it allows for any flexibility in application.

With all the intricacies of interpretation, implementation and application it certainly lends one to believe that the trust factor is nearly impossible. This is one area where good intentions may be an exercise in futility.

Camille P. Schuster, Ph.D.
Camille P. Schuster, Ph.D.
17 years ago

Without common data standards, synchronization of any kind is impossible. With common data standards, synchronization is possible, but problematic. The traditional competitive business model is based on holding information close not sharing information. As companies have moved toward creating seamless supply chains, the basic assumption of guarding information has to be replaced. Developing trust across business partners is difficult; throughout an industry improbable. There is no guideline or blueprint for how to be collaborative and competitive at the same time. Sharing information is fundamental to this dilemma. Until the dilemma is resolved it is unlikely that there will ever by industry-wide synchronization.

W. Frank Dell II, CMC
W. Frank Dell II, CMC
17 years ago

I will repeat again, Data Sync is great in theory but a failure in the real world. It will never be more than big retailer appeasement. A point often overlooked is that there are other software solutions. Using common data definitions, one data stream from a supplier can be accepted into each retailer’s unique software system. All one has to do is gather data from different manufacturing plants for the same item to see the differences. Further complicating the issues is the frequent package changes. Even using the same UPC code there will be differences in the distribution center. Data Sync will not solve this problem.

Ron Margulis
Ron Margulis
17 years ago

Frank is spot on, although it is important to note that master data management (MDM) is now all the rage in certain IT circles. This is because companies are moving to automate as many of their business processes as possible, resulting in the need for further tech sophistication to integrate all the applications. Both retailers and wholesalers can benefit from MDM because they have so many applications in multiple departments. In addition, with mergers and acquisitions on the rise in the industry, newly combined companies need MDM for data integration. The challenge is to create a common method for users to access information from all of the applications deployed while maintaining accurate (non-corrupted) master data.

Mark Lilien
Mark Lilien
17 years ago

More and more retailers use the same software packages for their supply chain functions, so standardizing data requirements shouldn’t be difficult. Any large CPG manufacturer could do it by surveying their customers’ requests.

Ryan Mathews
Ryan Mathews
17 years ago

Interesting that the four applications ASSUME the existence and acceptance of standards. Hmmm…. As to data sharing, the question isn’t can it be done, it’s will it be done?

Ted Gladson
Ted Gladson
17 years ago

It was interesting for me to read that the industry now recognizes the value of accurate data and master file information. Accurate master file data is what Gladson has been recording for years. Anyone who’s tried to implement a planogram at store level with bad dimensions or master file information knows it’s only a matter of time until the phone starts ringing off the hook with complaints from the stores that products don’t fit on the shelf.

You can talk all you want about software and hardware and infrastructure, but if the data is bad, nothing else matters. It truly becomes a garbage in, garbage out scenario.